Low-field MRI as a multiparametric tool for large engineered tissue characterization

Low-field MRI as a multiparametric tool for large engineered tissue characterization | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Low-field MRI as a multiparametric tool for large engineered tissue characterization Christophe Marquette, Yilbert Gimenez, Valernst Gilmus, Elliott Cowles, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8000297/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Advances in 3D bioprinting have enabled the fabrication of large and complex engineered tissues, but their increasing size demands non-invasive tools for monitoring structure, maturation, and perfusion. Magnetic Resonance Imaging (MRI) offers unique multiparametric capabilities, yet high-field systems remain costly and inaccessible for most laboratories. In this study, we evaluate the potential of low-field (LF, 0.3 T) MRI as an affordable and versatile alternative to high-field (HF, 7 T) MRI for characterizing bioprinted tissue constructs. Using standardized PLA and hydrogel scaffolds within a custom-designed perfusion chamber, we compared LF and HF imaging performance for morphology and flow visualization. Both modalities successfully resolved internal scaffold features, with morphometric deviations from reference CAD models remaining within quality control tolerances. Flow imaging demonstrated that LF MRI could capture velocity distributions consistent with HF measurements and computational fluid dynamics simulations, even revealing fabrication-induced defects such as channel collapse or occlusion. Finally, we applied LF MRI for longitudinal monitoring of a perfused adipose tissue construct over 34 days. This approach enabled repeated non-destructive assessments of morphology and perfusion, with final histological analyses confirming homogeneous adipogenic differentiation and extracellular matrix deposition. Together, these results establish LF MRI as a powerful tool for real-time, non-invasive evaluation of biofabricated tissues. By combining affordability, portability, and multiparametric imaging capacity, LF MRI broadens access to advanced monitoring strategies in tissue engineering and regenerative medicine, supporting both quality control and functional assessment of large-scale engineered constructs. Biological sciences/Biotechnology/Tissue engineering Biological sciences/Biotechnology/Regenerative medicine Adipose tissue Bioreactor Engineered tissue Magnetic resonance imaging Velocimetry Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Recent advancements in 3D bioprinting have enabled the fabrication of increasingly complex and large-scale tissue constructs, including vascularized models of organs such as the heart, liver, and kidney 1 . These developments hold significant promise for applications in regenerative medicine, disease modelling, and drug discovery 2 .​ As engineered tissues grow in size and complexity, there is a pressing need for non-invasive, non-destructive imaging methods to monitor their morphology, maturation, and internal functionality over time. Techniques like near-infrared fluorescence imaging have shown potential for tracking transplanted cells within tissue constructs​, and Raman spectroscopy has been employed for in vivo monitoring of tissue development in 3D printed scaffolds 3 , 4 .​Magnetic Resonance Imaging (MRI) stands out as a versatile modality capable of providing comprehensive insights into engineered tissues 5 . Its array of acquisition sequences allows for the extraction of diverse crucial information for tissue engineering (TE), including:​ Morphological details via T1- and T2-weighted imaging 6 ; Tissue content through quantitative relaxation mapping (T1, T2, T2*, proton density fat fraction); Tissue microarchitecture using Diffusion Weighted Imaging (DWI) or Diffusion Tensor Imaging (DTI); Perfusion and internal fluid flow using phase-contrast MRI or Arterial Spin Labelling (ASL); Chemical composition and metabolite detection via Magnetic Resonance Spectroscopy (MRS) or Chemical Exchange Saturation Transfer (CEST). MRS enables the detection of metabolites 7 such as lactate, choline, creatine, and glutamate, which are indicators of cellular activity and health 8 . CEST imaging extends this capability by indirectly detecting low-concentration metabolites and pH changes, enhancing the functional assessment of tissue constructs 9 Monitoring of oxygenation levels using BOLD contrast and T2* mapping 10 .​ Blood Oxygen Level Dependent (BOLD) imaging exploit the paramagnetic properties of deoxyhemoglobin to detect tissue oxygenation 11 while T2 mapping * reflects local variations in oxygen availability, and oxygen-enhanced MRI can dynamically assess oxygen diffusion and uptake 12 .​ Despite these strengths, the widespread adoption of MRI in tissue engineering is limited by the accessibility of high-field systems (≥ 7 T), which are expensive and require specialized infrastructure 13 . In contrast, low-field (0.1 to 1 T) and ultra-low field MRI (earth field to 0.1 T) 14–16 MRI systems have emerged as cost-effective alternatives. The drawback of these systems being a degradation of the signal-to-noise ratio (SNR) when moving from high-field to low-field and then ultra-low field. Commercially available low-field scanners, priced around €100,000, are compact, portable, and require minimal infrastructure 17 .​ One other advantage of working at low field is the possibility to perform continuum of assessment from in vitro to implanted stage with same contrast mechanisms which is not the case when preclinical data are acquired at high field (7T or even 11.7T). While low-field MRI systems inherently have reduced SNR, these limitations are mitigated in the context of engineered tissues. The biofabricated engineered tissues are static, highly hydrated, and can be scanned under optimized conditions. It allows to reduce motion artefacts and permit for extended acquisition times needed to reach higher spatial resolution or quantitative imaging 18 .​ In this study, we present a direct comparison between a high-field (HF: 7 T) and a low-field (LF: 0.3 T) MRI system for imaging bioprinted tissue constructs. Indeed, bioprinting is one of the tissue engineering biofabrication strategies which allow for in vitro fabrication of anatomic scale engineered tissues. Using standardized samples produced via extrusion-based bioprinting, we assess each MRI system's ability to resolve internal structure, characterize morphology, and visualize perfused fluid flow. Our goal is to determine whether low-field MRI can provide sufficient information to support modern biofabrication workflows, potentially democratizing advanced imaging capabilities for non-clinical research.​ As a final validation experiment, we performed the culture of a media perfused bioprinted adipose tissue, demonstrating the potential of LF MRI for longitudinal monitoring of tissue culture. Results Culture Chamber Design and 3D Samples Preparation To demonstrate the feasibility of using MRI for the non-invasive monitoring of tissue-engineered constructs, all imaging experiments were performed within a custom-built tissue culture chamber specifically designed for MRI compatibility (Fig. 1 -A). This chamber allowed for both aseptic perfusion and secure transport of the tissue samples. It was engineered to fit within the bores of both low-field (LF) and high-field (HF) MRI systems. The culture chamber, fabricated entirely via 3D printing using autoclavable Formlabs High Temp V2 resin, featured a cylindrical geometry with a 35 mm external diameter and a length of 185 mm. The central region, where the bioprinted engineered tissue was housed, measured 20 mm in diameter and 80 mm in height. The design incorporated two independent 3 mm fluidic channels for perfused media inlet and outlet. The inlet channel passed through the centre of a piston mechanism, while the outlet channel ran longitudinally along the chamber wall. The inlet was connected to a sprinkler distribution system composed of a 4×4 array of flow channels, intended to ensure homogeneous media perfusion across the engineered tissue. To maintain positioning and sterility, the chamber included triple internal grooves that allowed the piston to lock in place and serve simultaneously as a build platform for the scaffold. The assembly was made watertight by combining nitrile O-rings with 3D-printed threaded caps and fasteners, which applied precise axial pressure. This setup ensured leak-free imaging conditions and simplified post-printing insertion of the engineered tissue. In the present study, three different types of 3D printed sample were placed in the culture chamber. One type composed of polylactic acid (PLA), a rigid thermoplastic commonly used in fused deposition modelling. One type composed of a gelatine-alginate blend hydrogel, chosen for its relevance in soft tissue engineering applications 19 – 21 . One type composed of the previous hydrogel seeded with human adipose-derived mesenchymal stem cells (AD-MSCs). A single STL file was used to print all samples, ensuring strict geometric consistency across experiments. Each samples included a peripheral wall of three concentric layers and a central network of 4×4 channels surrounded by additional truncated-square ducts (Fig. 1 -B-D). The internal geometry was designed to align precisely with the chamber's flow distributor, facilitating controlled perfusion throughout the sample. Morphometric Evaluation of PLA Scaffolds MRI acquisitions were first carried out on PLA scaffolds using both HF (7 T) and LF (0.3 T) systems. Raw images examples are given in Fig. 2 -A, demonstrating comparable contrasts and resolution. For each systems, the imaged volumes were segmented (Fig. 2 -B) and quantitatively compared to the original STL file to assess morphological fidelity (Fig. 2 -C). Additional comparisons were performed directly between the volumes obtained at low and high field. Both MRI systems successfully resolved the internal features of the PLA sample, taking advantage of the material's high intrinsic contrast. The segmentations produced from the HF and LF images showed a high degree of concordance. The mean absolute surface distance (MASD) between the segmentation obtained at Low and High fields was 144 µm, with a root mean square (RMS) deviation of 248 µm. When each segmentation was compared to the reference STL geometry, the high-field MRI yielded a mean absolute deviation of 111 µm and an RMS of 150 µm, while the low-field dataset resulted in slightly higher but comparable values of 157 µm and 265 µm, respectively. These results indicate that the morphological information extracted from low-field MRI was sufficiently accurate for characterizing the rigid polymer scaffolds. Morphometric Evaluation of Hydrogel Scaffolds Hydrogel scaffolds, due to their soft, hydrated nature and lower MRI contrast, presented a more challenging case for imaging. Nevertheless, both HF and LF systems were able to acquire usable images (Fig. 2 -A) from which segmentation was possible (Fig. 2 -B). Morphometric comparison between the low-field segmented volumes and the reference STL model revealed a mean absolute distance of 340 µm and an RMS deviation of 400 µm (Fig. 2 -C). While these values were higher than those observed for PLA, they remain consistent with expectations based on the hydrogel printing process itself. Indeed, hydrogel 3D printing through microextrusion using an 800 µm nozzle is known to yield lower spatial resolution and greater variability than PLA printing using a 400 µm nozzle. This is clearly illustrated in Fig. 2 -D by the presence of unexpected channel (black arrow), reduced channel (red arrow) and deformed channel (blue arrow) in the hydrogel scaffold. The higher deviation observed is therefore likely related to the limitations of the fabrication technique rather than to imaging resolution. As described in literature 22 , among the panel of bioprinting techniques each exhibit very different resolutions. For further studies on MRI morphometric evaluation of bioprinted parts, it would be interesting to challenge this technology with production from more resolutive biofabrication technologies like photo-polymerization which achieve resolutions in the range of 50–100 µm. Across both materials, these results validate the ability of low-field MRI to capture the essential geometrical features of 3D printed samples with, in the case of the hydrogel, a degree of accuracy compatible with tissue engineering quality control requirements. Very interestingly, it was also possible to visualize the scaffold internal geometry using X-ray view of the hydrogel scaffold imaged using the LF system (Fig. 2 -E), accessing then information usually hidden to the operator. Such technology could enhance the definition of biofabrication quality by enabling the measurement of key parameters such as average surface roughness, smallest printable feature size, strand width, printing accuracy, and pore fidelity 22 . LF vs HF toward fluid flow visualisation Following the successful morphological evaluation of both polymeric and hydrogel scaffolds, we next investigated the capacity of low-field and high-field MRI systems to capture internal fluid flow within the same culture setup. These experiments aimed to determine whether low-field imaging, despite its lower spatial and temporal resolution, can provide relevant information regarding flow patterns and perfusion performance in engineered tissue. Results were compared to theoretical simulated flow patterns obtained either thanks to Poiseuille’s law fitting performed on single channels or full experimental set up computational flow dynamic (CFD) simulation. Flow characterization in PLA samples In the rigid polymer PLA samples, two flow rates, namely 7 mL/min and 20 mL/min, were tested to challenge the sensitivity and precision of the MRI systems. This corresponds to a 2.85 (20/7) velocity ratio between the two flow conditions. Clear flow paths were successfully imaged in both conditions, as depicted in Fig. 3 -A images of the central part of the perfusion chamber. From these images, velocities were measured within the 2.5x2.5 mm square channels thus allowing the determination of the velocity voxel population (Fig. 3 -B) and to compare median velocities between HF and LF in each flow conditions. At a perfusion rate of 20 mL/min, median velocities of 5.9 mm/sec and 4.4 mm/sec were measured using LF MRI and HF MRI, respectively. At the lower flow rate of 7 mL/min, median velocities of 1.7 mm/sec and 2.7 mm/sec were measured for LF and HF, respectively. These values were considered satisfactory to validate the comparability of the results obtained using the two MRI platforms. We then compared the ratio between the mean velocities measured at the two flow rates, with the 2.85 theoretical ratio. These ratio were 2.14 and 2.48 for LF and HF, respectively, comparing well with the theoretical ratio and validating that the velocities measured using both MRI systems scaled proportionally with the imposed flow rates. However, a more detailed analysis of the velocity distribution within the channels (Fig. 3 -B) revealed greater dispersion in the measurements obtained with the HF system. For the 20 mL/min flow rate, velocities ranged from 0.5 mm/s to 8.2 mm/s using HF MRI, and from 1.1 mm/sec to 6.9 mm/sec with LF MRI. At 7 mL/min, the HF system showed a velocity spread between 0.4 mm/sec and 3.8 mm/sec, while the LF system yielded a narrower range from 0.7 mm/sec to 2.9 mm/sec. This greater dispersion in HF measurements could be related to the increased sensitivity to local flow heterogeneities but also to ROI positioning error around the channel centre 23 . To complete this comparison, the velocity profiles within channel obtained using LF and HF MRI were plotted and compared to the Poiseuille theoretical velocity distribution (Fig. 3 -C). Thus, the velocity profile was fitted using Poiseuille’s law ( Eq. 1 ) using the three fitting parameters, the maximum velocity (Vmax), the channel’s hydraulic radius (R) and the offset of the maximum velocity. In this case it is theoretically positioned at the centre of the channel. \(\:V\left(x\right)=Vmax\left(1-\frac{{\left(x+offset\right)}^{2}}{{R}^{2}}\right)\) (Eq. 1), x being the position in the transvers plane along the centre of the channel. To find optimal fitting of the Poiseuille equation for our experimental data, the three fitting parameters had to derive from their theoretical values (Table 1 ). Mainly, the hydraulic radius of the channels had to be increased to fit the LF experiment data. In this case, first, the hydraulic radius was increased from 1.25 to 2 mm. This was mainly identified to be related to an overestimated measured velocity at the wall of the channel. This local heterogeneity was caused by the lower signal-to-noise ratio of LF system was originating from lower magnetic field intensity. Such a higher velocity at wall compelled the fitting to a larger channel dimension. Then, the positioning of the channel centre had also to be tuned to be compatible with the XY resolution of both MRI imaging. Indeed, spatial resolution being 195 µm and 250 µm for HF and LF, respectively, the precision of the experimental positioning cannot be better than these dimensions. This positioning was performed thanks to a channel centre offset in an expected ± 185 µm range. By fitting experimental results with optimised Poiseuille equation, we demonstrated that the flow profiles determined using the MRI velocimetry protocol were following a Poiseuille model. This is indicated with R-squares values all above 0.88 and fits presented in Fig. 3 -C. Still, discrepancy were identified on channel radius and centre positioning compared to theoretical values which demonstrated that MRI has intrinsic experimental limitations, mainly related to local heterogeneity and spatial resolution. Table 1 Optimal fitting parameters of the Poiseuille model for the different acquisitions at HF and LF for 7 and 20 ml/min flow. R represents the hydraulic radius of the square channels 24 . Offset represents the deviation from the theoretical position of the channel centre. Vmax represents the maximum velocity. R-square stands for the coefficient of determination. Fitted data R-square R (mm) Offset (µm) Vmax (m/s) HF20 0.88 1.3 + 184 7.6 x 10 − 3 HF7 0.96 1.4 + 63 3.4 x 10 − 3 LF20 0.98 2 + 53 5.6 x 10 − 3 LF7 0.99 2 -182 2.2 x 10 − 3 Theoretical 1 1.25 0 NA To conclude this evaluation, a second strategy was applied which consisted in comparison of the velocities measured using MRI with computational fluid dynamics (CFD) simulations of the perfusion within the culture chamber (Fig. 4 -A). The velocity fields obtained from both MRI systems were successfully segmented to obtain a three-dimensional representation of the flow paths (Fig. 4 -B and C ). When compared to computational fluid dynamics (CFD) simulations performed under laminar flow conditions (Fig. 4 -A), the reconstructed flow trajectories exhibited both spatial and intensity concordance. Three-dimensional reconstructions based on MRI velocimetry, at both 7 and 20 mL/min flow rates, confirmed the excellent overlap between the HF, LF, and CFD-derived flow maps, both in terms of channel positioning and relative velocity magnitudes. Flow characterization in hydrogel scaffolds Flow imaging was also performed in hydrogel scaffolds at a flow rate of 20 mL/min. In such a softer and more deformable material, both HF and LF MRI systems managed to successfully capture the principal perfusion paths within the structure. As previously observed during morphological imaging, the hydrogel scaffolds exhibited structural deformations, including local channel collapse, off-axis deviation, but also the presence of air bubbles (Fig. 5 ). These non-ideal features, which are representative of the practical challenges encountered in engineered tissues biofabrication, were readily visualized with both MRI systems. Deviations from the idealized flow paths seen in PLA were faithfully detected, demonstrating the ability of MRI, even at low field, to monitor real-world imperfections in tissue-engineered constructs. This capacity to identify defects such as structural deformation and air entrapment opens the path to the use of low-field MRI as a quality control tool for the fabrication and validation of functional tissue equivalents. Bioprinted tissue culture longitudinal experiment To further validate the utility of low-field MRI in tissue engineering workflows, we conducted a longitudinal experiment monitoring the development of a perfused, bioprinted adipose tissue. Adipose-derived mesenchymal stem cells (AD-MSCs) were embedded in a bioink at a concentration of 2 million cells per millilitres and bioprinted directly within the perfused culture chamber. The culture protocol included two distinct phases, an initial 7-day cell expansion phase in DMEM growth medium supplemented with 10% foetal bovine serum, followed by a 27-day adipogenic differentiation phase induced with a cocktail of dexamethasone (20 µg/mL), IBMX (0.5 mM), and indomethacin (50 µM). Throughout the experiment, the construct was continuously perfused at a flow rate of 3 mL/min to ensure adequate nutrient and oxygen delivery. Morphological assessments were performed at days-3, 17, 24, and 34 using LF MRI (Fig. 6 -A, B). The acquired images demonstrated excellent signal quality and clear delineation of the bioprinted cell-laden hydrogel within the culture chamber, enabling reliable longitudinal tissue morphology segmentation. Internal channels were visible all along the cultivation enabling to identify potential modification of the nutrient flow path (Fig. 6 -C). In addition, velocimetry measurement at day-3 was performed to visualize culture medium flow within the bioprinted tissue (Fig. 6 -D - F ). Such analysis permit to observe that in real experimental conditions only 7 channels among the 16 expected through scaffold design, were perfused by continuous culture medium flow. The channels perfused are identified in Fig. 6 -D by red labelling. Such experiment represents the first time that real morphology and nutritive flow distribution within porous, biofabricated tissue constructs can be assessed in a non-destructive and longitudinal manner. This finding is particularly impactful, given that most 3D bioprinting strategies rely on precisely designed geometries to replicate in vivo -like tissue characteristics In order to validate the biocompatibility of such a protocol, biofabricated scaffold level of adipose tissue maturation was evaluated through histochemical staining. The results revealed a homogeneous distribution of mature adipocytes throughout the construct, as indicated by Bodipy staining (Fig. 7 -A), alongside with strong collagen I deposition (Fig. 7 -B) and homogeneous cell distribution visualised by adipocyte nucleus staining with DAPI (Fig. 7 -C). These findings confirm not only the successful differentiation of mesenchymal stem cells into adipocytes but also the extensive matrix remodelling within the tissue with collagen I secretion. When analysing the cell distribution using the larger magnifications of the DAPI staining, a clear homogeneous presence of the cell all over the hydrogel was found, proof of the efficient perfusion of the entire tissue, even though some channels were collapsed or clogged (Fig. 6 -D). Similar results were obtained for the Bodipy staining, showing a homogeneous distribution of the differentiated and mature adipocytes. Finally, all together these longitudinal characterisation results and the final tissue properties underscore the value of low-field MRI to monitor dynamic tissue development in situ , without impairing biological development. This is a real step forward in the definition of a non-destructive monitoring strategy applicable to tissue production for human regenerative medicine. Discussion This study demonstrates that low-field (0.3 T) MRI systems can effectively monitor the morphology and perfusion of bioprinted tissue constructs, offering a cost-effective and accessible alternative to high-field (7T) MRI. The imaging conditions for both MRI systems are summarized in Table 2 , highlighting the differences in voxel size, acquisition time, and field of view between morphological and flow imaging protocols. Our results show that low-field MRI provides sufficient spatial resolution and contrast to accurately assess scaffold architecture and internal fluid dynamics, with velocity measurements correlating well with computational fluid dynamics (CFD) simulations and in agreement with simple analytical models such as Poiseuille distribution.​ This is the first study reporting perfusion imaging at low field without any contrast agent. Importantly, we extended this evaluation to a longitudinal tissue culture experiment, where low-field MRI was employed to monitor the development of a perfused adipose tissue construct over 34 days. This non-invasive imaging approach allowed for high-quality morphological assessment and confirmation of sustained perfusion throughout the culture period. The final histological analyses confirmed successful differentiation of mesenchymal stem cells into mature adipocytes and significant collagen I deposition, validating the biological relevance of the observed changes. These findings align with recent advancements in low-field MRI technology, which have expanded its applicability beyond traditional domains. For instance, low-field MRI has been successfully employed in musculoskeletal imaging, providing diagnostic quality comparable to high-field systems while offering advantages such as reduced cost and increased portability. Additionally, studies have highlighted the potential of low-field MRI in dynamic body imaging applications, including cardiac and pulmonary assessments, due to its flexibility in pulse sequence design and reduced susceptibility artifacts 25 . In the context of tissue engineering, our work complements efforts to develop non-invasive imaging techniques for monitoring scaffold development and function. For example, recent research has utilized MRI to visualize fluid flow within bone scaffolds, demonstrating its capability to capture complex internal dynamics 26 . Furthermore, the integration of low-field MRI with bioreactor systems has been explored to enable real-time monitoring of tissue maturation processes 27 . By validating the use of low-field MRI for detailed assessment of bioprinted constructs, our study supports its adoption as a practical tool for quality control and functional evaluation in tissue engineering. The accessibility and affordability of low-field MRI systems could facilitate broader implementation in research and clinical settings, promoting advancements in regenerative medicine and personalized healthcare. Table 2 Comparison of High-Field and Low-Field MRI performance for morphological and flow imaging of bioprinted scaffolds Parameter High-Field MRI (7 T) Low-Field MRI (0.3 T) Morphological imaging - voxel size 195 µm (XY) / 1 mm (Z) 234 µm (isotropic) Morphological imaging - acquisition time 15 minutes 14 minutes Morphological imaging - field of view 5 × 5 cm 3 × 3 cm Morphometric error (PLA) Mean absolute distance: 114 µm RMS : 278 µm Mean absolute distance: 157 µm RMS : 265 µm Morphometric error (Hydrogel) Mean absolute distance: (lower, value not specified) Mean absolute distance: 340 µm (RMS 400 µm) Flow imaging - voxel size 195 µm (XY) / 1 mm (Z) 250 µm (XYZ) Flow imaging - acquisition time 58 minutes 45 minutes Flow imaging - field of view 5 × 5 cm 6.4 × 6.4 cm Velocity dispersion (20 mL/min, PLA) 2.2–7.6 mm/s 3.2–8.2 mm/s Velocity dispersion (7 mL/min, PLA) 0.8–5.3 mm/s 1.2–3.6 mm/s Flow path 3D segmentation Achievable, good agreement with CFD simulations Achievable, good agreement with CFD simulations Ability to detect hydrogel defects Yes Yes Conclusion This study demonstrates that low-field MRI is a robust, non-invasive, and accessible modality for the characterization of large-scale biofabricated tissues. By directly comparing 0.3 T and 7 T MRI systems, we showed that low-field imaging provides sufficient spatial resolution and contrast for accurate morphological assessment of both polymeric and hydrogel-based scaffolds, with deviations remaining within quality control tolerances. Flow imaging further confirmed the ability of low-field MRI to visualize perfusion pathways and detect scaffold manufacturing-related defects, with results consistent with both high-field acquisitions and computational fluid dynamics simulations. Importantly, the longitudinal monitoring of a perfused adipose tissue construct highlighted the capacity of low-field MRI to support long term non-destructive evaluation of tissue development and maturation under culture conditions. These findings establish low-field MRI as a practical and scalable tool for integrating imaging into tissue engineering workflows, from construct validation to dynamic culture monitoring. This allow to envision its use for real-time engineered tissue monitoring. By lowering the cost and infrastructure barriers associated with high-field systems, low-field MRI opens new perspectives for quality control, bioreactor integration, and translational applications in regenerative medicine. It is now clear that further developments will followed, targeting more functional tissue characterisations such as oxygen levels, metabolic imaging but also cellularity, particularly adipocyte contents. We believe this is the future of large engineered tissues development and the key to their use in clinical applications. Material and Methods Low-field (LF) benchtop MRI scanner The LF MRI system used consists of a custom-built 300 mT permanent magnet system from Pure Device 310 x 400 x 270 mm/130 kg with a 1H frequency of 13.03 MHz. The magnet's temperature is stabilized at 29°C. The maximum field of view is 40 x 40 x 60 mm. This MRI system is controlled through Universal Serial Bus (USB) using MATLAB software for data acquisition and post-processing. A T/R loop gap coil (Ø 42 mm x 65 mm) geometry was selected for maximum B1 field homogeneity and low susceptibility to coil detuning by sample coupling 28 . The system employs a 3D gradient system (DC 600, Pure Devices GmbH) with a maximum gradient strength of 200 mT/m, enabling x, y, and z imaging. RF coils connect to the scanner's RF amplifier (RF-200, Pure Devices GmbH, Germany) via MB to SMB cable (RG316) wrapped around a ferrite filter to minimize noise and common-mode currents. High-field (HF) MRI scanner The HF MRI system used consists of a Bruker BioSpec 7 Tesla system (Bruker Biospin GbmH, Germany) equipped with a 400 mT/m maximal amplitude gradient set and controlled using a Bruker workstation interfaced with ParaVision 5.1 software for data acquisition and post-processing. A transmit-receive radio-frequency body coil with outer diameter of 112 mm and inner diameter of 72 mm was employed for in vitro Magnetic Resonance acquisition and post-processing. Internal structure and morphology characterization LF morphologic images were acquired by placing the filled and sealed culture chamber inside the transmit-receive body coil in a way that the scaffold was positioned horizontally in the centre of the bore. The morphology analysis of both PLA and hydrogel scaffolds was based on the Spin Echo Method. Each acquisition was set to cover a field of view of 3 cm and 128 pixels in the three directions, leading to images with an isotropic spatial resolution of 234 μm 3 . The echo time was set at 20 msec and 1 repetition average. Under this configuration, the acquisition time turned out to be 14 min. HF morphologic images were acquired by placing the sealed culture chamber inside the transmit-receive body coil in a way that the target is positioned horizontally in the centre of the bore. The morphology analysis of both PLA and hydrogel scaffolds was based on a T2-weighted spin-echo sequence. The imaging was carried out by sweeping consecutive sagittal planes separated by 1 mm. Each acquisition was set to cover a field of view of 5x5 cm discretized into 256x256 pixels, leading to images with an in-plane spatial resolution of 195 μm 2 . Under this configuration, the acquisition time turned out to be 15 min for each sample. The images were encapsulated both in DICOM and 2D sequence formats. Data treatment for internal structure and morphology segmentation LF acquisition data were obtained in MAT-files format (.mat), converted into Meta Image Meta Header file (.mhd), then in DICOM to be easily open and segmented using the open-source software 3D Slicer (v5.8.1). A segmentation threshold was selected in a way that the boundaries cover the known scaffold areas. The oversampling factor was set to 5 and the smoothing factor to 0.3. Some remaining noise signals were removed with a digital graphic tablet inspecting layer by layer. Finally, the segmented scaffold volumes were saved as STL file for visualization and further analysis. HF acquisition data were obtained as DICOM open and segmented using the open-source software 3D Slicer (v5.8.1). The segmented scaffold volumes were obtained with the same settings of 3D slicer used for data at LF. Morphometric analysis of the different segments was conducted using Artec Studio 16 Professional, comparing 1 million discrete points of two STL files. From these data, the mean absolute distance between the two files, together with the Root Mean Square of this distance were calculated and used as metrics. HF velocimetry perfusion imaging Perfusion within the culture chamber was performed using a Masterflex peristaltic pump (Ismatec ISM834C), having a speed range from 0.003 to 31 mL/min, connecting all the elements with autoclavable translucent hose (Saint Gobain Versilic flexible tube Ø 3 mm x Ø 6 mm). Scanning was carried out by placing the culture chamber connected to the perfusion system inside the transmit-receive body coil in a way that the target is positioned horizontally in the centre of the 7T MRI system. Velocity map sequence acquisition was carried out by using the flow-imaging technique known as Flow Map, in which bipolar gradient pulses are added during the encoding period to produce a flow-dependent signal phase 28 . The resulting images are scaled in m/s and represent the maps of flow velocity components. The maximum velocity range is -7.4:7.4 mm/sec, which means that higher velocity will lead to aliased data in the image that cannot be corrected after the acquisition. Pitfalls and limitation of such technique are well described in Guida et al. 23 . Each capture was acquired sweeping consecutive sagittal planes (XY) separated by 1 mm and encoding the velocity vector components in all directions (XYZ). Each acquisition was set to cover a field of view of 5x5 cm discretized into 256x256 pixels, leading to images with an in-plane spatial resolution of 195 μm (XY). The raw data were captured in the binary file format owned by BRUKER, namely, 2D sequence format data (.2dseqs) and then decoded using the BRUKER’s Application Programming Interface (API) installed in the programming language Python, to be finally converted from a four-dimensional array to segregated images in text file format. The set of 92 images corresponding to 23 cuts were loaded in MATLAB to be post-processed. The Euclidean norm of the vector velocity was calculated by combining the vector components in the three directions and scaling the result to obtain a result in meters per second according to equation 2 : LF velocimetry perfusion imaging Scanning was carried out by placing the filled and sealed culture chamber inside the transmit-receive body coil in a way that the scaffold was positioned horizontally in the centre of the bore. Velocity map sequence acquisition was carried out by using a modified Spin Echo sequence with the addition of gradient pulse on the either side of the 180-degree pulse. Flow velocity orthogonal to the image plane is encoded in phase image. This method allows to have direct relation between the phase and the velocity which limits the dependence to the maximum expected velocity as it is for the method implemented at HF. Each capture was acquired sweeping consecutive sagittal planes (XY) and encoding the velocity vector components in all directions (XYZ). The acquisitions were set to cover a field of view of 6.4x6.4 cm, represented in an isotropic 256x256x256-pixel 3D array with an isotropic spatial resolution of 250 μm 3 . The set of 768 images corresponding to 256 cuts were loaded in MATLAB to be post-processed. The Euclidean norm of the vector velocity was calculated by combining the vector components in the three directions, according to Equation 2 . For the 3D velocity maps rendering, the results encapsulated in DICOM, to be treated in the software 3D Slicer, defining two ranges of thresholds. Fitting velocity distribution to Poiseuille’s law Velocity data were fitted to Poiseuille’s equation ( Equation 1 ) adapted with three parameters. The fitting was performed in Matlab© software, using trust region algorithm with a starting value for R selected at 1.25 mm. This starting value corresponds to half width of the channel. This parameter can be considered as the equivalent hydraulic radius of a channel with a circular section. The data were weighted by the ratio of the velocity mean within the voxel divided by standard deviation. Computer fluid dynamics (CFD) The simulation of liquid flow throughout the culture chamber, including the scaffold and associated flow pathways, was performed using COMSOL Multiphysics (version 6.2, COMSOL AB, Sweden). This computational platform employs finite element analysis (FEA) to solve the governing equations of fluid dynamics, providing a comprehensive framework for modelling flow behaviour within the bioreactor system. The culture chamber geometry, including the scaffold and inlet/outlet manifolds, was imported into COMSOL as a CAD file. The liquid flow within the bioreactor was modelled using the Navier-Stokes equations for incompressible fluids. Where ρ is the fluid density, u is the velocity vector, p is pressure, μ is dynamic viscosity, and F represents external forces. Gravity was included in the model to simulate its effects on the horizontal bioreactor setup. The two flow rates of the used in the experiments, namely, 7 mL/min and 20 mL/min, were simulated to evaluate their effects on fluid dynamics within the culture chamber. These flow rates were applied as boundary conditions at the inlet, with zero-gauge pressure set at the outlet. A no-slip condition was enforced on all bioreactors. Water was used as a model fluid, with its density (ρ = 1000 kg/m) and dynamic viscosity (μ = 1 mPa) defined based on standard physical properties. The entire culture chamber geometry including the scaffold with 2.5 mm square channels, was discretized using a tetrahedral finite element mesh. Refinement of the mesh was performed near channel walls, bifurcations, and the inlet and outlet regions to accurately capture velocity gradients. Adaptive mesh refinement was employed to balance computational efficiency and numerical accuracy. Steady-state simulations were performed using segregated solvers to generate velocity fields and pressure distributions across the bioreactor system for both simulated flow rates. Flow regime was estimated and shown to be laminar, with a maximum Reynolds number of 415, well below the turbulence threshold of 2000 ( Supplementary Table 1 ). 3D printing of culture chamber Computer Assisted Design (CAD) of the tissue perfusion chamber was performed using SolidWorks (Dassault Systèmes - SolidWorks Corporation). CAD files were converted to .STL format to be transferred to an Object30 Pro inkjet printer (Stratasys, USA). The chamber was 3D-printed using VeroClear resin (Stratasys, USA). Once printed, the support material was removed using a high-pressure waterjet (Stratasys). The 3D-printed chamber was then incubated overnight in 70 % ethanol at room temperature to remove all leachable compounds, and then finally rinsed in milliQ water. Prior any use for living engineered tissue perfusion, the chamber was steam sterilized at 120°C, 2 Bars for 20 minutes. 3D printing of PLA and hydrogel scaffolds PLA constructs were 3D printed at room temperature with a Prusa i3 MK3S+ (Prusa Research ©, Czech Republic) and PRUSAment PLA filament (Prusa Research ©, Czech Republic), using the design depicted in Figure 1 . Printing parameters were set as layer width 400 µm, layer height 100 µm, 3 perimeters and an infill ratio of 38%. Hydrogel scaffolds were produced with identical CAD design as the PLA scaffold. Hydrogel scaffold were fabricated via extrusion-based 3D bioprinting using a composite bioink comprising 2% (w/v) fibrinogen (F8630-1G, Merck), 2.5% (w/v) porcine skin-derived gelatin (G1890, Sigma-Aldrich), and 2% (w/v) low-viscosity sodium alginate (A18565.36, Alfa Aesar) dissolved in calcium-free Dulbecco’s Modified Eagle Medium (DMEM; 21068028, ThermoFisher). Printing parameters were set as layer width 400µm, layer height 100µm, 3 perimeters and an infill ration of 38%. Printing was carried out at 21 °C using a COSMED333 bioprinter (TOBECA, France) equipped with a Vipro-head 3 extrusion system (VicoTec, Germany) and a 20mm long, 800 µm nozzle directly inside the sprinkler structure. Post-printing, structural integrity was achieved by crosslinking in a 270 mM calcium chloride bath. AD-MSC bioprinting and culture Adipose derived mesenchymal stem cells (AD-MSC) were obtained from the “Hopitaux Civils de Lyon”. AD-MSC were expanded in MSC GROWTH media 2 (Promocell, C-28009) and seeded in Flasks with a density of 2500 cells/cm². Once 80% of cell confluency was attained, cells were washed with PBS, and trypsinized 2 minutes with 3 mL of Trypsine-EDTA 0.05% (Sigma, T2601). After detachment, cells were centrifuged at 300g for 5 minutes and counted. 3D bioprinting bioink was formulated using bovine gelatine (G1890 Sigma, France), very low viscosity alginate (Alfa Aesar, Thermo Fisher France) and fibrinogen from bovine plasma (Sigma, France). All components were handled under sterile laminar flow to ensure sterility. Stock solution of 0.2 g/mL gelatine, 0.04 g/mL alginate and 0.08 g/mL fibrinogen were dissolved, without any stirring for 18 hours at 37 °C, in DMEM (without calcium, with glutamax-1, Invitrogen, France) supplemented with 10% foetal calf serum (HyClone, USA), 20 µg/ml gentamicin (Pantapharm, France), 100 UI/ml penicillin/streptomycin (Sarbach, France) and 1 µg/ml amphotericin B (Bristol Myers Squibb, France). For bioink preparation, trypsinated cells were first suspended in calcium-free DMEM supplemented with 10 % FBS and enumerated. Targeted cell concentration was adjusted by pelleting the appropriate number of cells at 300 g for 5 min. The pellet was suspended in the proper volume of 0.08 g/mL fibrinogen solution to reach 2 x 10 6 cells/mL. Then, to this cell suspension in fibrinogen, the appropriate volumes of alginate and gelatine stock solutions were added to reach a final composition of 0.02 mg/mL fibrinogen, 0.02 mg/mL alginate and 0.05 mg/mL gelatine. The bioink was homogenized and incubated for 15 min at 37°C. After homogenization, sterile cartridge (Nordson EFD, France) was filled with the bioink and incubated 30 minutes at 21°C to stabilise the bioink rheological properties. The cartridge was then loaded in a COSMED333 bioprinter (TOBECA, France) equipped with a Vipro-head 3 extrusion system (VicoTec, Germany) and an 800 µm diameter, 20 mm long needle (Nordson EFD, France). Once bioprinted, the engineered tissues were consolidated with a solution composed of 4 mg/mL transglutaminase (TAG) (ACTIVA WM, Ajinomoto, Japan), 10 U/mL thrombin from bovine plasma (Merck, France) and 90 mM CaCl 2 (Merck, France). The consolidation process was carried out at 37°C for 2 hours, under a 3 mL/min flow rate. They were then rinsed by flowing sterile physiological serum (Versol, France) at a 3 mL/min flow rate within the culture chamber. Once consolidated and rinsed, the tissue was cultured for 7 days under de constant 3 mL/min flow of DMEM (Invitrogen, France) supplemented with 10% foetal calf serum (HyClone, USA). Then, to initiate the AD-MSC differentiation toward adipose tissue, the perfused culture medium was supplemented with 20 μg/ml (Sigma, France), 0.5 mM 3-Isobutyl-1-Methylxanthine (IBMX, Sigma, France) and 50 μM indomethacin (Sigma, France). Differentiation was conducted for an additional 27 days under constant 3 mL/min flow rate. Histological characterisation After the end of the culture, the tissue was fixed during 24 hours at 4°C using Antigenfix (P0014, Diapath, France). It was then prepared for freezing with 2 successive incubations in 15% (w/v) and 30% (w/v) sucrose for 24 hours at 4°C and immerged in a mix of OCT (U/TFM-C, Microm Microtech, France) and 30% sucrose (50%/50%, v/v) for 1h30 at room temperature. The tissue was frozen in the OCT/30% sucrose mix using liquid nitrogen and sliced in 20 µm sections. Tissue sections were stained with 1.8 µg/mL bodipy (790389, Sigma, France) during 20 minutes, washed with PBS and stained with DAPI (D3571, Thermofisher, France) for 10 minutes. For collagen I immunofluorescence, the sections were rinsed in PBS to remove the OCT/30% sucrose, then treated with 0.2% (w/v) Triton X100 (T8787, Sigma, France), saturated using BlockAID blocking solution (B10710, Invitrogen, France) during 20 minutes and then incubated with collagen I primary antibody (dilution 1/300, 20111, Novotec, France) for 2 hours at room temperature. Finally, tissue sections were covered with AlexaFluor488-coupled secondary antibody (dilution 1/500, A11008, Invitrogen, France) during 1 hour at room temperature and incubated with DAPI (D3571, Thermofisher, France) during 10 minutes. All the sections were mounted with Fluoromount (F4680, Sigma, France) and imaged with an Axioscan 7 slide scanner (Zeiss,France). Declarations Acknowledgment: The authors gratefully acknowledge Pure Devices GmbH, manufacturer of the low-field MRI scanner used in this work. Their invaluable technical support and collaboration were crucial in developing and optimizing the specialized sequences necessary for this project. References Lee, S.J., Jeong, W. & Atala, A. 3D Bioprinting for Engineered Tissue Constructs and Patient-Specific Models: Current Progress and Prospects in Clinical Applications. Adv Mater 36 , 2408032 (2024). Ramadan, Q. & Zourob, M. 3D Bioprinting at the Frontier of Regenerative Medicine, Pharmaceutical, and Food Industries. Frontiers in Medical Technology Volume 2 - 2020 (2021). Zhang, Y.S. & Yao, J. Imaging Biomaterial–Tissue Interactions. Trends Biotechnol 36 , 403-414 (2018). Berry, D.B., Englund, E.K., Chen, S., Frank, L.R. & Ward, S.R. Medical imaging of tissue engineering and regenerative medicine constructs. Biomaterials Science 9 , 301-314 (2021). Kotecha, M. 21-48 (2017). Nitzsche, H. et al. Characterization of Scaffolds for Tissue Engineering by Benchtop-Magnetic Resonance Imaging. Tissue Engineering Part C: Methods 15 , 513-521 (2009). Befroy, D.E. & Shulman, G.I. Magnetic resonance spectroscopy studies of human metabolism. Diabetes 60 , 1361-1369 (2011). He, L., Jiang, B., Peng, Y., Zhang, X. & Liu, M. NMR Based Methods for Metabolites Analysis. Anal Chem 97 , 5393-5406 (2025). Cai, Z. et al. Z-Spectral MRI Quantifies the Mass and Metabolic Activity of Adipose Tissues With Fat-Water-Fraction and Amide-Proton-Transfer Contrasts. Journal of Magnetic Resonance Imaging 61 , 1905-1913 (2025). Guensch, D.P. et al. The blood oxygen level dependent (BOLD) effect of in-vitro myoglobin and hemoglobin. Scientific Reports 11 , 11464 (2021). Biondetti, E., Cho, J. & Lee, H. Cerebral oxygen metabolism from MRI susceptibility. Neuroimage 276 , 120189 (2023). McCabe, A., Martin, S., Shah, J., Morgan, P.S. & Panek, R. T(1) based oxygen-enhanced MRI in tumours; a scoping review of current research. The British journal of radiology 96 , 20220624 (2023). Hori, M., Hagiwara, A., Goto, M., Wada, A. & Aoki, S. Low-Field Magnetic Resonance Imaging: Its History and Renaissance. Invest Radiol 56 , 669-679 (2021). Breton, E., Goetz, C., Choquet, P. & Constantinesco, A. Low field magnetic resonance imaging in rat in vivo. Irbm 29 , 366-374 (2008). Nitzsche, H. et al. Characterization of scaffolds for tissue engineering by benchtop-magnetic resonance imaging. Tissue Eng Part C Methods 15 , 513-521 (2009). Caysa, H., Metz, H., Mäder, K. & Mueller, T. Application of Benchtop-magnetic resonance imaging in a nude mouse tumor model. Journal of experimental & clinical cancer research : CR 30 , 69 (2011). Arnold, T.C., Freeman, C.W., Litt, B. & Stein, J.M. Low-field MRI: Clinical promise and challenges. Journal of magnetic resonance imaging : JMRI 57 , 25-44 (2023). Cooley, C.Z. et al. Design and implementation of a low-cost, tabletop MRI scanner for education and research prototyping. J. Magn. Reson. 310 , 106625 (2020). Marquette, C.A. et al. Unlocking the potential of bio-inspired bioinks: A collective breakthrough in mammalian tissue bioprinting. Bioprinting 41 , e00351 (2024). Chastagnier, L. et al. Deciphering dermal fibroblast behavior in 3D bioprinted dermis constructs. Bioprinting 32 , e00275 (2023). Pragnere, S., Essayan, L., El-Kholti, N., Petiot, E. & Pailler-Mattei, C. In vitro bioprinted 3D model enhancing osteoblast-to-osteocyte differentiation. Biofabrication 17 , 015021 (2025). Guida, L., Cavallaro, M. & Levi, M. Advancements in high-resolution 3D bioprinting: Exploring technological trends, bioinks and achieved resolutions. Bioprinting 44 , e00376 (2024). Chai, P. & Mohiaddin, R. How we perform cardiovascular magnetic resonance flow assessment using phase-contrast velocity mapping. Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance 7 , 705-716 (2005). Curd, E.F. et al. in Industrial Ventilation Design Guidebook. (eds. H. Goodfellow & E. Tähti) 677-806 (Academic Press, San Diego; 2001). Pogarell, T. et al. Modern low-field MRI. Skeletal radiology 53 , 1751-1760 (2024). Han, S. et al. Flow inside a bone scaffold: Visualization using 3D phase contrast MRI and comparison with numerical simulations. J Biomech 126 , 110625 (2021). Mohapatra, S.R. et al. Novel Bioreactor Design for Non-invasive Longitudinal Monitoring of Tissue-Engineered Heart Valves in 7T MRI and Ultrasound. Ann Biomed Eng 53 , 383-397 (2025). Staat, C., Mützel, M. & Haase, A. A bridged loop gap resonator (BLGR) for small animal imaging by 1.5 T MRI systems. The Review of scientific instruments 91 3 , 033704 (2020). LaNasa, P.J. & Upp, E.L. in Fluid Flow Measurement (Third Edition). (eds. P.J. LaNasa & E.L. Upp) 19-29 (Butterworth-Heinemann, Oxford; 2014). Additional Declarations There is NO Competing Interest. Supplementary Files TableS1.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8000297","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":540985541,"identity":"dfc5e803-bf63-4541-b70b-9910d4b9c4f7","order_by":0,"name":"Christophe Marquette","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0003-3019-0696","institution":"3d.FAB - Université Lyon1","correspondingAuthor":true,"prefix":"","firstName":"Christophe","middleName":"","lastName":"Marquette","suffix":""},{"id":540985542,"identity":"41ec1204-df14-4759-ac84-bf28921076d1","order_by":1,"name":"Yilbert Gimenez","email":"","orcid":"","institution":"Ampère, UMR5005","correspondingAuthor":false,"prefix":"","firstName":"Yilbert","middleName":"","lastName":"Gimenez","suffix":""},{"id":540985543,"identity":"c645f8f0-6863-4e53-8f04-95a25afea738","order_by":2,"name":"Valernst Gilmus","email":"","orcid":"","institution":"Ampère, UMR5005","correspondingAuthor":false,"prefix":"","firstName":"Valernst","middleName":"","lastName":"Gilmus","suffix":""},{"id":540985544,"identity":"54d248e4-5420-4c9a-a143-469cbc5925a9","order_by":3,"name":"Elliott Cowles","email":"","orcid":"","institution":"3d.FAB - Université Lyon1","correspondingAuthor":false,"prefix":"","firstName":"Elliott","middleName":"","lastName":"Cowles","suffix":""},{"id":540985545,"identity":"27413221-5923-44c7-bb3c-f3a44d82557f","order_by":4,"name":"Lucie Essayan","email":"","orcid":"https://orcid.org/0000-0001-5500-9406","institution":"3d.FAB - Université Lyon1","correspondingAuthor":false,"prefix":"","firstName":"Lucie","middleName":"","lastName":"Essayan","suffix":""},{"id":540985546,"identity":"a881cf86-eaa9-410a-8f58-9499e7f224da","order_by":5,"name":"Vincent Semet","email":"","orcid":"","institution":"Ampère, UMR5005","correspondingAuthor":false,"prefix":"","firstName":"Vincent","middleName":"","lastName":"Semet","suffix":""},{"id":540985547,"identity":"a7a128f4-410b-45ff-b9dd-b6b9fa130367","order_by":6,"name":"Michel Cabrera","email":"","orcid":"","institution":"Ampère, UMR5005","correspondingAuthor":false,"prefix":"","firstName":"Michel","middleName":"","lastName":"Cabrera","suffix":""},{"id":540985548,"identity":"0a0b6014-b963-419a-969a-1bed44166923","order_by":7,"name":"Simon Lambert","email":"","orcid":"","institution":"Ampère, UMR5005","correspondingAuthor":false,"prefix":"","firstName":"Simon","middleName":"","lastName":"Lambert","suffix":""},{"id":540985549,"identity":"dcdc2ec8-33a2-4d07-bc8d-0d4120b8cdac","order_by":8,"name":"Emma Petiot","email":"","orcid":"","institution":"3d.FAB - Université Lyon1","correspondingAuthor":false,"prefix":"","firstName":"Emma","middleName":"","lastName":"Petiot","suffix":""}],"badges":[],"createdAt":"2025-10-31 16:26:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8000297/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8000297/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":95504626,"identity":"c209c51a-00d8-4f77-8c95-aca40ed26b3f","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5843510,"visible":true,"origin":"","legend":"","description":"","filename":"Gimenezetal.Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/970594bfd54ade6ae4ee163f.docx"},{"id":95529660,"identity":"4fc0d308-ca0c-46ab-a754-9ca17040af6a","added_by":"auto","created_at":"2025-11-10 10:17:21","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":342,"visible":true,"origin":"","legend":"","description":"","filename":"rs8000297.json","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/41242051d0b2dd1fc72aaf8e.json"},{"id":95504627,"identity":"51b32bd0-e1e3-4ad5-ba4a-645b6243a929","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":108045,"visible":true,"origin":"","legend":"","description":"","filename":"rs80002972enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/ac4463032d5715e584287df0.xml"},{"id":95529548,"identity":"c32519a8-e7ef-4aa1-9a76-05ede4100370","added_by":"auto","created_at":"2025-11-10 10:17:13","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":708993,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/cda9ccb49cb635d1bba56cd3.png"},{"id":95504629,"identity":"98a0bf0b-b4c7-4720-a7f3-2f9035f2246f","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":983976,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/23438e002a9473434837acd1.png"},{"id":95504640,"identity":"d82303cf-667f-4b8b-92f1-62292215efcc","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":467906,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/bc39350b06ae08bd2095f4ae.png"},{"id":95504639,"identity":"c3713906-0cda-4bf9-853d-d78022e27153","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":470109,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/a087bbdbe0ae99cce36a403b.png"},{"id":95504649,"identity":"a50249fc-0da3-42ea-a3e2-ecc0a4c21ac2","added_by":"auto","created_at":"2025-11-10 05:50:27","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":639502,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/0f0ef5838a9e9b141efc4f8b.png"},{"id":95504641,"identity":"c1ba53c7-2d3e-4961-8cc6-00f46ce6b826","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":953055,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/bb89d782dfc044b8242c7d73.png"},{"id":95504631,"identity":"c0027cb6-5147-4ba6-9fd7-7270588233b7","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1513334,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/e1ffe2ff65a36d86157b77fb.png"},{"id":95504652,"identity":"f756f033-2da2-4970-8a72-1b95e83094a2","added_by":"auto","created_at":"2025-11-10 05:50:27","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2944,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/9880523d91a2d23fe744d384.png"},{"id":95504643,"identity":"87893708-192d-4410-9454-4a7d96271990","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":56596,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/269943fc68b893538bf611d0.png"},{"id":95528847,"identity":"a0538471-0dea-4cc8-ac76-c7670c9cb14e","added_by":"auto","created_at":"2025-11-10 10:16:32","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":109522,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/6cb33b2f54b97bb7b7ec2708.png"},{"id":95504651,"identity":"7964f385-f4de-4818-a53f-a4daf2824996","added_by":"auto","created_at":"2025-11-10 05:50:27","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64277,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/57b2b5fb4195d9ed58df72c5.png"},{"id":95504647,"identity":"4430acc1-6dbf-4122-bd61-34295baf550a","added_by":"auto","created_at":"2025-11-10 05:50:27","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":61217,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/fca90dabd5fd4ec7e4711c44.png"},{"id":95529169,"identity":"026c6e67-254c-42eb-8fdd-6a050edaf76a","added_by":"auto","created_at":"2025-11-10 10:16:51","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64808,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/651b37255fd2014a56b92c69.png"},{"id":95504648,"identity":"3b4ef9b2-a77a-4bdb-af65-eadc75978a29","added_by":"auto","created_at":"2025-11-10 05:50:27","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":160504,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/5ccd723c7373e8b1c832c92c.png"},{"id":95504634,"identity":"b34019fc-a2c1-4c56-9d47-a201ccf02f63","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":209797,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/e0646e54584ac092045ee807.png"},{"id":95529029,"identity":"15999029-2460-45c1-92ee-0f73afbf91aa","added_by":"auto","created_at":"2025-11-10 10:16:42","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1867,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/5cac50081abcf64c7235933a.png"},{"id":95504646,"identity":"07ebedd7-4603-4d9d-8caf-28777f6cc548","added_by":"auto","created_at":"2025-11-10 05:50:27","extension":"xml","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":106886,"visible":true,"origin":"","legend":"","description":"","filename":"rs80002972structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/dcfaeaafe51be8faa7181279.xml"},{"id":95504638,"identity":"7d82d857-4e83-4d64-9e97-341119c19df5","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"html","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":115187,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/d6ec16a254b691be68bcaa5e.html"},{"id":95504624,"identity":"b9af2bb5-5ed9-4ca1-8433-dc266ca0d0ff","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":195159,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003ePerfusion chamber.\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e A. \u003c/strong\u003eView of the culture perfusion chamber designed to host a porous engineered tissue and to fit into HF and LF MRI bore. \u003cstrong\u003eB.\u003c/strong\u003e Top view of the designed scaffold depicting channel dimension. \u003cstrong\u003eC.\u003c/strong\u003e Tilted view of the designed scaffold. \u003cstrong\u003eD.\u003c/strong\u003e X-ray view of the designed scaffold depicting internal channels.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/3986058b38dff7f9365f2ae3.png"},{"id":95504623,"identity":"e142c85a-505e-4b26-b8d9-2d21c047cfb1","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":212722,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eMorphological assessment of PLA and hydrogel samples.\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e \u003c/em\u003e\u003cstrong\u003eA.\u003c/strong\u003e Axial, sagittal and coronal views of HF and LF MR imaging of PAL and hydrogel scaffolds. \u003cstrong\u003eB.\u003c/strong\u003e 3D view of PLA and hydrogel scaffolds after automatic segmentation. \u003cstrong\u003eC.\u003c/strong\u003e Morphometric comparison between segments obtained from HF, LF and the initial STL file for both PLA and hydrogel scaffolds (Mean Absolute Surface Distance: MASD; Root Mean Square: RMS). \u003cstrong\u003eD.\u003c/strong\u003e Cross section view of the PLA and hydrogel scaffolds, depicting unexpected channel (black arrow), reduced channel (red arrow) and deformed channel (blue arrow) in the hydrogel scaffold. \u003cstrong\u003eE.\u003c/strong\u003e X-ray view of hydrogel scaffold LF MR imaging after automatic segmentation, enabling the visualisation of the scaffold internal geometry.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/a7a8777c2e77d002d97863a1.png"},{"id":95504633,"identity":"42babf9c-77fc-4657-8887-770a185b06a9","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":141856,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eVelocimetry assessment in PLA scaffold.\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e A.\u003c/strong\u003e 2D views and flow velocity analysis, the region of interest (ROI), represented as a red rectangle, is placed within one perfused channel of the scaffold. \u003cstrong\u003eB.\u003c/strong\u003e Velocity voxel populations within the ROI. \u003cstrong\u003eC.\u003c/strong\u003eVelocity profiles within the ROI with fitted Poiseuille theoretical flow velocity distribution.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/80bd2be5e7767d5bc3477dda.png"},{"id":95528278,"identity":"df6a23d9-30d9-458d-b435-eafd17bd50ce","added_by":"auto","created_at":"2025-11-10 10:15:49","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":100217,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e3D visualisation of velocimetry assessments within the PLA scaffold.\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003e \u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003eA.\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003e \u003c/strong\u003e\u003c/em\u003eFlow volume representation based on CFD simulation. \u003cstrong\u003eB.\u003c/strong\u003e Flow volume representation based on HF measurements.\u003cstrong\u003e C.\u003c/strong\u003e Flow volume representation based on LF measurements.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/8a5d15b98000a1429e876d8b.png"},{"id":95504625,"identity":"19e773d4-a71b-4076-8099-d24ec60a2e7e","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":119095,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVelocimetry assessment in hydrogel scaffold at 20 mL/min flow rate.\u003c/strong\u003e \u003cstrong\u003eA.\u003c/strong\u003e 2D view of one LF MRI velocimetry image. \u003cstrong\u003eB.\u003c/strong\u003e 3D reconstruction of the imaged velocities. Red arrow shows air bubble presence; Black arrows show blocked hydrogel channels.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/4b33988a48bbb58609cbd5bb.png"},{"id":95504636,"identity":"7d493b02-1bd6-4f0f-9724-5790ff50a8b2","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":239496,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMorphological and nutritive flow velocimetry assessments in cell-laden bioprinted tissue.\u003c/strong\u003e \u003cstrong\u003eA.\u003c/strong\u003e and \u003cstrong\u003eB.\u003c/strong\u003e views of the automatically segmented tissue volume during longitudinal LF MR imaging. \u003cstrong\u003eC.\u003c/strong\u003e X-ray views of the tissues volume longitudinal segments showing internal geometries of the channels. \u003cstrong\u003eD.\u003c/strong\u003e View of the 3D reconstruction of the imaged velocities at day-3 indicating the 16 different channels. Channels tagged in red indicate the completely perfused channels. \u003cstrong\u003eE.\u003c/strong\u003e and \u003cstrong\u003eF.\u003c/strong\u003e views of the entire flow distribution within the tissue at day-3.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/0002303dde233a9c80d91722.png"},{"id":95529449,"identity":"315b72bb-7a1b-4c86-b77e-34641eb93ece","added_by":"auto","created_at":"2025-11-10 10:17:07","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":306942,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdipose tissue characterisation after 34 days of perfused culture and MRI monitoring. A.\u003c/strong\u003e Mature adipocyte imaging using Bodipy staining at three different magnitudes. \u003cstrong\u003eB.\u003c/strong\u003e Collagen type I immunofluorescence characterisation at three different magnitudes. \u003cstrong\u003eC.\u003c/strong\u003e Cell nucleus imaging using DAPI staining at three different magnitudes.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/dd076c02fa27059b0a041632.png"},{"id":95653928,"identity":"0d11fec7-2efe-4643-aa40-f5a439e240b4","added_by":"auto","created_at":"2025-11-11 16:06:13","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2461077,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/9f29618a-8630-4b63-bb59-80380cbf62c7.pdf"},{"id":95504622,"identity":"d3739b8e-4af8-44ab-a173-63f9df8e81b4","added_by":"auto","created_at":"2025-11-10 05:50:26","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":15694,"visible":true,"origin":"","legend":"","description":"","filename":"TableS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8000297/v1/e6028977b13cb08488fddfd6.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Low-field MRI as a multiparametric tool for large engineered tissue characterization","fulltext":[{"header":"Introduction","content":"\u003cp\u003eRecent advancements in 3D bioprinting have enabled the fabrication of increasingly complex and large-scale tissue constructs, including vascularized models of organs such as the heart, liver, and kidney\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. These developments hold significant promise for applications in regenerative medicine, disease modelling, and drug discovery\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e.​\u003c/p\u003e\u003cp\u003eAs engineered tissues grow in size and complexity, there is a pressing need for non-invasive, non-destructive imaging methods to monitor their morphology, maturation, and internal functionality over time. Techniques like near-infrared fluorescence imaging have shown potential for tracking transplanted cells within tissue constructs​, and Raman spectroscopy has been employed for \u003cem\u003ein vivo\u003c/em\u003e monitoring of tissue development in 3D printed scaffolds\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e.​Magnetic Resonance Imaging (MRI) stands out as a versatile modality capable of providing comprehensive insights into engineered tissues\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Its array of acquisition sequences allows for the extraction of diverse crucial information for tissue engineering (TE), including:​\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eMorphological details via T1- and T2-weighted imaging\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e;\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eTissue content through quantitative relaxation mapping (T1, T2, T2*, proton density fat fraction);\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eTissue microarchitecture using Diffusion Weighted Imaging (DWI) or Diffusion Tensor Imaging (DTI);\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003ePerfusion and internal fluid flow using phase-contrast MRI or Arterial Spin Labelling (ASL);\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eChemical composition and metabolite detection via Magnetic Resonance Spectroscopy (MRS) or Chemical Exchange Saturation Transfer (CEST). MRS enables the detection of metabolites\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e such as lactate, choline, creatine, and glutamate, which are indicators of cellular activity and health\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. CEST imaging extends this capability by indirectly detecting low-concentration metabolites and pH changes, enhancing the functional assessment of tissue constructs\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eMonitoring of oxygenation levels using BOLD contrast and T2* mapping\u003csup\u003e10\u003c/sup\u003e.​ Blood Oxygen Level Dependent (BOLD) imaging exploit the paramagnetic properties of deoxyhemoglobin to detect tissue oxygenation\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e while \u003cem\u003eT2 mapping\u003c/em\u003e* reflects local variations in oxygen availability, and oxygen-enhanced MRI can dynamically assess oxygen diffusion and uptake\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e.​\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eDespite these strengths, the widespread adoption of MRI in tissue engineering is limited by the accessibility of high-field systems (\u0026ge;\u0026thinsp;7 T), which are expensive and require specialized infrastructure\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. In contrast, low-field (0.1 to 1 T) and ultra-low field MRI (earth field to 0.1 T)\u003csup\u003e14\u0026ndash;16\u003c/sup\u003e MRI systems have emerged as cost-effective alternatives. The drawback of these systems being a degradation of the signal-to-noise ratio (SNR) when moving from high-field to low-field and then ultra-low field. Commercially available low-field scanners, priced around \u0026euro;100,000, are compact, portable, and require minimal infrastructure\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e.​ One other advantage of working at low field is the possibility to perform continuum of assessment from \u003cem\u003ein vitro\u003c/em\u003e to implanted stage with same contrast mechanisms which is not the case when preclinical data are acquired at high field (7T or even 11.7T).\u003c/p\u003e\u003cp\u003eWhile low-field MRI systems inherently have reduced SNR, these limitations are mitigated in the context of engineered tissues. The biofabricated engineered tissues are static, highly hydrated, and can be scanned under optimized conditions. It allows to reduce motion artefacts and permit for extended acquisition times needed to reach higher spatial resolution or quantitative imaging\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e.​\u003c/p\u003e\u003cp\u003eIn this study, we present a direct comparison between a high-field (HF: 7 T) and a low-field (LF: 0.3 T) MRI system for imaging bioprinted tissue constructs. Indeed, bioprinting is one of the tissue engineering biofabrication strategies which allow for \u003cem\u003ein vitro\u003c/em\u003e fabrication of anatomic scale engineered tissues. Using standardized samples produced via extrusion-based bioprinting, we assess each MRI system's ability to resolve internal structure, characterize morphology, and visualize perfused fluid flow. Our goal is to determine whether low-field MRI can provide sufficient information to support modern biofabrication workflows, potentially democratizing advanced imaging capabilities for non-clinical research.​ As a final validation experiment, we performed the culture of a media perfused bioprinted adipose tissue, demonstrating the potential of LF MRI for longitudinal monitoring of tissue culture.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eCulture Chamber Design and 3D Samples Preparation\u003c/h2\u003e\u003cp\u003eTo demonstrate the feasibility of using MRI for the non-invasive monitoring of tissue-engineered constructs, all imaging experiments were performed within a custom-built tissue culture chamber specifically designed for MRI compatibility (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-A). This chamber allowed for both aseptic perfusion and secure transport of the tissue samples. It was engineered to fit within the bores of both low-field (LF) and high-field (HF) MRI systems. The culture chamber, fabricated entirely via 3D printing using autoclavable Formlabs High Temp V2 resin, featured a cylindrical geometry with a 35 mm external diameter and a length of 185 mm. The central region, where the bioprinted engineered tissue was housed, measured 20 mm in diameter and 80 mm in height. The design incorporated two independent 3 mm fluidic channels for perfused media inlet and outlet. The inlet channel passed through the centre of a piston mechanism, while the outlet channel ran longitudinally along the chamber wall. The inlet was connected to a sprinkler distribution system composed of a 4\u0026times;4 array of flow channels, intended to ensure homogeneous media perfusion across the engineered tissue.\u003c/p\u003e\u003cp\u003eTo maintain positioning and sterility, the chamber included triple internal grooves that allowed the piston to lock in place and serve simultaneously as a build platform for the scaffold. The assembly was made watertight by combining nitrile O-rings with 3D-printed threaded caps and fasteners, which applied precise axial pressure. This setup ensured leak-free imaging conditions and simplified post-printing insertion of the engineered tissue.\u003c/p\u003e\u003cp\u003eIn the present study, three different types of 3D printed sample were placed in the culture chamber. One type composed of polylactic acid (PLA), a rigid thermoplastic commonly used in fused deposition modelling. One type composed of a gelatine-alginate blend hydrogel, chosen for its relevance in soft tissue engineering applications\u003csup\u003e\u003cspan additionalcitationids=\"CR20\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. One type composed of the previous hydrogel seeded with human adipose-derived mesenchymal stem cells (AD-MSCs).\u003c/p\u003e\u003cp\u003eA single STL file was used to print all samples, ensuring strict geometric consistency across experiments. Each samples included a peripheral wall of three concentric layers and a central network of 4\u0026times;4 channels surrounded by additional truncated-square ducts (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-B-D). The internal geometry was designed to align precisely with the chamber's flow distributor, facilitating controlled perfusion throughout the sample.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eMorphometric Evaluation of PLA Scaffolds\u003c/h3\u003e\n\u003cp\u003eMRI acquisitions were first carried out on PLA scaffolds using both HF (7 T) and LF (0.3 T) systems. Raw images examples are given in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-A, demonstrating comparable contrasts and resolution. For each systems, the imaged volumes were segmented (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-B) and quantitatively compared to the original STL file to assess morphological fidelity (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-C). Additional comparisons were performed directly between the volumes obtained at low and high field.\u003c/p\u003e\u003cp\u003eBoth MRI systems successfully resolved the internal features of the PLA sample, taking advantage of the material's high intrinsic contrast. The segmentations produced from the HF and LF images showed a high degree of concordance. The mean absolute surface distance (MASD) between the segmentation obtained at Low and High fields was 144 \u0026micro;m, with a root mean square (RMS) deviation of 248 \u0026micro;m. When each segmentation was compared to the reference STL geometry, the high-field MRI yielded a mean absolute deviation of 111 \u0026micro;m and an RMS of 150 \u0026micro;m, while the low-field dataset resulted in slightly higher but comparable values of 157 \u0026micro;m and 265 \u0026micro;m, respectively. These results indicate that the morphological information extracted from low-field MRI was sufficiently accurate for characterizing the rigid polymer scaffolds.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eMorphometric Evaluation of Hydrogel Scaffolds\u003c/h3\u003e\n\u003cp\u003eHydrogel scaffolds, due to their soft, hydrated nature and lower MRI contrast, presented a more challenging case for imaging. Nevertheless, both HF and LF systems were able to acquire usable images (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-A) from which segmentation was possible (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-B). Morphometric comparison between the low-field segmented volumes and the reference STL model revealed a mean absolute distance of 340 \u0026micro;m and an RMS deviation of 400 \u0026micro;m (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-C). While these values were higher than those observed for PLA, they remain consistent with expectations based on the hydrogel printing process itself. Indeed, hydrogel 3D printing through microextrusion using an 800 \u0026micro;m nozzle is known to yield lower spatial resolution and greater variability than PLA printing using a 400 \u0026micro;m nozzle. This is clearly illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-D by the presence of unexpected channel (black arrow), reduced channel (red arrow) and deformed channel (blue arrow) in the hydrogel scaffold. The higher deviation observed is therefore likely related to the limitations of the fabrication technique rather than to imaging resolution. As described in literature\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, among the panel of bioprinting techniques each exhibit very different resolutions. For further studies on MRI morphometric evaluation of bioprinted parts, it would be interesting to challenge this technology with production from more resolutive biofabrication technologies like photo-polymerization which achieve resolutions in the range of 50\u0026ndash;100 \u0026micro;m.\u003c/p\u003e\u003cp\u003eAcross both materials, these results validate the ability of low-field MRI to capture the essential geometrical features of 3D printed samples with, in the case of the hydrogel, a degree of accuracy compatible with tissue engineering quality control requirements. Very interestingly, it was also possible to visualize the scaffold internal geometry using X-ray view of the hydrogel scaffold imaged using the LF system (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-E), accessing then information usually hidden to the operator.\u003c/p\u003e\u003cp\u003eSuch technology could enhance the definition of biofabrication quality by enabling the measurement of key parameters such as average surface roughness, smallest printable feature size, strand width, printing accuracy, and pore fidelity\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003ch3\u003eLF vs HF toward fluid flow visualisation\u003c/h3\u003e\n\u003cp\u003eFollowing the successful morphological evaluation of both polymeric and hydrogel scaffolds, we next investigated the capacity of low-field and high-field MRI systems to capture internal fluid flow within the same culture setup. These experiments aimed to determine whether low-field imaging, despite its lower spatial and temporal resolution, can provide relevant information regarding flow patterns and perfusion performance in engineered tissue. Results were compared to theoretical simulated flow patterns obtained either thanks to Poiseuille\u0026rsquo;s law fitting performed on single channels or full experimental set up computational flow dynamic (CFD) simulation.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eFlow characterization in PLA samples\u003c/h3\u003e\n\u003cp\u003eIn the rigid polymer PLA samples, two flow rates, namely 7 mL/min and 20 mL/min, were tested to challenge the sensitivity and precision of the MRI systems. This corresponds to a 2.85 (20/7) velocity ratio between the two flow conditions. Clear flow paths were successfully imaged in both conditions, as depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-A images of the central part of the perfusion chamber. From these images, velocities were measured within the 2.5x2.5 mm square channels thus allowing the determination of the velocity voxel population (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-B) and to compare median velocities between HF and LF in each flow conditions.\u003c/p\u003e\u003cp\u003eAt a perfusion rate of 20 mL/min, median velocities of 5.9 mm/sec and 4.4 mm/sec were measured using LF MRI and HF MRI, respectively. At the lower flow rate of 7 mL/min, median velocities of 1.7 mm/sec and 2.7 mm/sec were measured for LF and HF, respectively. These values were considered satisfactory to validate the comparability of the results obtained using the two MRI platforms. We then compared the ratio between the mean velocities measured at the two flow rates, with the 2.85 theoretical ratio. These ratio were 2.14 and 2.48 for LF and HF, respectively, comparing well with the theoretical ratio and validating that the velocities measured using both MRI systems scaled proportionally with the imposed flow rates.\u003c/p\u003e\u003cp\u003eHowever, a more detailed analysis of the velocity distribution within the channels (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-B) revealed greater dispersion in the measurements obtained with the HF system. For the 20 mL/min flow rate, velocities ranged from 0.5 mm/s to 8.2 mm/s using HF MRI, and from 1.1 mm/sec to 6.9 mm/sec with LF MRI. At 7 mL/min, the HF system showed a velocity spread between 0.4 mm/sec and 3.8 mm/sec, while the LF system yielded a narrower range from 0.7 mm/sec to 2.9 mm/sec. This greater dispersion in HF measurements could be related to the increased sensitivity to local flow heterogeneities but also to ROI positioning error around the channel centre\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e .\u003c/p\u003e\u003cp\u003eTo complete this comparison, the velocity profiles within channel obtained using LF and HF MRI were plotted and compared to the Poiseuille theoretical velocity distribution (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-C). Thus, the velocity profile was fitted using Poiseuille\u0026rsquo;s law (\u003cb\u003eEq.\u0026nbsp;1\u003c/b\u003e) using the three fitting parameters, the maximum velocity (Vmax), the channel\u0026rsquo;s hydraulic radius (R) and the offset of the maximum velocity. In this case it is theoretically positioned at the centre of the channel.\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:V\\left(x\\right)=Vmax\\left(1-\\frac{{\\left(x+offset\\right)}^{2}}{{R}^{2}}\\right)\\)\u003c/span\u003e\u003c/span\u003e (Eq.\u0026nbsp;1), x being the position in the transvers plane along the centre of the channel.\u003c/p\u003e\u003cp\u003eTo find optimal fitting of the Poiseuille equation for our experimental data, the three fitting parameters had to derive from their theoretical values (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Mainly, the hydraulic radius of the channels had to be increased to fit the LF experiment data. In this case, first, the hydraulic radius was increased from 1.25 to 2 mm. This was mainly identified to be related to an overestimated measured velocity at the wall of the channel. This local heterogeneity was caused by the lower signal-to-noise ratio of LF system was originating from lower magnetic field intensity. Such a higher velocity at wall compelled the fitting to a larger channel dimension.\u003c/p\u003e\u003cp\u003eThen, the positioning of the channel centre had also to be tuned to be compatible with the XY resolution of both MRI imaging. Indeed, spatial resolution being 195 \u0026micro;m and 250 \u0026micro;m for HF and LF, respectively, the precision of the experimental positioning cannot be better than these dimensions. This positioning was performed thanks to a channel centre offset in an expected\u0026thinsp;\u0026plusmn;\u0026thinsp;185 \u0026micro;m range.\u003c/p\u003e\u003cp\u003eBy fitting experimental results with optimised Poiseuille equation, we demonstrated that the flow profiles determined using the MRI velocimetry protocol were following a Poiseuille model. This is indicated with R-squares values all above 0.88 and fits presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-C. Still, discrepancy were identified on channel radius and centre positioning compared to theoretical values which demonstrated that MRI has intrinsic experimental limitations, mainly related to local heterogeneity and spatial resolution.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cb\u003eOptimal fitting parameters of the Poiseuille model\u003c/b\u003e for the different acquisitions at HF and LF for 7 and 20 ml/min flow. \u003cb\u003eR\u003c/b\u003e represents the hydraulic radius of the square channels\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. \u003cb\u003eOffset\u003c/b\u003e represents the deviation from the theoretical position of the channel centre. \u003cb\u003eVmax\u003c/b\u003e represents the maximum velocity. R-square stands for the coefficient of determination.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFitted data\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eR-square\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eR (mm)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eOffset (\u0026micro;m)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eVmax (m/s)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHF20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.88\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e+\u0026thinsp;184\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e7.6 x 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHF7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.96\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e+\u0026thinsp;63\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.4 x 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLF20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.98\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e+\u0026thinsp;53\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.6 x 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLF7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.99\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-182\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.2 x 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eTheoretical\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003e1\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e1.25\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003e0\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003eNA\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eTo conclude this evaluation, a second strategy was applied which consisted in comparison of the velocities measured using MRI with computational fluid dynamics (CFD) simulations of the perfusion within the culture chamber (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e-A). The velocity fields obtained from both MRI systems were successfully segmented to obtain a three-dimensional representation of the flow paths (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e-B and \u003cb\u003eC\u003c/b\u003e). When compared to computational fluid dynamics (CFD) simulations performed under laminar flow conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e-A), the reconstructed flow trajectories exhibited both spatial and intensity concordance. Three-dimensional reconstructions based on MRI velocimetry, at both 7 and 20 mL/min flow rates, confirmed the excellent overlap between the HF, LF, and CFD-derived flow maps, both in terms of channel positioning and relative velocity magnitudes.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eFlow characterization in hydrogel scaffolds\u003c/h2\u003e\u003cp\u003eFlow imaging was also performed in hydrogel scaffolds at a flow rate of 20 mL/min. In such a softer and more deformable material, both HF and LF MRI systems managed to successfully capture the principal perfusion paths within the structure. As previously observed during morphological imaging, the hydrogel scaffolds exhibited structural deformations, including local channel collapse, off-axis deviation, but also the presence of air bubbles (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThese non-ideal features, which are representative of the practical challenges encountered in engineered tissues biofabrication, were readily visualized with both MRI systems. Deviations from the idealized flow paths seen in PLA were faithfully detected, demonstrating the ability of MRI, even at low field, to monitor real-world imperfections in tissue-engineered constructs. This capacity to identify defects such as structural deformation and air entrapment opens the path to the use of low-field MRI as a quality control tool for the fabrication and validation of functional tissue equivalents.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eBioprinted tissue culture longitudinal experiment\u003c/h3\u003e\n\u003cp\u003eTo further validate the utility of low-field MRI in tissue engineering workflows, we conducted a longitudinal experiment monitoring the development of a perfused, bioprinted adipose tissue. Adipose-derived mesenchymal stem cells (AD-MSCs) were embedded in a bioink at a concentration of 2\u0026nbsp;million cells per millilitres and bioprinted directly within the perfused culture chamber.\u003c/p\u003e\u003cp\u003eThe culture protocol included two distinct phases, an initial 7-day cell expansion phase in DMEM growth medium supplemented with 10% foetal bovine serum, followed by a 27-day adipogenic differentiation phase induced with a cocktail of dexamethasone (20 \u0026micro;g/mL), IBMX (0.5 mM), and indomethacin (50 \u0026micro;M). Throughout the experiment, the construct was continuously perfused at a flow rate of 3 mL/min to ensure adequate nutrient and oxygen delivery.\u003c/p\u003e\u003cp\u003eMorphological assessments were performed at days-3, 17, 24, and 34 using LF MRI (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-A, B). The acquired images demonstrated excellent signal quality and clear delineation of the bioprinted cell-laden hydrogel within the culture chamber, enabling reliable longitudinal tissue morphology segmentation. Internal channels were visible all along the cultivation enabling to identify potential modification of the nutrient flow path (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-C).\u003c/p\u003e\u003cp\u003eIn addition, velocimetry measurement at day-3 was performed to visualize culture medium flow within the bioprinted tissue (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-D\u003cb\u003e- F\u003c/b\u003e). Such analysis permit to observe that in real experimental conditions only 7 channels among the 16 expected through scaffold design, were perfused by continuous culture medium flow. The channels perfused are identified in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-D by red labelling. Such experiment represents the first time that real morphology and nutritive flow distribution within porous, biofabricated tissue constructs can be assessed in a non-destructive and longitudinal manner. This finding is particularly impactful, given that most 3D bioprinting strategies rely on precisely designed geometries to replicate \u003cem\u003ein vivo\u003c/em\u003e-like tissue characteristics\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn order to validate the biocompatibility of such a protocol, biofabricated scaffold level of adipose tissue maturation was evaluated through histochemical staining. The results revealed a homogeneous distribution of mature adipocytes throughout the construct, as indicated by Bodipy staining (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e-A), alongside with strong collagen I deposition (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e-B) and homogeneous cell distribution visualised by adipocyte nucleus staining with DAPI (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e-C). These findings confirm not only the successful differentiation of mesenchymal stem cells into adipocytes but also the extensive matrix remodelling within the tissue with collagen I secretion. When analysing the cell distribution using the larger magnifications of the DAPI staining, a clear homogeneous presence of the cell all over the hydrogel was found, proof of the efficient perfusion of the entire tissue, even though some channels were collapsed or clogged (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-D). Similar results were obtained for the Bodipy staining, showing a homogeneous distribution of the differentiated and mature adipocytes.\u003c/p\u003e\u003cp\u003eFinally, all together these longitudinal characterisation results and the final tissue properties underscore the value of low-field MRI to monitor dynamic tissue development \u003cem\u003ein situ\u003c/em\u003e, without impairing biological development. This is a real step forward in the definition of a non-destructive monitoring strategy applicable to tissue production for human regenerative medicine.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study demonstrates that low-field (0.3 T) MRI systems can effectively monitor the morphology and perfusion of bioprinted tissue constructs, offering a cost-effective and accessible alternative to high-field (7T) MRI. The imaging conditions for both MRI systems are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, highlighting the differences in voxel size, acquisition time, and field of view between morphological and flow imaging protocols.\u003c/p\u003e\u003cp\u003eOur results show that low-field MRI provides sufficient spatial resolution and contrast to accurately assess scaffold architecture and internal fluid dynamics, with velocity measurements correlating well with computational fluid dynamics (CFD) simulations and in agreement with simple analytical models such as Poiseuille distribution.​ This is the first study reporting perfusion imaging at low field without any contrast agent.\u003c/p\u003e\u003cp\u003eImportantly, we extended this evaluation to a longitudinal tissue culture experiment, where low-field MRI was employed to monitor the development of a perfused adipose tissue construct over 34 days. This non-invasive imaging approach allowed for high-quality morphological assessment and confirmation of sustained perfusion throughout the culture period. The final histological analyses confirmed successful differentiation of mesenchymal stem cells into mature adipocytes and significant collagen I deposition, validating the biological relevance of the observed changes.\u003c/p\u003e\u003cp\u003eThese findings align with recent advancements in low-field MRI technology, which have expanded its applicability beyond traditional domains. For instance, low-field MRI has been successfully employed in musculoskeletal imaging, providing diagnostic quality comparable to high-field systems while offering advantages such as reduced cost and increased portability. Additionally, studies have highlighted the potential of low-field MRI in dynamic body imaging applications, including cardiac and pulmonary assessments, due to its flexibility in pulse sequence design and reduced susceptibility artifacts\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eIn the context of tissue engineering, our work complements efforts to develop non-invasive imaging techniques for monitoring scaffold development and function. For example, recent research has utilized MRI to visualize fluid flow within bone scaffolds, demonstrating its capability to capture complex internal dynamics\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Furthermore, the integration of low-field MRI with bioreactor systems has been explored to enable real-time monitoring of tissue maturation processes\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eBy validating the use of low-field MRI for detailed assessment of bioprinted constructs, our study supports its adoption as a practical tool for quality control and functional evaluation in tissue engineering. The accessibility and affordability of low-field MRI systems could facilitate broader implementation in research and clinical settings, promoting advancements in regenerative medicine and personalized healthcare.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eComparison of High-Field and Low-Field MRI performance for morphological and flow imaging of bioprinted scaffolds\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eParameter\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHigh-Field MRI (7 T)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eLow-Field MRI (0.3 T)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eMorphological imaging - voxel size\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e195 \u0026micro;m (XY) / 1 mm (Z)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e234 \u0026micro;m (isotropic)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eMorphological imaging - acquisition time\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e15 minutes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e14 minutes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eMorphological imaging - field of view\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5 \u0026times; 5 cm\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3 \u0026times; 3 cm\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eMorphometric error (PLA)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMean absolute distance: 114 \u0026micro;m\u003c/p\u003e\u003cp\u003eRMS : 278 \u0026micro;m\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMean absolute distance: 157 \u0026micro;m\u003c/p\u003e\u003cp\u003eRMS : 265 \u0026micro;m\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eMorphometric error (Hydrogel)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMean absolute distance: (lower, value not specified)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMean absolute distance: 340 \u0026micro;m (RMS 400 \u0026micro;m)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFlow imaging - voxel size\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e195 \u0026micro;m (XY) / 1 mm (Z)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e250 \u0026micro;m (XYZ)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFlow imaging - acquisition time\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e58 minutes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e45 minutes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFlow imaging - field of view\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5 \u0026times; 5 cm\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e6.4 \u0026times; 6.4 cm\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eVelocity dispersion (20 mL/min, PLA)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.2\u0026ndash;7.6 mm/s\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.2\u0026ndash;8.2 mm/s\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eVelocity dispersion (7 mL/min, PLA)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.8\u0026ndash;5.3 mm/s\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.2\u0026ndash;3.6 mm/s\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFlow path 3D segmentation\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAchievable, good agreement with CFD simulations\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAchievable, good agreement with CFD simulations\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAbility to detect hydrogel defects\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study demonstrates that low-field MRI is a robust, non-invasive, and accessible modality for the characterization of large-scale biofabricated tissues. By directly comparing 0.3 T and 7 T MRI systems, we showed that low-field imaging provides sufficient spatial resolution and contrast for accurate morphological assessment of both polymeric and hydrogel-based scaffolds, with deviations remaining within quality control tolerances. Flow imaging further confirmed the ability of low-field MRI to visualize perfusion pathways and detect scaffold manufacturing-related defects, with results consistent with both high-field acquisitions and computational fluid dynamics simulations. Importantly, the longitudinal monitoring of a perfused adipose tissue construct highlighted the capacity of low-field MRI to support long term non-destructive evaluation of tissue development and maturation under culture conditions. These findings establish low-field MRI as a practical and scalable tool for integrating imaging into tissue engineering workflows, from construct validation to dynamic culture monitoring. This allow to envision its use for real-time engineered tissue monitoring. By lowering the cost and infrastructure barriers associated with high-field systems, low-field MRI opens new perspectives for quality control, bioreactor integration, and translational applications in regenerative medicine. It is now clear that further developments will followed, targeting more functional tissue characterisations such as oxygen levels, metabolic imaging but also cellularity, particularly adipocyte contents. We believe this is the future of large engineered tissues development and the key to their use in clinical applications.\u003c/p\u003e"},{"header":"Material and Methods","content":"\u003cp\u003e\u003cem\u003eLow-field (LF) benchtop MRI scanner\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe LF MRI system used consists of a custom-built 300 mT permanent magnet system from Pure Device 310 x 400 x 270 mm/130 kg with a 1H frequency of 13.03 MHz. The magnet\u0026apos;s temperature is stabilized at 29\u0026deg;C. The maximum field of view is 40 x 40 x 60 mm. This MRI system is controlled through Universal Serial Bus (USB) using MATLAB software for data acquisition and post-processing. A T/R loop gap coil (\u0026Oslash; 42 mm x 65 mm) geometry was selected for maximum B1 field homogeneity and low susceptibility to coil detuning by sample coupling\u003csup\u003e28\u003c/sup\u003e. The system employs a 3D gradient system (DC 600, Pure Devices GmbH) with a maximum gradient strength of 200 mT/m, enabling x, y, and z imaging. RF coils connect to the scanner\u0026apos;s RF amplifier (RF-200, Pure Devices GmbH, Germany) via MB to SMB cable (RG316) wrapped around a ferrite filter to minimize noise and common-mode currents.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eHigh-field (HF) MRI scanner\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe HF MRI system used consists of a Bruker BioSpec 7 Tesla system (Bruker Biospin GbmH, Germany) equipped with a 400 mT/m maximal amplitude gradient set and controlled using a Bruker workstation interfaced with ParaVision 5.1 software for data acquisition and post-processing. A transmit-receive radio-frequency body coil with outer diameter of 112 mm and inner diameter of 72 mm was employed for \u003cem\u003ein vitro\u003c/em\u003e Magnetic Resonance acquisition and post-processing.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eInternal structure and morphology characterization\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eLF morphologic images were acquired by placing the filled and sealed culture chamber inside the transmit-receive body coil in a way that the scaffold was positioned horizontally in the centre of the bore. The morphology analysis of both PLA and hydrogel scaffolds was based on the Spin Echo Method. Each acquisition was set to cover a field of view of 3 cm and 128 pixels in the three directions, leading to images with an isotropic spatial resolution of 234 \u0026mu;m\u003csup\u003e3\u003c/sup\u003e. The echo time was set at 20 msec and 1 repetition average. Under this configuration, the acquisition time turned out to be 14 min.\u003c/p\u003e\n\u003cp\u003eHF morphologic images were acquired by placing the sealed culture chamber inside the transmit-receive body coil in a way that the target is positioned horizontally in the centre of the bore. The morphology analysis of both PLA and hydrogel scaffolds was based on a T2-weighted spin-echo sequence. The imaging was carried out by sweeping consecutive sagittal planes separated by 1 mm. Each acquisition was set to cover a field of view of 5x5 cm discretized into 256x256 pixels, leading to images with an in-plane spatial resolution of 195 \u0026mu;m\u003csup\u003e2\u003c/sup\u003e. Under this configuration, the acquisition time turned out to be 15 min for each sample. The images were encapsulated both in DICOM and 2D sequence formats.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eData treatment for\u003c/em\u003e \u003cem\u003einternal structure and morphology segmentation\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eLF acquisition data were obtained in MAT-files format (.mat), converted into Meta Image Meta Header file (.mhd), then in DICOM to be easily open and segmented using the open-source software 3D Slicer (v5.8.1). A segmentation threshold was selected in a way that the boundaries cover the known scaffold areas. The oversampling factor was set to 5 and the smoothing factor to 0.3. Some remaining noise signals were removed with a digital graphic tablet inspecting layer by layer. Finally, the segmented scaffold volumes were saved as STL file for visualization and further analysis.\u003c/p\u003e\n\u003cp\u003eHF acquisition data were obtained as DICOM open and segmented using the open-source software 3D Slicer (v5.8.1). The segmented scaffold volumes were obtained with the same settings of 3D slicer used for data at LF.\u003c/p\u003e\n\u003cp\u003eMorphometric analysis of the different segments was conducted using Artec Studio 16 Professional, comparing 1 million discrete points of two STL files. From these data, the mean absolute distance between the two files, together with the Root Mean Square of this distance were calculated and used as metrics.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eHF velocimetry perfusion imaging\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003ePerfusion within the culture chamber was performed using a Masterflex peristaltic pump (Ismatec ISM834C), having a speed range from 0.003 to 31 mL/min, connecting all the elements with autoclavable translucent hose (Saint Gobain Versilic flexible tube \u0026Oslash; 3 mm x \u0026Oslash; 6 mm).\u003c/p\u003e\n\u003cp\u003eScanning was carried out by placing the culture chamber connected to the perfusion system inside the transmit-receive body coil in a way that the target is positioned horizontally in the centre of the 7T MRI system. Velocity map sequence acquisition was carried out by using the flow-imaging technique known as Flow Map, in which bipolar gradient pulses are added during the encoding period to produce a flow-dependent signal phase\u003csup\u003e28\u003c/sup\u003e. The resulting images are scaled in m/s and represent the maps of flow velocity components. The maximum velocity range is -7.4:7.4 mm/sec, which means that higher velocity will lead to aliased data in the image that cannot be corrected after the acquisition. Pitfalls and limitation of such technique are well described in Guida et al. \u003csup\u003e23\u003c/sup\u003e. Each capture was acquired sweeping consecutive sagittal planes (XY) separated by 1 mm and encoding the velocity vector components in all directions (XYZ). Each acquisition was set to cover a field of view of 5x5 cm discretized into 256x256 pixels, leading to images with an in-plane spatial resolution of 195 \u0026mu;m (XY).\u003c/p\u003e\n\u003cp\u003eThe raw data were captured in the binary file format owned by BRUKER, namely, 2D sequence format data (.2dseqs) and then decoded using the BRUKER\u0026rsquo;s Application Programming Interface (API) installed in the programming language Python, to be finally converted from a four-dimensional array to segregated images in text file format. The set of 92 images corresponding to 23 cuts were loaded in MATLAB to be post-processed.\u003c/p\u003e\n\u003cp\u003eThe Euclidean norm of the vector velocity was calculated by combining the vector components in the three directions and scaling the result to obtain a result in meters per second according to \u003cstrong\u003eequation 2\u003c/strong\u003e:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eLF velocimetry perfusion imaging\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eScanning was carried out by placing the filled and sealed culture chamber inside the transmit-receive body coil in a way that the scaffold was positioned horizontally in the centre of the bore. Velocity map sequence acquisition was carried out by using a modified Spin Echo sequence with the addition of gradient pulse on the either side of the 180-degree pulse. Flow velocity orthogonal to the image plane is encoded in phase image. This method allows to have direct relation between the phase and the velocity which limits the dependence to the maximum expected velocity as it is for the method implemented at HF. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eEach capture was acquired sweeping consecutive sagittal planes (XY) and encoding the velocity vector components in all directions (XYZ). The acquisitions were set to cover a field of view of 6.4x6.4 cm, represented in an isotropic 256x256x256-pixel 3D array with an isotropic spatial resolution of 250 \u0026mu;m\u003csup\u003e3\u003c/sup\u003e. The set of 768 images corresponding to 256 cuts were loaded in MATLAB to be post-processed. The Euclidean norm of the vector velocity was calculated by combining the vector components in the three directions, according to \u003cstrong\u003eEquation 2\u003c/strong\u003e. For the 3D velocity maps rendering, the results encapsulated in DICOM, to be treated in the software 3D Slicer, defining two ranges of thresholds.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eFitting velocity distribution to Poiseuille\u0026rsquo;s law\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eVelocity data were fitted to Poiseuille\u0026rsquo;s equation (\u003cstrong\u003eEquation 1\u003c/strong\u003e) adapted with three parameters.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe fitting was performed in Matlab\u0026copy; software, using trust region algorithm with a starting value for R selected at 1.25 mm. This starting value corresponds to half width of the channel. This parameter can be considered as the equivalent hydraulic radius of a channel with a circular section. The data were weighted by the ratio of the velocity mean within the voxel divided by standard deviation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eComputer fluid dynamics (CFD)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe simulation of liquid flow throughout the culture chamber, including the scaffold and associated flow pathways, was performed using COMSOL Multiphysics (version 6.2, COMSOL AB, Sweden). This computational platform employs finite element analysis (FEA) to solve the governing equations of fluid dynamics, providing a comprehensive framework for modelling flow behaviour within the bioreactor system.\u003c/p\u003e\n\u003cp\u003eThe culture chamber geometry, including the scaffold and inlet/outlet manifolds, was imported into COMSOL as a CAD file. The liquid flow within the bioreactor was modelled using the Navier-Stokes equations for incompressible fluids.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhere \u0026rho; is the fluid density, u is the velocity vector, p is pressure, \u0026mu; is dynamic viscosity, and F represents external forces. Gravity was included in the model to simulate its effects on the horizontal bioreactor setup. The two flow rates of the used in the experiments, namely, 7 mL/min and 20 mL/min, were simulated to evaluate their effects on fluid dynamics within the culture chamber. These flow rates were applied as boundary conditions at the inlet, with zero-gauge pressure set at the outlet. A no-slip condition was enforced on all bioreactors. Water was used as a model fluid, with its density (\u0026rho; = 1000 kg/m) and dynamic viscosity (\u0026mu; = 1 mPa) defined based on standard physical properties.\u003c/p\u003e\n\u003cp\u003eThe entire culture chamber geometry including the scaffold with 2.5 mm square channels, was discretized using a tetrahedral finite element mesh. Refinement of the mesh was performed near channel walls, bifurcations, and the inlet and outlet regions to accurately capture velocity gradients. Adaptive mesh refinement was employed to balance computational efficiency and numerical accuracy. Steady-state simulations were performed using segregated solvers to generate velocity fields and pressure distributions across the bioreactor system for both simulated flow rates.\u003c/p\u003e\n\u003cp\u003eFlow regime was estimated and shown to be laminar, with a maximum Reynolds number of 415, well below the turbulence threshold of 2000 (\u003cstrong\u003e\u003cem\u003eSupplementary Table 1\u003c/em\u003e\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e3D printing of culture chamber\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eComputer Assisted Design (CAD) of the tissue perfusion chamber was performed using SolidWorks (Dassault Syst\u0026egrave;mes - SolidWorks Corporation). CAD files were converted to .STL format to be transferred to an Object30 Pro inkjet printer (Stratasys, USA). The chamber was 3D-printed using VeroClear resin (Stratasys, USA). Once printed, the support material was removed using a high-pressure waterjet (Stratasys). The 3D-printed chamber was then incubated overnight in 70 % ethanol at room temperature to remove all leachable compounds, and then finally rinsed in milliQ water. Prior any use for living engineered tissue perfusion, the chamber was steam sterilized at 120\u0026deg;C, 2 Bars for 20 minutes.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e3D printing of PLA and hydrogel scaffolds\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003ePLA constructs were 3D printed at room temperature with a Prusa i3 MK3S+ (Prusa Research \u0026copy;, Czech Republic) and PRUSAment PLA filament (Prusa Research \u0026copy;, Czech Republic), using the design depicted in \u003cstrong\u003eFigure 1\u003c/strong\u003e. Printing parameters were set as layer width 400 \u0026micro;m, layer height 100 \u0026micro;m, 3 perimeters and an infill ratio of 38%.\u003c/p\u003e\n\u003cp\u003eHydrogel scaffolds were produced with identical CAD design as the PLA scaffold. Hydrogel scaffold were fabricated via extrusion-based 3D bioprinting using a composite bioink comprising 2% (w/v) fibrinogen (F8630-1G, Merck), 2.5% (w/v) porcine skin-derived gelatin (G1890, Sigma-Aldrich), and 2% (w/v) low-viscosity sodium alginate (A18565.36, Alfa Aesar) dissolved in calcium-free Dulbecco\u0026rsquo;s Modified Eagle Medium (DMEM; 21068028, ThermoFisher). Printing parameters were set as layer width 400\u0026micro;m, layer height 100\u0026micro;m, 3 perimeters and an infill ration of 38%. Printing was carried out at 21 \u0026deg;C using a COSMED333 bioprinter (TOBECA, France) equipped with a Vipro-head 3 extrusion system (VicoTec, Germany) and a 20mm long, 800 \u0026micro;m nozzle directly inside the sprinkler structure. Post-printing, structural integrity was achieved by crosslinking in a 270 mM calcium chloride bath.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAD-MSC bioprinting and culture\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAdipose derived mesenchymal stem cells (AD-MSC) were obtained from the \u0026ldquo;Hopitaux Civils de Lyon\u0026rdquo;. AD-MSC were expanded in MSC GROWTH media 2 (Promocell, C-28009) and seeded in Flasks with a density of 2500 cells/cm\u0026sup2;. Once 80% of cell confluency was attained, cells were washed with PBS, and trypsinized 2 minutes with 3 mL of Trypsine-EDTA 0.05% (Sigma, T2601). After detachment, cells were centrifuged at 300g for 5 minutes and counted.\u003c/p\u003e\n\u003cp\u003e3D bioprinting bioink was formulated using bovine gelatine (G1890 Sigma, France), very low viscosity alginate (Alfa Aesar, Thermo Fisher France) and fibrinogen from bovine plasma (Sigma, France). All components were handled under sterile laminar flow to ensure sterility. Stock solution of 0.2 g/mL gelatine, 0.04 g/mL alginate and 0.08 g/mL fibrinogen were dissolved, without any stirring for 18 hours at 37 \u0026deg;C, in DMEM (without calcium, with glutamax-1, Invitrogen, France) supplemented with 10% foetal calf serum (HyClone, USA), 20 \u0026micro;g/ml gentamicin (Pantapharm, France), 100 UI/ml penicillin/streptomycin (Sarbach, France) and 1 \u0026micro;g/ml amphotericin B (Bristol Myers Squibb, France).\u003c/p\u003e\n\u003cp\u003eFor bioink preparation, trypsinated cells were first suspended in calcium-free DMEM supplemented with 10\u0026nbsp;% FBS and enumerated. Targeted cell concentration was adjusted by pelleting the appropriate number of cells at 300 g for 5 min. The pellet was suspended in the proper volume of 0.08 g/mL fibrinogen solution to reach 2 x 10\u003csup\u003e6\u003c/sup\u003e cells/mL. Then, to this cell suspension in fibrinogen, the appropriate volumes of alginate and gelatine stock solutions were added to reach a final composition of 0.02 mg/mL fibrinogen, 0.02 mg/mL\u003csub\u003e\u0026nbsp;\u003c/sub\u003ealginate and 0.05 mg/mL\u003csub\u003e\u0026nbsp;\u003c/sub\u003egelatine. The bioink was homogenized and incubated for 15 min at 37\u0026deg;C.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAfter homogenization, sterile cartridge (Nordson EFD, France) was filled with the bioink and incubated 30 minutes at 21\u0026deg;C to stabilise the bioink rheological properties. The cartridge was then loaded in a COSMED333 bioprinter (TOBECA, France) equipped with a Vipro-head 3 extrusion system (VicoTec, Germany) and an 800 \u0026micro;m diameter, 20 mm long needle (Nordson EFD, France).\u003c/p\u003e\n\u003cp\u003eOnce bioprinted, the engineered tissues were consolidated with a solution composed of 4 mg/mL transglutaminase (TAG) (ACTIVA WM, Ajinomoto, Japan), 10 U/mL thrombin from bovine plasma (Merck, France) and 90 mM CaCl\u003csub\u003e2\u003c/sub\u003e (Merck, France). The consolidation process was carried out at 37\u0026deg;C for 2 hours, under a 3 mL/min flow rate. They were then rinsed by flowing sterile physiological serum (Versol, France) at a 3 mL/min flow rate within the culture chamber.\u003c/p\u003e\n\u003cp\u003eOnce consolidated and rinsed, the tissue was cultured for 7 days under de constant 3 mL/min flow of DMEM (Invitrogen, France) supplemented with 10% foetal calf serum (HyClone, USA). Then, to initiate the AD-MSC differentiation toward adipose tissue, the perfused culture medium was supplemented with 20 \u0026mu;g/ml (Sigma, France), 0.5 mM 3-Isobutyl-1-Methylxanthine (IBMX, Sigma, France) and 50 \u0026mu;M indomethacin (Sigma, France). Differentiation was conducted for an additional 27 days under constant 3 mL/min flow rate.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eHistological characterisation\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAfter the end of the culture, the tissue was fixed during 24 hours at 4\u0026deg;C using Antigenfix (P0014, Diapath, France). It was then prepared for freezing with 2 successive incubations in 15% (w/v) and 30% (w/v) sucrose for 24 hours at 4\u0026deg;C and immerged in a mix of OCT (U/TFM-C, Microm Microtech, France) and 30% sucrose (50%/50%, v/v) for 1h30 at room temperature. The tissue was frozen in the OCT/30% sucrose mix using liquid nitrogen and sliced in 20 \u0026micro;m sections.\u003c/p\u003e\n\u003cp\u003eTissue sections were stained with 1.8 \u0026micro;g/mL bodipy (790389, Sigma, France) during 20 minutes, washed with PBS and stained with DAPI (D3571, Thermofisher, France) for 10 minutes.\u003c/p\u003e\n\u003cp\u003eFor collagen I immunofluorescence, the sections were rinsed in PBS to remove the OCT/30% sucrose, then treated with 0.2% (w/v) Triton X100 (T8787, Sigma, France), saturated using BlockAID blocking solution (B10710, Invitrogen, France) during 20 minutes and then incubated with collagen I primary antibody (dilution 1/300, 20111, Novotec, France) for 2 hours at room temperature. Finally, tissue sections were covered with AlexaFluor488-coupled secondary antibody (dilution 1/500, A11008, Invitrogen, France) during 1 hour at room temperature and incubated with DAPI (D3571, Thermofisher, France) during 10 minutes.\u003c/p\u003e\n\u003cp\u003eAll the sections were mounted with Fluoromount (F4680, Sigma, France) and imaged with an Axioscan 7 slide scanner (Zeiss,France).\u003c/p\u003e\n"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgment:\u003c/h2\u003e\u003cp\u003eThe authors gratefully acknowledge Pure Devices GmbH, manufacturer of the low-field MRI scanner used in this work. Their invaluable technical support and collaboration were crucial in developing and optimizing the specialized sequences necessary for this project.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eLee, S.J., Jeong, W. \u0026amp; Atala, A. 3D Bioprinting for Engineered Tissue Constructs and Patient-Specific Models: Current Progress and Prospects in Clinical Applications. \u003cem\u003eAdv Mater\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 2408032 (2024).\u003c/li\u003e\n\u003cli\u003eRamadan, Q. \u0026amp; Zourob, M. 3D Bioprinting at the Frontier of Regenerative Medicine, Pharmaceutical, and Food Industries. \u003cem\u003eFrontiers in Medical Technology\u003c/em\u003e \u003cstrong\u003eVolume 2 - 2020\u003c/strong\u003e (2021).\u003c/li\u003e\n\u003cli\u003eZhang, Y.S. \u0026amp; Yao, J. Imaging Biomaterial\u0026ndash;Tissue Interactions. \u003cem\u003eTrends Biotechnol\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 403-414 (2018).\u003c/li\u003e\n\u003cli\u003eBerry, D.B., Englund, E.K., Chen, S., Frank, L.R. \u0026amp; Ward, S.R. Medical imaging of tissue engineering and regenerative medicine constructs. \u003cem\u003eBiomaterials Science\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 301-314 (2021).\u003c/li\u003e\n\u003cli\u003eKotecha, M. 21-48 (2017).\u003c/li\u003e\n\u003cli\u003eNitzsche, H. et al. Characterization of Scaffolds for Tissue Engineering by Benchtop-Magnetic Resonance Imaging. \u003cem\u003eTissue Engineering Part C: Methods\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 513-521 (2009).\u003c/li\u003e\n\u003cli\u003eBefroy, D.E. \u0026amp; Shulman, G.I. Magnetic resonance spectroscopy studies of human metabolism. \u003cem\u003eDiabetes\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, 1361-1369 (2011).\u003c/li\u003e\n\u003cli\u003eHe, L., Jiang, B., Peng, Y., Zhang, X. \u0026amp; Liu, M. NMR Based Methods for Metabolites Analysis. \u003cem\u003eAnal Chem\u003c/em\u003e \u003cstrong\u003e97\u003c/strong\u003e, 5393-5406 (2025).\u003c/li\u003e\n\u003cli\u003eCai, Z. et al. Z-Spectral MRI Quantifies the Mass and Metabolic Activity of Adipose Tissues With Fat-Water-Fraction and Amide-Proton-Transfer Contrasts. \u003cem\u003eJournal of Magnetic Resonance Imaging\u003c/em\u003e \u003cstrong\u003e61\u003c/strong\u003e, 1905-1913 (2025).\u003c/li\u003e\n\u003cli\u003eGuensch, D.P. et al. The blood oxygen level dependent (BOLD) effect of in-vitro myoglobin and hemoglobin. \u003cem\u003eScientific Reports\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 11464 (2021).\u003c/li\u003e\n\u003cli\u003eBiondetti, E., Cho, J. \u0026amp; Lee, H. Cerebral oxygen metabolism from MRI susceptibility. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cstrong\u003e276\u003c/strong\u003e, 120189 (2023).\u003c/li\u003e\n\u003cli\u003eMcCabe, A., Martin, S., Shah, J., Morgan, P.S. \u0026amp; Panek, R. T(1) based oxygen-enhanced MRI in tumours; a scoping review of current research. \u003cem\u003eThe British journal of radiology\u003c/em\u003e \u003cstrong\u003e96\u003c/strong\u003e, 20220624 (2023).\u003c/li\u003e\n\u003cli\u003eHori, M., Hagiwara, A., Goto, M., Wada, A. \u0026amp; Aoki, S. Low-Field Magnetic Resonance Imaging: Its History and Renaissance. \u003cem\u003eInvest Radiol\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, 669-679 (2021).\u003c/li\u003e\n\u003cli\u003eBreton, E., Goetz, C., Choquet, P. \u0026amp; Constantinesco, A. Low field magnetic resonance imaging in rat in vivo. \u003cem\u003eIrbm\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 366-374 (2008).\u003c/li\u003e\n\u003cli\u003eNitzsche, H. et al. Characterization of scaffolds for tissue engineering by benchtop-magnetic resonance imaging. \u003cem\u003eTissue Eng Part C Methods\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 513-521 (2009).\u003c/li\u003e\n\u003cli\u003eCaysa, H., Metz, H., M\u0026auml;der, K. \u0026amp; Mueller, T. Application of Benchtop-magnetic resonance imaging in a nude mouse tumor model. \u003cem\u003eJournal of experimental \u0026amp; clinical cancer research : CR\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 69 (2011).\u003c/li\u003e\n\u003cli\u003eArnold, T.C., Freeman, C.W., Litt, B. \u0026amp; Stein, J.M. Low-field MRI: Clinical promise and challenges. \u003cem\u003eJournal of magnetic resonance imaging : JMRI\u003c/em\u003e \u003cstrong\u003e57\u003c/strong\u003e, 25-44 (2023).\u003c/li\u003e\n\u003cli\u003eCooley, C.Z. et al. Design and implementation of a low-cost, tabletop MRI scanner for education and research prototyping. \u003cem\u003eJ. Magn. Reson.\u003c/em\u003e \u003cstrong\u003e310\u003c/strong\u003e, 106625 (2020).\u003c/li\u003e\n\u003cli\u003eMarquette, C.A. et al. Unlocking the potential of bio-inspired bioinks: A collective breakthrough in mammalian tissue bioprinting. \u003cem\u003eBioprinting\u003c/em\u003e \u003cstrong\u003e41\u003c/strong\u003e, e00351 (2024).\u003c/li\u003e\n\u003cli\u003eChastagnier, L. et al. Deciphering dermal fibroblast behavior in 3D bioprinted dermis constructs. \u003cem\u003eBioprinting\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, e00275 (2023).\u003c/li\u003e\n\u003cli\u003ePragnere, S., Essayan, L., El-Kholti, N., Petiot, E. \u0026amp; Pailler-Mattei, C. In vitro bioprinted 3D model enhancing osteoblast-to-osteocyte differentiation. \u003cem\u003eBiofabrication\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 015021 (2025).\u003c/li\u003e\n\u003cli\u003eGuida, L., Cavallaro, M. \u0026amp; Levi, M. Advancements in high-resolution 3D bioprinting: Exploring technological trends, bioinks and achieved resolutions. \u003cem\u003eBioprinting\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, e00376 (2024).\u003c/li\u003e\n\u003cli\u003eChai, P. \u0026amp; Mohiaddin, R. How we perform cardiovascular magnetic resonance flow assessment using phase-contrast velocity mapping. \u003cem\u003eJournal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 705-716 (2005).\u003c/li\u003e\n\u003cli\u003eCurd, E.F. et al. in Industrial Ventilation Design Guidebook. (eds. H. Goodfellow \u0026amp; E. T\u0026auml;hti) 677-806 (Academic Press, San Diego; 2001).\u003c/li\u003e\n\u003cli\u003ePogarell, T. et al. Modern low-field MRI. \u003cem\u003eSkeletal radiology\u003c/em\u003e \u003cstrong\u003e53\u003c/strong\u003e, 1751-1760 (2024).\u003c/li\u003e\n\u003cli\u003eHan, S. et al. Flow inside a bone scaffold: Visualization using 3D phase contrast MRI and comparison with numerical simulations. \u003cem\u003eJ Biomech\u003c/em\u003e \u003cstrong\u003e126\u003c/strong\u003e, 110625 (2021).\u003c/li\u003e\n\u003cli\u003eMohapatra, S.R. et al. Novel Bioreactor Design for Non-invasive Longitudinal Monitoring of Tissue-Engineered Heart Valves in 7T MRI and Ultrasound. \u003cem\u003eAnn Biomed Eng\u003c/em\u003e \u003cstrong\u003e53\u003c/strong\u003e, 383-397 (2025).\u003c/li\u003e\n\u003cli\u003eStaat, C., M\u0026uuml;tzel, M. \u0026amp; Haase, A. A bridged loop gap resonator (BLGR) for small animal imaging by 1.5 T MRI systems. \u003cem\u003eThe Review of scientific instruments\u003c/em\u003e \u003cstrong\u003e91 3\u003c/strong\u003e, 033704 (2020).\u003c/li\u003e\n\u003cli\u003eLaNasa, P.J. \u0026amp; Upp, E.L. in Fluid Flow Measurement (Third Edition). (eds. P.J. LaNasa \u0026amp; E.L. Upp) 19-29 (Butterworth-Heinemann, Oxford; 2014).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Adipose tissue, Bioreactor, Engineered tissue, Magnetic resonance imaging, Velocimetry","lastPublishedDoi":"10.21203/rs.3.rs-8000297/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8000297/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAdvances in 3D bioprinting have enabled the fabrication of large and complex engineered tissues, but their increasing size demands non-invasive tools for monitoring structure, maturation, and perfusion. Magnetic Resonance Imaging (MRI) offers unique multiparametric capabilities, yet high-field systems remain costly and inaccessible for most laboratories. In this study, we evaluate the potential of low-field (LF, 0.3 T) MRI as an affordable and versatile alternative to high-field (HF, 7 T) MRI for characterizing bioprinted tissue constructs. Using standardized PLA and hydrogel scaffolds within a custom-designed perfusion chamber, we compared LF and HF imaging performance for morphology and flow visualization. Both modalities successfully resolved internal scaffold features, with morphometric deviations from reference CAD models remaining within quality control tolerances. Flow imaging demonstrated that LF MRI could capture velocity distributions consistent with HF measurements and computational fluid dynamics simulations, even revealing fabrication-induced defects such as channel collapse or occlusion. Finally, we applied LF MRI for longitudinal monitoring of a perfused adipose tissue construct over 34 days. This approach enabled repeated non-destructive assessments of morphology and perfusion, with final histological analyses confirming homogeneous adipogenic differentiation and extracellular matrix deposition. Together, these results establish LF MRI as a powerful tool for real-time, non-invasive evaluation of biofabricated tissues. By combining affordability, portability, and multiparametric imaging capacity, LF MRI broadens access to advanced monitoring strategies in tissue engineering and regenerative medicine, supporting both quality control and functional assessment of large-scale engineered constructs.\u003c/p\u003e","manuscriptTitle":"Low-field MRI as a multiparametric tool for large engineered tissue characterization","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-10 05:50:20","doi":"10.21203/rs.3.rs-8000297/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b97bdfa1-6eab-4090-b5d6-ee5696a2eebd","owner":[],"postedDate":"November 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":57550606,"name":"Biological sciences/Biotechnology/Tissue engineering"},{"id":57550607,"name":"Biological sciences/Biotechnology/Regenerative medicine"}],"tags":[],"updatedAt":"2025-11-10T05:50:20+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-10 05:50:20","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8000297","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8000297","identity":"rs-8000297","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}