Spinothalamic Tract Microstructure as a Common Neural Substrate of Pain Sensitivity Across Modalities: A Combined Brain-Spinal Cord Diffusion Imaging Study

Spinothalamic Tract Microstructure as a Common Neural Substrate of Pain Sensitivity Across Modalities: A Combined Brain-Spinal Cord Diffusion Imaging Study | 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 Spinothalamic Tract Microstructure as a Common Neural Substrate of Pain Sensitivity Across Modalities: A Combined Brain-Spinal Cord Diffusion Imaging Study Jixin Liu, Leiming Wu, Binglan Li, Zhaoxing Wei, Xiaomin Lin, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8640197/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract The spinothalamic tract (STT) is the principal ascending pathway for nociceptive transmission and a key integrative hub for pain. However, whether inter‑individual microstructural variability constitutes a shared neuroanatomical basis for cross‑modal pain sensitivity remains unclear, partly due to limitations in tract‑level imaging along the brain-spinal axis. We applied a combined brain-spinal cord diffusion tensor imaging framework to characterize STT microstructure across the central nervous system. Multivariate analyses identified a distributed STT microstructural pattern robustly associated with individual sensitivity to heat and mechanical pain across ten experimental measures. Although STT microstructure was also related to tactile sensitivity, pain and tactile measures were linked to largely distinct multivariate patterns, indicating a modality-differentiated organization within the tract. Importantly, an STT-derived microstructural pattern predicted individual differences in experimental pain sensitivity and generalized to clinical pain severity in cohorts with zoster and irritable bowel syndrome. Together, these findings provide that variability in STT microstructure shapes individual pain vulnerability and establishes a structural framework linking ascending nociceptive pathways to experimental and clinical pain in humans. Biological sciences/Neuroscience/Neural circuits Health sciences/Neurology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Pain sensitivity, which reflects the magnitude of an individual's response to noxious stimulation, is a fundamental dimension of human somatosensory function with broad implications for clinical practice and neuroscience research 1 , 2 . Although heat, mechanical, and chemical pain arise from distinct peripheral mechanisms, pain thresholds across modalities are moderately to strongly intercorrelated, suggesting shared central substrates that regulate generalized nociceptive responsiveness 3 . A principal anatomical basis for such cross-modal integration lies in the ascending nociceptive pathways, in which diverse peripheral nociceptive inputs are relayed via the dorsal horn before projecting to the thalamus and cortex 4 . Despite the well-characterized anatomy of nociceptive pathways in animal models, the microstructural basis underlying shared pain sensitivity in humans remains insufficiently characterized. Among the ascending nociceptive pathways, the spinothalamic tract (STT) is a central conduit for transmitting nociceptive signals from the periphery to the brain and is therefore a prime candidate for shaping individual differences in pain sensitivity. Its role has been demonstrated in rodent models using neuroanatomical tracing and electrophysiological techniques 5 , 6 . Studies of spinothalamic neurons, including wide-dynamic-range and nociceptive‑specific populations, have exhibited graded responses to heat, mechanical, cold, and chemical stimulation 7 . Clinically, structural damage to the STT correlates with pain severity in several chronic pain syndromes, including chronic neck and shoulder pain, as well as neuropathic pain following spinal cord injury or stroke 8 , 9 . These findings suggest that the STT is anatomically and physiologically positioned to integrate nociceptive signals across modalities. However, it remains unknown whether inter-individual variability in STT microstructure in humans constitutes a common biological substrate underlying pain sensitivity across different pain modalities. Beyond nociception, the STT is a polymodal pathway for sensory integration, mediating sensations such as touch and itch 10 , 11 , 12 . Neurophysiological studies indicate that wide-dynamic-range neurons integrate both non-noxious and noxious inputs within overlapping populations, raising the question of whether STT microstructural properties are selective for pain or instead reflect a more general capacity for sensory encoding across domains 13 . In humans, tract-level characterization of the STT has remained challenging because of technical limitations in neuroimaging 14 . Diffusion MRI studies of pain have predominantly focused on the brain, whereas reliable imaging of the cervical spinal cord has been hindered by physiological noise related to respiration and cardiac pulsation, motion artifacts, and susceptibility-induced signal loss. Consequently, the human spinothalamic tract has rarely been examined as a continuous pathway spanning the spinal cord and brain, hindering efforts to link tract-level microstructural variability along the ascending nociceptive pathway to individual differences in pain sensitivity. While integrated brain-spinal cord imaging has been developed in functional MRI studies of pain 15 , 16 , diffusion-based tract-level characterization of the spinothalamic pathway remains underdeveloped. In this study, we develop and apply a combined brain-spinal cord diffusion MRI framework to map STT microstructure across the entire central nervous system. We analyzed three datasets comprising healthy participants exposed to multiple laboratory-based painful stimuli, including heat threshold and tolerance, mechanical pain, and tactile sensations (Dataset 1), and two consisting of patients with acute zoster pain (AZP; Datasets 2) and irritable bowel syndrome (IBS; Datasets 3).We aimed to determine whether multimodal pain sensitivity shares a common microstructural correlate with the STT, whether this relationship is selective for painful stimuli, and whether STT microstructure can predict both experimental and clinical pain. We first applied partial least squares correlation (PLSC) to identify multivariate associations between STT microstructure and pain sensitivity across modalities in Dataset 1. Pain selectivity was then assessed by contrasting associations for painful versus non-painful tactile measurements. We next applied a multivariate prediction framework combining PLSC and partial least squares regression (PLSR) to evaluate whether this shared substrate could predict individual differences in multimodal pain sensitivity. Finally, after establishing the predictive performance in healthy participants, we evaluated model generalizability by predicting clinical pain severity in patients with AZP and IBS. Results Quality check All diffusion datasets underwent a standardized combined brain-spinal cord preprocessing pipeline prior to tract reconstruction. Because long-echo-time diffusion acquisitions are particularly susceptible to magnetic susceptibility-induced distortions, eddy currents, and motion-related artifacts, we applied FSL's topup and eddy for susceptibility and eddy-current correction (Fig. 1 A) 17 . Furthermore, an automated DTI quality-control framework was implemented to compute the voxel-wise temporal signal-to-noise ratio (tSNR) for the b0 (b = 0 s/mm2) images across all participants 18 . Across the full sample, the preprocessed diffusion data yielded a mean b0 tSNR of 35.09 ± 7.42 (mean ± standard deviation [SD]; Figure S1 ). An example participant with the median b0 tSNR is presented in Fig. 1 . Distortion correction effectively transformed raw images into undistorted anatomical space, removing major susceptibility- and eddy-related artifacts (Fig. 1 A). SNR values were consistently higher in the brain than in the spinal cord (Fig. 1 B). Following preprocessing, the brain-spinal cord diffusion data supported robust fiber-pathway reconstruction along the full brain-spinal axis. Individual STT streamlines were visualized in native space, while their corresponding voxel-wise tract masks were accurately registered to standard space (Fig. 1 C). White-Matter Microstructure of the STT Reflects individual Pain Sensitivity Dataset 1 included ten painful stimulus conditions, comprising heat-pain threshold and tolerance as well as mechanical-pain thresholds at the C5, C6, and C7 dermatomes bilaterally. Group-level pain ratings exhibited the expected variability, with heat-pain thresholds and tolerance of 41.04 ± 3.31°C and 45.94 ± 2.92°C, respectively, while mechanical-pain thresholds of 5.02 ± 0.85, 5.17 ± 0.95, and 5.09 ± 0.97 a.u. for the C5, C6, and C7 dermatomes (Fig. 2 A; Table S1 ). Pearson correlation analysis across all ten stimuli revealed two distinct clusters, showing strong within-modality correlations for heat and mechanical pain but substantially weaker cross-modality relationships (all FDR-correct p < 0.05; Fig. 2 B). Due to methodological challenges associated with registering brain-spinal cord imaging data between native and standard spaces, all tract-level analyses were conducted in native space. Diffusion metrics were extracted along each participant's STT streamlines, including fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD). These tract-level values formed the multivariate feature set used for subsequent analyses. Using PLSC, a robust multivariate association was observed between STT microstructural properties and pain sensitivity across individuals (Fig. 3 A). The first latent variable (LV1) explained 63.08% of the cross-block covariance, indicating a consistent directional relationship across all ten pain measures. Image scores for LV1 showed significant associations with all pain measures (|bootstrap ratio [BSR]| > 1.96, p 1.96) were distributed across the STT, including segments in the left C1, C4, and C6 levels, the right C2 and C5 levels, the substantia nigra, and the bilateral superior and inferior medullary reticular formations (Fig. 3 B). To summarize regions consistently contributed across diffusion metrics, we generated an overlap map retaining features present in > 50% of features, thereby revealing an anatomical core shared across metrics (Fig. 3 C). We further investigated whether heat- and mechanical-pain modalities contributed similarly to the multivariate pattern. Modality-specific PLSC analyses identified significant associations for both heat and mechanical pain, revealing convergent anatomical loci, including left C6, right C5, right C1, and the substantia nigra (Figs. 3 D, 3 E, and Figure S2). Dice similarity analyses indicated substantial overlap between the full-model results (Fig. 3 B) and the modality-specific maps, with Dice coefficients of 0.41 for heat pain and 0.77 for mechanical pain (Figures S3A, S3B), indicating shared yet distinct anatomical substrates across modalities. To account for demographic differences (Dataset 1 showed sex equivalence: χ2(1) = 0.96, p = 0.33, but a significant age difference: t(50) = 47.75, p < 0.001, Cohen's d = 3.21), we reexamined tract-pain correlations using partial correlations controlling for sex and age. Although effect sizes changed slightly, the spatial pattern remained highly consistent with the original results (Figure S3), suggesting that demographic variation did not account for the observed microstructure-pain associations. White-Matter Microstructure of the STT Differentially Reflect Pain and Tactile Sensitivity Having demonstrated that STT microstructural properties are associated with pain measurements, we next examined whether this relationship observed in Dataset 1 was specific to pain by assessing associations between STT microstructural properties and tactile sensitivity. Although the segregation of pain and tactile processing has been proposed in prior research 19 , how this distinction is reflected in the microstructural organization of the human STT remains poor characterized. PLSC analysis identified a distinct multivariate pattern linked to tactile sensitivity (Fig. 4 A). Notably, the spatial distribution of tactile-related regions showed minimal overlap with the pain-related regions identified earlier (pain: Fig. 2 B; tactile: Fig. 4 B), as quantified by a Dice coefficient of 0.28. This association remained consistent after controlling for sex and age (Figure S4). Multivariate STT Microstructural Pattern Predicts Individual Pain Sensitivity A multivariate pattern analysis (MVPA) framework was developed based on STT microstructural properties, and leave-one-out cross-validation (LOOCV) was employed to simultaneously predict individual pain sensitivity across all ten pain measurements. For Dataset 1, data were iteratively partitioned into a discovery set (n = 50) and a holdout test set (n = 1), such that each participant served as the holdout test case once. Initially, the model utilized PLSC to identify the multivariate STT microstructural pattern jointly associated with all pain measurements, and then used PLSR to simultaneously predict each pain measurement in the holdout participant (PLSC-PLSR; Fig. 5 A). Across holdout tests, the PLSC-PLSR model simultaneously and significantly predicted all ten pain measurements, as shown by LOOCV performance across participants (n = 51; p < 0.05; Fig. 5 A). To derive a unified behavioral target reflecting shared variance across pain modalities, principal component regression (PCR) was applied. Principal component analysis (PCA) was performed on the ten pain measurements to extract a latent pain-sensitivity score that captured their common variance (Fig. 5 B). The first principal component was used as the outcome variable in a regression model with STT microstructural features as predictors. The resulting PCR model showed robust predictive performance under LOOCV (n = 51; r = 0.406; p = 0.003; Fig. 5 B), supporting the existence of a shared STT-based substrate underlying individual differences in pain sensitivity. Finally, to assess modality specificity, we applied the same model to tactile sensitivity measurements. Consistent with earlier results, the pain-derived STT pattern did not predict tactile sensitivity (p > 0.05; Figure S5), supporting its selectivity for nociceptive rather than somatosensory-general processing. STT-Based Pain Sensitivity Pattern Generalizes to Clinical Pain in AZP and IBS Having established in Dataset 1 that STT microstructure predicts individual pain sensitivity, encompassing both modality-specific measures and a unified latent pain-sensitivity score, we applied the PLSC-PCR pain-sensitivity model to two independent clinical cohorts. In patients with AZP (Dataset 2), the model significantly predicted the intensity of acute zoster pain intensity (n = 18; r = 0.710; p = 0.003; Fig. 5 C). Similarly, in patients with IBS (Dataset 3), the model significantly predicted the severity of abdominal pain severity (n = 76; r = 0.333; p = 0.003; Fig. 5 D). Discussion Using combined brain-spinal cord diffusion MRI, this study provides a tract-level characterization of STT microstructure across the central nervous system and demonstrates that inter-individual variability along this pathway systematically correlates with pain sensitivity across multiple experimental modalities. Three key findings emerged. First, strong associations were identified between STT microstructural properties and pain sensitivity across ten experimentally derived measures, spanning both thermal and mechanical pain modalities, supporting the existence of a shared structural substrate underlying generalized nociceptive responsiveness. Second, although STT microstructure was associated with tactile sensitivity, pain and tactile measures exhibited largely distinct multivariate patterns, indicating a modality-differentiated organization within the tract. Third, an STT-derived multivariate pattern predicted individual differences in experimental pain sensitivity in healthy participants and generalized to predict clinical pain severity in independent cohorts of patients with AZP and IBS. Collectively, these findings identify the STT microstructure as a common neuroanatomical substrate of pain sensitivity and establish a structural framework linking individual pain vulnerability to ascending nociceptive pathways in humans. STT Microstructure as a Shared Neuroanatomical Substrate for Multimodal Pain Sensitivity The STT serves as the principal ascending pathway responsible for transmitting nociceptive information from the spinal dorsal horn to supraspinal structures, particularly the brainstem and thalamus 20 , 21 . Our results demonstrate that inter-individual variability in STT white-matter microstructure is strongly associated with pain sensitivity across multiple experimental modalities, including heat pain threshold, heat pain tolerance, and mechanical pain sensitivity. Importantly, these associations were not confined to any single spinal segment but instead reflected a tract-level pattern distributed along the STT, spanning cervical spinal levels, and extending into brainstem and thalamic regions. This spatially extended and modality-convergent pattern supports the interpretation that STT microstructure constitutes a unified tract-level substrate underlying generalized pain sensitivity in humans, rather than modalities or segment-specific phenomena. Heat and mechanical pain rely on distinct peripheral transduction mechanisms and afferent fibers 22 , yet their sensitivities converge in STT microstructural features. Further evidence indicates that high-threshold nociceptive (NOX) neurons in the spinal cord's anterolateral tract (including the STT) are broadly responsive to both thermal and mechanical stimuli and can be activated by low-intensity warmth following peripheral sensitization 23 . This convergence suggests that STT microstructure does not primarily encode modality-specific pain representations 24 , but instead reflects a higher-order property of the nociceptive system, such as general nociceptive gain or amplification capacity. Individuals with specific STT microstructural profiles may therefore exhibit heightened responsiveness across various pain modalities, providing a structural basis for generalized pain susceptibility. Extensive animal electrophysiological studies have established that spinothalamic neurons encode nociceptive input in a graded, intensity-dependent manner across thermal and mechanical modalities, providing a cellular and circuit-level account of how pain signals are centrally transmitted 20 , 21 . However, these mechanistic studies were not designed to explain why human pain sensitivity varies so markedly across individuals. By integrating tract-level STT microstructural analysis with multivariate modeling in humans, the present study complements animal findings by demonstrating that inter-individual differences in STT organization underlie variability in pain sensitivity, thereby bridging the gap between cellular nociceptive mechanisms and observed behavioral differences. Modality-differentiated organization of pain and tactile sensitivity within the STT Although the spinothalamic tract is traditionally regarded as a major nociceptive pathway, it is also known to convey non-noxious somatosensory signals such as touch and itch 25 , 26 . Anatomical and physiological studies have shown that STT fibers originate from distinct spinal laminae and follow partially differentiated trajectories within the tract 10 , 11 , 12 , which reflects heterogeneity in the types of sensory information transmitted, rather than an absolute segregation between painful and non-painful modalities 27 . Consistent with this background, our results indicate that pain and tactile sensitivity are underpinned by partially distinct microstructural features within the STT. While white-matter microstructural properties within the tract were significantly mapped onto both pain and tactile sensitivity, these associations were associated with largely distinct multivariate patterns. The spatial overlap between pain-related and tactile-related STT features was limited (Dice = 0.28), suggesting only modest convergence at the tract level. Crucially, this dissociation was further bolstered by the finding that the STT-based pattern derived from pain sensitivity failed to predict individual differences in tactile sensitivity, despite its robust predictive utility for pain across multiple modalities. This pattern of partial segregation aligns with prior experimental evidence demonstrating functional heterogeneity within the STT. For instance, electrophysiological recordings have revealed that individual spinothalamic neurons can exhibit distinct response profiles to graded mechanical and thermal stimulation, including separate classes of nociception-specific and polymodal units, each characterized by distinct stimulus-response curves 28 . In primates, spinothalamic neurons responsive to pruritic versus other somatosensory inputs project to partially distinct thalamic nuclei, further illustrating functional heterogeneity within the pathway 29 . By demonstrating that such functional distinctions are reflected in separable multivariate microstructural patterns in humans, our findings provide tract-level evidence that STT organization cannot be reduced to a single, modality-independent encoding of sensory intensity. Instead, they suggest that alterations in STT microstructure may selectively influence pain vulnerability without uniformly affecting other somatosensory functions. From Experimental Pain to Clinical Pain: Generalizability of STT-Based Patterns Beyond correlational findings in healthy participants, the multivariate STT pattern derived from Dataset 1 was applied to two independent patient cohorts characterized by distinct etiologies and symptom profiles. Despite heterogeneity in pain type, clinical context, rating scales, and MRI acquisition parameters, the model significantly predicted clinical pain severity in both AZP and IBS. These findings converge with accumulating neuroimaging evidence that white matter microstructural alterations within the ascending nociceptive pathway are clinically meaningful. For example, patients with chronic neck and shoulder pain exhibit localized reductions in STT FA and increases in MD relative to healthy controls, with these changes correlating with both pain intensity and duration 30 . In central post-stroke pain, diffusion tensor tractography has reported decreased FA and tract volume in the affected STT, suggesting a role for STT injury in the pathogenesis of persistent pain following cerebral infarction 9 . Additionally, patients with chronic back pain show reduced fiber density in white-matter tracts including the spinothalamic tracts, with microstructural differences related to clinical symptom measures 31 . These clinical examples illustrate that white-matter microstructural features similar to those identified in our predictive model are systematically associated with pain severity across diverse chronic pain conditions. Limitation Several limitations of this study should be acknowledged. First, although DTI provides a unique, noninvasive window into in-vivo white-matter microstructure, the biological specificity of conventional DTI metrics remains limited. Metrics such as FA, MD, RD, and AD represent a composite of microstructural features, including axonal density, myelination, extracellular water content, and fiber geometry, particularly in regions with complex fiber crossings or fanning. Thus, these findings should not be attributed to a single cellular property of the STT. Instead, the observed associations likely reflect an aggregate microstructural signature pertaining to the conduction efficiency or structural integrity of the STT. Future studies using advanced diffusion models with improved biological specificity may help clarify the underlying microstructural substrates. Second, although our analyses identified spatially distributed STT segments spanning spinal, brainstem, and thalamic levels that were associated with pain sensitivity, the spatial resolution of spinal diffusion imaging remains limited. Confounds such as partial-volume effects, physiological motion, and susceptibility artifacts in spinal cord and lower brainstem imaging substantially limit precise localization of microstructural effects at specific laminar or sub-tract levels. Therefore, while our results support a tract-level interpretation, caution is warranted when inferring fine-scale anatomical subdivisions within the STT. Third, although we evaluated the generalizability of the STT-based pattern in independent clinical cohorts, the number of external datasets and the diversity of chronic pain conditions examined remain limited. Additionally, sample sizes were modest. Future studies with larger, well-characterized cohorts spanning a broader spectrum of chronic pain conditions are essential to validate the robustness and clinical applicability of STT-based microstructural biomarkers. Methods Participants Three DTI datasets acquired in China were analyzed in this study (Fig. 1 A). Dataset 1 was collected by the Institute of Psychology, Chinese Academy of Sciences. Participants were excluded if they had physical illnesses (e.g., brain tumors, hepatitis, epilepsy), chronic pain conditions (e.g., tension-type headache, fibromyalgia), neurological or psychiatric disorders, pregnancy, use of prescription medication within the past month, substance abuse (alcohol, nicotine, illicit drug), or claustrophobia precluding MRI. Based on these criteria, 53 individuals were initially recruited. Two participants were excluded due to poor DTI data quality, yielding a final cohort of 51 healthy participants (29 females and 22 males; mean age ± SD = 21.45 ± 3.21 years). Dataset 2 comprised patients with AZP recruited at Xuanwu Hospital from a longitudinal study 32 . Patients who fulfilled the AZP classification were eligible if they presented with the acute phase, defined as pain occurring within 30 days after the onset of the herpes zoster rash. A total of 26 patients were initially recruited. After excluding participants with missing DTI data (n = 2), poor image quality (n = 5), or incomplete clinical assessments (n = 1), the final sample comprised 18 patients (7 females and 11 males; mean age ± SD = 54.72 ± 10.61 years). Dataset 3 included patients aged 20–40 years with diarrhea-predominant IBS who met Rome IV criteria 33 . Eligibility required meeting predefined stool-form and abdominal pain thresholds during a two-week screening period. Detailed exclusion criteria and definitions of alarm features are provided in the Supplementary Information. Of the 80 patients initially recruited, four were excluded due to poor DTI data quality, resulting in a final sample of 76 patients with IBS (26 females and 50 males; mean age ± SD = 26.71 ± 6.73 years). Basic demographic and imaging information for all three datasets is summarized in Table S1 (Supporting Information). All participants gave written informed consent prior to the experiments. Local ethics committees approved the experimental procedures for the original studies (Dataset 1: Ethics Committee of the Institute of Psychology at the Chinese Academy of Science H19023; Dataset 2: Ethics Committee of the Beijing Xuanwu hospital No. 2024-005-001; Dataset 3: Ethics Committees of the Shaanxi Provincial Hospital of Traditional Chinese Medicine No. 2023-027). Sensory Stimulation and Pain Assessment Pain and somatosensory measurements included experimental paradigms in healthy participants (Dataset 1) and clinical pain ratings in patient cohorts (Datasets 2 and 3). Dataset 1 included both painful and non-painful stimuli, whereas Datasets 2 and 3 used standardized clinical pain intensity ratings. In Dataset 1, heat-pain threshold and tolerance were measured using a contact heat-evoked potential stimulator (CHEPS; Medoc Ltd., Ramat Yishai, Israel). A thermode (contact area: 573 mm2; diameter: 27 mm) was placed on the ventrolateral forearm, targeting the C5-C6 dermatomes. Starting from a baseline temperature of 32°C, thermal stimuli were applied using the method of limits with a ramp rate of 0.5°C/s. Participants were instructed to terminate the stimulus by pressing a button at initial perception of pain (threshold) or when the sensation became intolerable (tolerance). Each measurement was repeated four times, with termination temperatures recorded for subsequent analysis. Mechanical tactile and pain sensitivity were assessed using Semmes-Weinstein monofilaments (Touch-Test 20-Piece Kit; North Coast Medical, Inc., USA), with applied forces ranging from 0.008 to 300 g. Filaments were applied perpendicularly to the ipsilateral forearm for approximately one second per trial, with each application repeated three times. A stimulus was considered perceived if it was reported at least once across the three applications. Tactile and pain thresholds were determined using a standardized ascending-descending protocol, with averages calculated across four sequences. In Dataset 2, acute zoster pain intensity was assessed using a 0–10 Numeric Rating Scale (NRS), in accordance with the NeuPSIG grading system at the time of AZP diagnosis. In Dataset 3, abdominal pain severity was quantified using a 0–10 NRS, collected during the two-week screening phase prior to baseline assessment. For both clinical datasets, pain ratings served as the primary indices of clinical pain severity. Imaging Acquisition All three datasets underwent MRI scanning using identical parameters. Imaging was performed on a 3T MRI scanner (Siemens Prisma, Erlangen, Germany) equipped with a 64-channel head and neck coil and a 3D-MPRAGE sequence 34 . Participants were positioned with foam pads beneath and alongside the head, together with neck braces, to elevate and stabilize the head and to align the cervical spinal cord, thereby reducing motion and physiological noise. High-resolution anatomical images covering the whole brain and cervical spinal cord (C1-C7) were subsequently acquired in the sagittal orientation with the following parameters: phase-encoding direction: anterior-to-posterior (A→P), repetition time (TR) = 2300 ms, echo time (TE) = 2.28 ms, inversion time (TI) = 900 ms, field of view (FOV) = 192 × 288 × 288 mm3, phase FOV = 100%, voxel size = 1 × 1 × 1 mm3, and flip angle = 8°. DTI images were then acquired using a dedicated corticospinal protocol with simultaneous multi-slice (SMS) acquisition and the following parameters: TR = 7500 ms, TE = 87 ms, FOV = 230 × 230 mm2, number of slices = 70, voxel size = 1.2 × 1.2 × 4.0 mm3, base resolution = 192 × 192, diffusion directions = 32, b-value = 1500 s/mm2, GRAPPA factor = 2, acceleration factor = 2, and slice order = interleaved. Data were collected with phase-encoding direction = A→P, and an additional reversed phase-encoding b0 image (P→A, b = 0 s/mm2) was acquired for susceptibility-induced distortion correction. T1-weighted image preprocessing Three-dimensional T1-weighted images were corrected for bias fields using Advanced Normalization Tools with N4BiasFieldCorrection to reduce intensity inhomogeneity (Figure S6A) 35 . Images were then divided at the caudal end of the brainstem into brain and spinal-cord sections. Brain images were skull-stripped using deepbet to facilitate registration and alignment with DTI and template images 36 . Each skull-stripped image was visually inspected, and manual adjustments were made when automatic extraction was inadequate. Spinal-cord centerline detection and binary mask generation were performed using sct_deepseg spinalcord in the Spinal Cord Toolbox (SCT) 37 , with manual adjustments as needed to ensure segmentation quality. To identify spinal cord levels, seven intervertebral disc points (C1-C7) were manually labeled on T1-weighted images using SCT sct_label_utils. Spatial registration was performed with SCT sct_register_to_template, aligning individual spinal-cord images to the PAM50 template using the binary mask and intervertebral labels to guide the nonlinear registration 38 . This procedure generated a warp field that enabled bidirectional transformations between the template and individual structural images, ensuring precise alignment for subsequent DTI registration. DTI preprocessing DTI images were preprocessed using the FMRIB Software Library (FSL, v6.0.7) 39 . Two experienced neuroradiologists visually inspected the corticospinal diffusion-weighted images for noise artifacts (Figure S6B). Marchenko-Pastur principal component denoising was applied with MRtrix3 dwidenoise 40 , 41 , 42 , 43 , and Gibbs ringing artifacts were removed with MRtrix3 degibbs 43 , 44 . Susceptibility-induced distortions were corrected using paired b0 images with opposite phase-encoding directions (A→P and P→A) processed with FSL topup 45 , 46 . Eddy current distortions and head motion were corrected using FSL eddy 18 . Finally, FSL dtifit was applied to fit the preprocessed images to a diffusion tensor model, generating diffusion metrics including FA, MD, RD, and AD. The preprocessed images were divided at the caudal end of the brainstem for separate processing of the brain and spinal cord sections. For the brain section, skull stripping was performed with the FSL Brain Extraction Tool (bet), followed by alignment to the T1-weighted images using FSL flirt 47 . For the spinal cord section, DTI images were first separated into b0 and diffusion-weighted (DWI) volumes using SCT sct_dmri_separate_b0_and_dwi. Spinal cord centerline detection and binary mask generation from the mean DWI image were then performed with SCT sct_propseg 48 , with manual corrections applied when necessary. Finally, mean DWIs were aligned directly to the T1-weighted images using SCT sct_register_multimodal to ensure anatomical correspondence and to obtain registration parameters for subsequent tractography (Figure S6B). DTI Quality Check An automated DTI quality-control framework based on the FSL eddy ( https://fsl.fmrib.ox.ac.uk/fsl/docs/diffusion/eddyqc.html ) was employed 18. The voxel-wise tSNR was calculated as the mean signal intensity divided by its standard deviation 49 . Analysis of White-Matter Microstructural Properties Within the STT in Native Space The analysis was performed in two primary steps (Figure S7). First, for the brain, the thalamus mask from the Automated Anatomical Labeling 3 atlas was transformed into each participant's native DTI space by applying concatenated registration parameters from DTI to T1-weighted and from T1-wighted to template space (Figure S7), generating ROI A 50 . For the spinal cord, geometry was segmented and cervical vertebral levels were labeled on the T1-weighted images using SCT sct_label_vertebrae (Figure S7). Anatomical label maps derived from T1-weighted images were subsequently warped onto DTI images using the DTI-to-T1 registration parameters (Figure S7), producing ROI B (spinal segment C6) and ROI C (spinal segment C1) in native space. Second, spinothalamic tractography was conducted separately for the left and right hemispheres using MRtrix3 (v3.0.6) 43 . Tracking parameters included a seed point resolution of 1.2 × 1.2 × 4.0 mm3, a step size of 1 mm, an angle threshold of 15° or 30°, a FA threshold of 0.1, and a maximum streamline length of 250 mm. Tractography used ROI A (thalamus) as the seed region, with ROI B (spinal segment C6) and ROI C (spinal segment C1) as inclusion regions to extract the corresponding streamlines (Figure S7). Extracted streamlines were cropped such that the inferior boundary of the thalamus serving as the upper limit and the inferior boundary of spinal segment C6 serving as the lower limit. Based on the number of points along each streamline, the 70th percentile was used as the center, a window of ± 5 points was applied, and streamlines that were excessively long or short were excluded. Each streamline was divided into 185 points, as determined by the tracking step size. A representative streamline was selected and modeled using a continuous arc-length coordinate, which was subsequently transformed to align with all other streamlines for consistency. Given the slice thickness of the data, each spinothalamic streamline was resampled to 60 equally spaced points, and diffusion measures were obtained at each point along the streamlines, with mean values calculated across participants at each point (Figure S7). Multivariate Correlation Analysis Between White-Matter Microstructural Properties Within the STT and Pain Measurements PLSC analysis was performed to identify structural patterns within spinothalamic fiber streamlines and their multivariate relationships with pain measurements. The analysis was implemented using the myPLS toolbox ( https://github.com/MIPLabCH/myPLS ). PLSC extracts latent variables (LVs) that maximize covariance between two data matrices: an image matrix (X) and a behavioral matrix (Y) (Figure S8). Each LV comprises a behavioral salience vector, reflecting the contribution weights of behavioral measures to the image-behavior relationship, and an image-salience vector, representing the contribution weights of individual points along fiber streamlines. For each participant, an image score was computed by projecting image features onto the corresponding image-salience vector. High absolute image scores indicated strong contributions of specific image features to the image-behavior correlation, whereas scores near zero indicated negligible contribution. Correlation analyses between image scores and individual behavioral measures in matrix Y were performed to visualize the associations captured by the behavioral-salience vectors. To assess the reliability of salience estimates, bootstrap resampling with 5000 iterations was performed, and a BSR was computed as the observed salience divided by its standard error. To examine the pain selectivity of the relationship between white-matter microstructural properties within the STT and sensory measurements, we performed correlational analysis for tactile sensitivity in Dataset 1. To evaluate overlap between regions encoding pain and non-pain measurements, we calculated the Dice similarity coefficient between maps of spatial patterns. Multivariate Pattern Analysis for Predicting Individual Pain Measurements Two multivariate pattern analyses (MVPA) using LOOCV were performed, referred to as PLSC-PLSR and PLSC-PCR. The PLSC-PLSR analysis aimed to simultaneously predict ten pain measurements across individuals (Figure S9A), while the PLSC-PCR constructed a univariate prediction model across individuals (Figure S9B). LOOCV provides a robust estimate of model generalizability, particularly for studies with limited sample sizes. In each iteration, n-1 participants formed the training set, and the remaining participant constituted the test set. For the first model, PLSC was performed on each training set, generating BSR values through 5000 bootstrap resamples. Features with absolute BSR values consistently exceeding predetermined thresholds (1.96, 2.58, 3.29) were selected as candidate predictors. PLSR was subsequently applied using these features to build a multivariate prediction model, which was then tested on the held-out participant. For the PLSC-PCR model, the PLSC procedure was identical to that described above. PCR was performed by first extracting the first principal component (PC1) of pain measurements to capture the primary variance, followed by PLSR using PC1 and the selected diffusion features. Model performance was assessed across different combinations of BSR thresholds and numbers of PLSR components. Predicted pain scores were calculated as the dot product of model weights and STT microstructural features. The models were trained using heat and mechanical pain data from Dataset 1. To evaluate generalizability, the PLSC-PCR model was applied to Datasets 2 (AZP) and 3 (IBS). Because pain measures differed across datasets, predictions were aligned to the target behavioral space using PLSR to correct for scale discrepancies. Declarations Author Contributions Statement L-M W, B-L L, Z-X W, X-M L, and B N contributed to investigation and data curation; L-M W, Z-X W, X-M L, and B N contributed to methodology development; L-M W contributed to conceptualization, formal analysis, original draft writing, and visualization. Y-Z K and J-X L contributed to funding acquisition, supervision, methodology, resources, project administration, and review & editing; J-X L also contributed to conceptualization. All authors had access to a summary of all data. Y-Z K and J-X reviewed and verified the data in the study. All authors were responsible for the final decision to submit the manuscript. Competing Interests Statement All authors declare no competing interests. Acknowledgements This work was supported by the National Key R&D Program of China (No.2022YFC3500603), National Natural Science Foundation of China under Grant (82572197), Fundamental Research Funds for the Central Universities (QTZX25103), and Xidian University Specially Funded Project for Interdisciplinary Exploration (TZJH2024016). Thanks to Zhaoxing Wei, Yupu Zhu, Yunyun Duan, Xianchang Zhang, Yunyun Duan, Xiaomin Lin, and others for their support in data collection and methodology. Data Availability Statement The corresponding author takes full responsibility for the integrity of the data and the accuracy of the analyses. The anonymized dataset used in this study is available upon reasonable request from the corresponding author. References Nahman-Averbuch H et al (2019) Increased pain sensitivity but normal pain modulation in adolescents with migraine. Pain 160:1019–1028 Meints SM et al (2019) The relationship between catastrophizing and altered pain sensitivity in patients with chronic low back pain. Pain 160:833 Lee YC, Nassikas NJ, Clauw DJ (2011) The role of the central nervous system in the generation and maintenance of chronic pain in rheumatoid arthritis, osteoarthritis and fibromyalgia. Arthritis Res Therapy 13:211 Fields H (2004) State-dependent opioid control of pain. Nat Rev Neurosci 5:565–575 Lu Y et al (2013) A feed-forward spinal cord glycinergic neural circuit gates mechanical allodynia. J Clin Invest 123:4050–4062 Liu Y et al (2018) Touch and tactile neuropathic pain sensitivity are set by corticospinal projections. Nature 561:547–550 Gieré C, Melchior M, Dufour A, Poisbeau P (2021) Spinal integration of hot and cold nociceptive stimuli by wide-dynamic-range neurons in anesthetized adult rats. Pain Rep 6:e983 Seo JP, Jang SH (2013) Traumatic thalamic injury demonstrated by diffusion tensor tractography of the spinothalamic pathway. Brain Inj 27:749–753 Jang SH, Lee J, Yeo SS (2017) Central post-stroke pain due to injury of the spinothalamic tract in patients with cerebral infarction: a diffusion tensor tractography imaging study. Neural Regen Res 12:2021–2024 Allam AK et al (2025) The human spinothalamic tract: lessons from cordotomy. Brain Commun 7:fcaf237 Nathan PW (1990) Touch and surgical division of the anterior quadrant of the spinal cord. J Neurol Neurosurg Psychiatry 53:935–939 Chen H et al (2024) The functional and anatomical characterization of three spinal output pathways of the anterolateral tract. Cell Rep 43 Zhang Z, Zheng H, Yu Q, Jing X (2024) Understanding of Spinal Wide Dynamic Range Neurons and Their Modulation on Pathological Pain. J Pain Res 17:441–457 Dauleac C, Boukhari A, Jacquesson T, Criton G, Cotton F, Frindel C (2025) Stitching brain and spinal cord DTI using cross-correlation registration: toward an atlas of spinal tracts. Eur Radiol Experimental 9:112 Tinnermann A, Büchel C, Cohen-Adad J (2021) Cortico-spinal imaging to study pain. NeuroImage 224:117439 Wei Z et al (2025) CoSpine open access simultaneous cortico-spinal fMRI database of thermal pain and motor tasks. Sci Data 12:1696 Jenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM (2012) FSL Neuroimage 62:782–790 Andersson JLR, Sotiropoulos SN (2016) An integrated approach to correction for off-resonance effects and subject movement in diffusion MR imaging. NeuroImage 125:1063–1078 Yam MF, Loh YC, Tan CS, Khadijah Adam S, Abdul Manan N, Basir R (2018) General Pathways of Pain Sensation and the Major Neurotransmitters Involved in Pain Regulation. Int J Mol Sci 19 Ab Aziz CB, Ahmad AH (2006) The role of the thalamus in modulating pain. Malaysian J Med sciences: MJMS 13:11 Sciacca S, Lynch J, Davagnanam I, Barker R (2019) Midbrain, pons, and medulla: anatomy and syndromes. Radiographics 39:1110–1125 Basbaum AI, Bautista DM, Scherrer G, Julius D (2009) Cellular and Molecular Mechanisms of Pain. Cell 139:267–284 Yarmolinsky DA et al (2025) Differential modification of ascending spinal outputs in acute and chronic pain states. Neuron 113:1223–1239e1225 Lenz FA, Ohara S, Gracely RH, Dougherty PM, Patel SH (2004) Pain encoding in the human forebrain: binary and analog exteroceptive channels. J Neurosci 24:6540–6544 Sun YG (2025) Central neural circuits underlying itch sensation. Nat Rev Neurosci 26:765–777 Marshall AG, Sharma ML, Marley K, Olausson H, McGlone FP (2019) Spinal signalling of C-fiber mediated pleasant touch in humans. Elife 8 Chisholm KI, Lo Re L, Polgár E, Gutierrez-Mecinas M, Todd AJ, McMahon SB (2021) Encoding of cutaneous stimuli by lamina I projection neurons. Pain 162:2405–2417 Andrew D, Craig AD (2002) Quantitative responses of spinothalamic lamina I neurones to graded mechanical stimulation in the cat. J Physiol 545:913–931 Davidson S et al (2012) Pruriceptive spinothalamic tract neurons: physiological properties and projection targets in the primate. J Neurophysiol 108:1711–1723 Qiu Z, Liu T, Zeng C, Yang M, Xu X (2025) Local abnormal white matter microstructure in the spinothalamic tract in people with chronic neck and shoulder pain. Front NeuroSci 18–2024 Robertson JW, Aristi G, Hashmi JA (2023) White matter microstructure predicts measures of clinical symptoms in chronic back pain patients. Neuroimage Clin 37:103309 Lin X-M et al (2025) A corticospinal signature for interindividual pain sensitivity. Nat Commun Wang C et al (2025) The role of pain catastrophizing in modulating irritable bowel syndrome symptoms through brain-spinal cord integration. Meta-Radiology, 100187 Mugler IIIJP, Brookeman JR (1990) Three-dimensional magnetization-prepared rapid gradient-echo imaging (3D MP RAGE). Magn Reson Med 15:152–157 Tustison NJ et al (2010) N4ITK: improved N3 bias correction. IEEE Trans Med Imaging 29:1310–1320 Fisch L et al (2024) deepbet: Fast brain extraction of T1-weighted MRI using Convolutional Neural Networks. Comput Biol Med 179:108845 De Leener B et al (2017) Spinal Cord Toolbox, an open-source software for processing spinal cord MRI data. NeuroImage 145:24–43 De Leener B, Fonov VS, Collins DL, Callot V, Stikov N, Cohen-Adad J (2018) PAM50: Unbiased multimodal template of the brainstem and spinal cord aligned with the ICBM152 space. NeuroImage 165:170–179 Jenkinson M, Beckmann CF, Behrens TEJ, Woolrich MW, Smith SM (2012) FSL NeuroImage 62:782–790 Veraart J, Novikov DS, Christiaens D, Ades-Aron B, Sijbers J, Fieremans E (2016) Denoising of diffusion MRI using random matrix theory. NeuroImage 142:394–406 Cordero-Grande L, Christiaens D, Hutter J, Price AN, Hajnal JV (2019) Complex diffusion-weighted image estimation via matrix recovery under general noise models. NeuroImage 200, 391–404 Veraart J, Fieremans E, Novikov DS (2016) Diffusion MRI noise mapping using random matrix theory. Magn Reson Med 76:1582–1593 Tournier JD et al (2019) MRtrix3: A fast, flexible and open software framework for medical image processing and visualisation. NeuroImage 202:116137 Kellner E, Dhital B, Kiselev VG, Reisert M (2016) Gibbs-ringing artifact removal based on local subvoxel-shifts. Magn Reson Med 76:1574–1581 Andersson JL, Skare S, Ashburner J (2003) How to correct susceptibility distortions in spin-echo echo-planar images: application to diffusion tensor imaging. NeuroImage 20:870–888 Smith SM et al (2004) Advances in functional and structural MR image analysis and implementation as FSL. NeuroImage 23(Suppl 1):S208–219 Greve DN, Fischl B (2009) Accurate and robust brain image alignment using boundary-based registration. NeuroImage 48:63–72 Bédard S et al (2025) Towards contrast-agnostic soft segmentation of the spinal cord. Med Image Anal 101:103473 Welvaert M, Rosseel Y (2013) On the definition of signal-to-noise ratio and contrast-to-noise ratio for FMRI data. PLoS ONE 8:e77089 Rolls ET, Huang C-C, Lin C-P, Feng J, Joliot M (2020) Automated anatomical labelling atlas 3. NeuroImage 206, 116189 Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryInformation.docx Supplementary Information Cite Share Download PDF Status: Under Review 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-8640197","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":579403004,"identity":"b6367df6-041f-43fc-9a0a-bdd55e040cb3","order_by":0,"name":"Jixin Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA50lEQVRIiWNgGAWjYBACPmYGBoMEBgkGfgbGBokEsFgCfi1sMC2SDURrgTEMDjAANRKlhZ3HoOBhm0We8Y3kxhsPau4w8LPnGDD83IHPYTwGBglnJIrNzhxstkg49oxBsueNAWPvGUJaKiQStx1vbJNIYDvMYHAjx4CZsY2QFgOJxM3NjEAt/w4z2BOnBWjLBnagLYltQFskCGphKwD5JXEGyC+JfYd5JM48KzjYi0cLP//hbYY/2+oS+2ekP7z549thOf725I0PfuLRArLIAJnHAyIO4NXAwMD8gICCUTAKRsEoGOkAAIz7S3Ws0Zm1AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-4239-1201","institution":"School of Life Science and Technology, Xidian University","correspondingAuthor":true,"prefix":"","firstName":"Jixin","middleName":"","lastName":"Liu","suffix":""},{"id":579403005,"identity":"12644248-9dbf-44f1-92cc-c06c760a8fd8","order_by":1,"name":"Leiming Wu","email":"","orcid":"https://orcid.org/0009-0007-3575-678X","institution":"Xidian University","correspondingAuthor":false,"prefix":"","firstName":"Leiming","middleName":"","lastName":"Wu","suffix":""},{"id":579403006,"identity":"f1e2ec9a-9cb9-4845-ac86-be4746148240","order_by":2,"name":"Binglan Li","email":"","orcid":"","institution":"School of Life Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Binglan","middleName":"","lastName":"Li","suffix":""},{"id":579403007,"identity":"32a54fca-6843-44e6-a5be-93b2120b121b","order_by":3,"name":"Zhaoxing Wei","email":"","orcid":"https://orcid.org/0000-0001-5629-4384","institution":"Dartmouth College","correspondingAuthor":false,"prefix":"","firstName":"Zhaoxing","middleName":"","lastName":"Wei","suffix":""},{"id":579403008,"identity":"1ba5b636-4f10-4864-828c-fc8c216ad74e","order_by":4,"name":"Xiaomin Lin","email":"","orcid":"https://orcid.org/0000-0002-0081-662X","institution":"CAS Key Laboratory of Mental Health, Institute of Psychology, Chinese Academy of Sciences. Department of Psychology, University of Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Xiaomin","middleName":"","lastName":"Lin","suffix":""},{"id":579403009,"identity":"948dbccb-81c9-4666-8cad-19f195c6b88f","order_by":5,"name":"Bing Ni","email":"","orcid":"https://orcid.org/0000-0001-5548-9200","institution":"Functional neurosurgery Department, Xuanwu Hospital","correspondingAuthor":false,"prefix":"","firstName":"Bing","middleName":"","lastName":"Ni","suffix":""},{"id":579403010,"identity":"d58f5f5e-bd02-47ba-b6a1-3c120363dc51","order_by":6,"name":"Yazhuo Kong","email":"","orcid":"https://orcid.org/0000-0002-8249-4723","institution":"Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Yazhuo","middleName":"","lastName":"Kong","suffix":""}],"badges":[],"createdAt":"2026-01-19 14:09:36","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8640197/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8640197/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101189199,"identity":"99f87b76-3a7c-42f8-aa49-3bbeda4ea504","added_by":"auto","created_at":"2026-01-27 06:51:21","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":94615,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/6b9f857708f1babfd8b621a4.docx"},{"id":101189204,"identity":"3b4b3e24-ffeb-4103-b9c7-f021983d1254","added_by":"auto","created_at":"2026-01-27 06:51:21","extension":"tif","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8163100,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/1fe0be88a0ed986fde905d4f.tif"},{"id":101189202,"identity":"91fb74b7-6bb7-46f7-b269-a36b470def53","added_by":"auto","created_at":"2026-01-27 06:51:21","extension":"tif","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2047116,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/e317f10395e3754ee2992ec4.tif"},{"id":101189205,"identity":"f12330fe-c4cf-49df-92ad-7ea56f55e5ce","added_by":"auto","created_at":"2026-01-27 06:51:21","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7178740,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/b7479c139dc88b04bda6575e.tif"},{"id":101206809,"identity":"0d825dab-dc22-4002-9e34-a8b072c22dfa","added_by":"auto","created_at":"2026-01-27 09:56:47","extension":"tif","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2352552,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/ac957383650f9e54fd44b476.tif"},{"id":101189203,"identity":"29c94674-c09a-4b31-b36e-a5a53719348e","added_by":"auto","created_at":"2026-01-27 06:51:21","extension":"tif","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1809032,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/dee6d7d497f4b6c9f26def44.tif"},{"id":101189214,"identity":"4ff22377-ee31-4b81-992d-bc5f6c9b603d","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"json","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8426,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS26005387.json","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/d2a89a3c8093b5d840e39807.json"},{"id":101206178,"identity":"167fc54e-e5ab-4fe6-9ac9-0a98dc927d9a","added_by":"auto","created_at":"2026-01-27 09:55:33","extension":"docx","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29775461,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/2c8d4953d7339f8a7b9b5f15.docx"},{"id":101189215,"identity":"5c82f253-840f-4f92-bb9f-b4898ed90b02","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"xml","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":117731,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS260053870enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/e419ec67ee544927201d816c.xml"},{"id":101189224,"identity":"3c1bf82d-f7c7-4e52-88da-5224abf596c4","added_by":"auto","created_at":"2026-01-27 06:51:23","extension":"tif","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8163100,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/e9c92faf8c2e7ea78b053c78.tif"},{"id":101189222,"identity":"5c4b1543-b63a-44a5-95e8-615b6d6269c1","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"tif","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2047116,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/3d12d8d85e1d4e18162e4d81.tif"},{"id":101189212,"identity":"a30f8507-8614-463e-b4cf-8dc8fb4ff912","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"tif","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7178740,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/b93dca8096bee00351e9543f.tif"},{"id":101209476,"identity":"37b47604-14ee-4487-9791-4cce260fc597","added_by":"auto","created_at":"2026-01-27 10:17:03","extension":"tif","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2352552,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/589c692f55c492d94b2ddef9.tif"},{"id":101189218,"identity":"8fb1fd43-e774-4e58-a56a-21c56ea9cfe1","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"tif","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1809032,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.tif","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/20950d5eda0905667cf45a31.tif"},{"id":101189216,"identity":"71f630e6-9fb7-49c8-afeb-ce4840b62b51","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":215734,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/9bf7d8abe9c23d9407c80898.png"},{"id":101189217,"identity":"5a535996-97f3-4e2d-9e25-6a4f51d459f0","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":145957,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/eb9192eeed521c66b5faec4f.png"},{"id":101189219,"identity":"f7d43bb4-6995-40e4-a2ea-15cf55e53fe9","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":333650,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/be90d9ac14b39ec7de46c7a7.png"},{"id":101189209,"identity":"9455e502-9c31-4ee3-9270-07bc41b32f07","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":176593,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/af201bf70ba5a4fc75c2d840.png"},{"id":101206956,"identity":"5b432699-ecd5-4313-a249-1b1533f695dd","added_by":"auto","created_at":"2026-01-27 09:57:03","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":141924,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure5.png","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/6e11effb5ffec3a1652a7b30.png"},{"id":101189223,"identity":"36fbd7a1-5017-4bfa-9622-a19b20768dc7","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"xml","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":114886,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS260053870structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/d7f801c719a2172b5dfefcd7.xml"},{"id":101189220,"identity":"734acd6b-a4e3-4830-a562-7b68b8c70810","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"html","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":127583,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/6c863032d8e1e97bd4734418.html"},{"id":101206249,"identity":"1f6170c3-aa2f-4530-9c52-67d814481c60","added_by":"auto","created_at":"2026-01-27 09:55:46","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1249920,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eQuality control of DTI data. \u003c/strong\u003e(A) Geometric distortions were corrected using FSL. (B) Voxel-wise temporal signal-to-noise ratio was calculated as a summary measure of overall dataset quality. (C) Tracts of interest were visualized in native space, and the corresponding binary voxel-wise masks were registered and displayed in standard space. SNR, signal-to-noise ratio.\u003c/p\u003e","description":"","filename":"Figure1.tif.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/e93948eb167ecc8916cdb8d4.jpg"},{"id":101189198,"identity":"798e9f7f-7a53-4518-80b5-37a24985a94b","added_by":"auto","created_at":"2026-01-27 06:51:21","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1006042,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePain sensitivity across experimental conditions.\u003c/strong\u003e(A, B) Scatter plots showing individual left- and right-hand scores for heat pain threshold, heat pain tolerance and pain sensitivity at C5, C6 and C7 dermatomes. (C) Pearson correlation matrix among laboratory measures. *, p \u0026lt; 0.05; ***, p \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure2.tif.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/c06e49165f8193b2766d9114.jpg"},{"id":101189201,"identity":"3dad1ea9-1468-4059-9b2f-01d06187f677","added_by":"auto","created_at":"2026-01-27 06:51:21","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1937195,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelations between STT microstructural properties and pain measurements.\u003c/strong\u003e(A, B) Correlations between pain sensitivity and STT microstructural properties across all pain conditions in Dataset 1. (C) Overlap map of significant features across four white-matter measurements. To highlight anatomical commonalities within the STT, only features contributing more than 50% were retained. (D, E) Correlations between pain sensitivity and STT microstructural properties in heat or mechanical pain conditions in Dataset 1. FA, fractional anisotropy; AD, axial diffusivity; MD, mean diffusivity; RD, radial diffusivity.\u003c/p\u003e","description":"","filename":"Figure3.tif.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/6c1673f460a6c8abbe112402.jpg"},{"id":101189210,"identity":"8270ba2e-bb30-4c30-bd90-7ac072cc96da","added_by":"auto","created_at":"2026-01-27 06:51:22","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":707971,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelations between STT microstructural properties and tactile measurements.\u003c/strong\u003e(A) Behavioral salience of the STT microstructural properties as identified by PLSC analysis. (B) Spatial map about the regions of the STT with statistically significant correlations (|BSR| \u0026gt; 1.96) with tactile measurements. FA, Fractional anisotropy; AD, Axial diffusivity; MD, Mean diffusivity; RD, Radial diffusivity.\u003c/p\u003e","description":"","filename":"Figure4.tif.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/39c0a2bbbf913ba1c5ccec1c.jpg"},{"id":101189208,"identity":"0f42f16d-5177-486c-a6b0-733ee5e87916","added_by":"auto","created_at":"2026-01-27 06:51:21","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1060418,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDevelopment and performance of the prediction model. \u003c/strong\u003e(A, B) Development and performance of prediction models using STT microstructural properties as features. We developed two models based on PLSC-PLSR and PLSC-PCR, and their cross-validated performance in Dataset 1 was assessed. (C, D) Generalizability of the PLSC-PCR model in external datasets of patients with AZP (Dataset 2) and IBS (Dataset 3). PLSC, Partial least squares correlation; PCA, principal component analysis; PLSR, Partial least squares regression; HP, heat pain; TO, tolerance; TH, threshold.\u003c/p\u003e","description":"","filename":"Figure5.tif.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/159d5ce5b407154aa7e81f88.jpg"},{"id":101398835,"identity":"a8f775c7-4b95-47c8-bc37-7224e22d06a8","added_by":"auto","created_at":"2026-01-29 09:49:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6951554,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/908de4a0-a2d2-4870-b227-c04a614e4980.pdf"},{"id":101189207,"identity":"eaf1a772-4a30-495b-95e8-de0553e301ed","added_by":"auto","created_at":"2026-01-27 06:51:21","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":29775461,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-8640197/v1/74b272560cbda24136f353ee.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Spinothalamic Tract Microstructure as a Common Neural Substrate of Pain Sensitivity Across Modalities: A Combined Brain-Spinal Cord Diffusion Imaging Study","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePain sensitivity, which reflects the magnitude of an individual's response to noxious stimulation, is a fundamental dimension of human somatosensory function with broad implications for clinical practice and neuroscience research \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Although heat, mechanical, and chemical pain arise from distinct peripheral mechanisms, pain thresholds across modalities are moderately to strongly intercorrelated, suggesting shared central substrates that regulate generalized nociceptive responsiveness \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. A principal anatomical basis for such cross-modal integration lies in the ascending nociceptive pathways, in which diverse peripheral nociceptive inputs are relayed via the dorsal horn before projecting to the thalamus and cortex \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Despite the well-characterized anatomy of nociceptive pathways in animal models, the microstructural basis underlying shared pain sensitivity in humans remains insufficiently characterized.\u003c/p\u003e \u003cp\u003eAmong the ascending nociceptive pathways, the spinothalamic tract (STT) is a central conduit for transmitting nociceptive signals from the periphery to the brain and is therefore a prime candidate for shaping individual differences in pain sensitivity. Its role has been demonstrated in rodent models using neuroanatomical tracing and electrophysiological techniques \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Studies of spinothalamic neurons, including wide-dynamic-range and nociceptive‑specific populations, have exhibited graded responses to heat, mechanical, cold, and chemical stimulation \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Clinically, structural damage to the STT correlates with pain severity in several chronic pain syndromes, including chronic neck and shoulder pain, as well as neuropathic pain following spinal cord injury or stroke \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. These findings suggest that the STT is anatomically and physiologically positioned to integrate nociceptive signals across modalities. However, it remains unknown whether inter-individual variability in STT microstructure in humans constitutes a common biological substrate underlying pain sensitivity across different pain modalities.\u003c/p\u003e \u003cp\u003eBeyond nociception, the STT is a polymodal pathway for sensory integration, mediating sensations such as touch and itch \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Neurophysiological studies indicate that wide-dynamic-range neurons integrate both non-noxious and noxious inputs within overlapping populations, raising the question of whether STT microstructural properties are selective for pain or instead reflect a more general capacity for sensory encoding across domains \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. In humans, tract-level characterization of the STT has remained challenging because of technical limitations in neuroimaging \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Diffusion MRI studies of pain have predominantly focused on the brain, whereas reliable imaging of the cervical spinal cord has been hindered by physiological noise related to respiration and cardiac pulsation, motion artifacts, and susceptibility-induced signal loss. Consequently, the human spinothalamic tract has rarely been examined as a continuous pathway spanning the spinal cord and brain, hindering efforts to link tract-level microstructural variability along the ascending nociceptive pathway to individual differences in pain sensitivity.\u003c/p\u003e \u003cp\u003eWhile integrated brain-spinal cord imaging has been developed in functional MRI studies of pain \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, diffusion-based tract-level characterization of the spinothalamic pathway remains underdeveloped. In this study, we develop and apply a combined brain-spinal cord diffusion MRI framework to map STT microstructure across the entire central nervous system. We analyzed three datasets comprising healthy participants exposed to multiple laboratory-based painful stimuli, including heat threshold and tolerance, mechanical pain, and tactile sensations (Dataset 1), and two consisting of patients with acute zoster pain (AZP; Datasets 2) and irritable bowel syndrome (IBS; Datasets 3).We aimed to determine whether multimodal pain sensitivity shares a common microstructural correlate with the STT, whether this relationship is selective for painful stimuli, and whether STT microstructure can predict both experimental and clinical pain. We first applied partial least squares correlation (PLSC) to identify multivariate associations between STT microstructure and pain sensitivity across modalities in Dataset 1. Pain selectivity was then assessed by contrasting associations for painful versus non-painful tactile measurements. We next applied a multivariate prediction framework combining PLSC and partial least squares regression (PLSR) to evaluate whether this shared substrate could predict individual differences in multimodal pain sensitivity. Finally, after establishing the predictive performance in healthy participants, we evaluated model generalizability by predicting clinical pain severity in patients with AZP and IBS.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eQuality check\u003c/h2\u003e \u003cp\u003eAll diffusion datasets underwent a standardized combined brain-spinal cord preprocessing pipeline prior to tract reconstruction. Because long-echo-time diffusion acquisitions are particularly susceptible to magnetic susceptibility-induced distortions, eddy currents, and motion-related artifacts, we applied FSL's topup and eddy for susceptibility and eddy-current correction (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA) \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Furthermore, an automated DTI quality-control framework was implemented to compute the voxel-wise temporal signal-to-noise ratio (tSNR) for the b0 (b\u0026thinsp;=\u0026thinsp;0 s/mm2) images across all participants \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Across the full sample, the preprocessed diffusion data yielded a mean b0 tSNR of 35.09\u0026thinsp;\u0026plusmn;\u0026thinsp;7.42 (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation [SD]; Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). An example participant with the median b0 tSNR is presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Distortion correction effectively transformed raw images into undistorted anatomical space, removing major susceptibility- and eddy-related artifacts (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). SNR values were consistently higher in the brain than in the spinal cord (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). Following preprocessing, the brain-spinal cord diffusion data supported robust fiber-pathway reconstruction along the full brain-spinal axis. Individual STT streamlines were visualized in native space, while their corresponding voxel-wise tract masks were accurately registered to standard space (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eWhite-Matter Microstructure of the STT Reflects individual Pain Sensitivity\u003c/h3\u003e\n\u003cp\u003eDataset 1 included ten painful stimulus conditions, comprising heat-pain threshold and tolerance as well as mechanical-pain thresholds at the C5, C6, and C7 dermatomes bilaterally. Group-level pain ratings exhibited the expected variability, with heat-pain thresholds and tolerance of 41.04\u0026thinsp;\u0026plusmn;\u0026thinsp;3.31\u0026deg;C and 45.94\u0026thinsp;\u0026plusmn;\u0026thinsp;2.92\u0026deg;C, respectively, while mechanical-pain thresholds of 5.02\u0026thinsp;\u0026plusmn;\u0026thinsp;0.85, 5.17\u0026thinsp;\u0026plusmn;\u0026thinsp;0.95, and 5.09\u0026thinsp;\u0026plusmn;\u0026thinsp;0.97 a.u. for the C5, C6, and C7 dermatomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA; Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Pearson correlation analysis across all ten stimuli revealed two distinct clusters, showing strong within-modality correlations for heat and mechanical pain but substantially weaker cross-modality relationships (all FDR-correct p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDue to methodological challenges associated with registering brain-spinal cord imaging data between native and standard spaces, all tract-level analyses were conducted in native space. Diffusion metrics were extracted along each participant's STT streamlines, including fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD). These tract-level values formed the multivariate feature set used for subsequent analyses. Using PLSC, a robust multivariate association was observed between STT microstructural properties and pain sensitivity across individuals (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). The first latent variable (LV1) explained 63.08% of the cross-block covariance, indicating a consistent directional relationship across all ten pain measures. Image scores for LV1 showed significant associations with all pain measures (|bootstrap ratio [BSR]| \u0026gt; 1.96, p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). Significant microstructural features (|BSR| \u0026gt; 1.96) were distributed across the STT, including segments in the left C1, C4, and C6 levels, the right C2 and C5 levels, the substantia nigra, and the bilateral superior and inferior medullary reticular formations (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). To summarize regions consistently contributed across diffusion metrics, we generated an overlap map retaining features present in \u0026gt;\u0026thinsp;50% of features, thereby revealing an anatomical core shared across metrics (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe further investigated whether heat- and mechanical-pain modalities contributed similarly to the multivariate pattern. Modality-specific PLSC analyses identified significant associations for both heat and mechanical pain, revealing convergent anatomical loci, including left C6, right C5, right C1, and the substantia nigra (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD, \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE, and Figure S2). Dice similarity analyses indicated substantial overlap between the full-model results (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB) and the modality-specific maps, with Dice coefficients of 0.41 for heat pain and 0.77 for mechanical pain (Figures S3A, S3B), indicating shared yet distinct anatomical substrates across modalities. To account for demographic differences (Dataset 1 showed sex equivalence: χ2(1)\u0026thinsp;=\u0026thinsp;0.96, p\u0026thinsp;=\u0026thinsp;0.33, but a significant age difference: t(50)\u0026thinsp;=\u0026thinsp;47.75, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, Cohen's d\u0026thinsp;=\u0026thinsp;3.21), we reexamined tract-pain correlations using partial correlations controlling for sex and age. Although effect sizes changed slightly, the spatial pattern remained highly consistent with the original results (Figure S3), suggesting that demographic variation did not account for the observed microstructure-pain associations.\u003c/p\u003e\n\u003ch3\u003eWhite-Matter Microstructure of the STT Differentially Reflect Pain and Tactile Sensitivity\u003c/h3\u003e\n\u003cp\u003eHaving demonstrated that STT microstructural properties are associated with pain measurements, we next examined whether this relationship observed in Dataset 1 was specific to pain by assessing associations between STT microstructural properties and tactile sensitivity. Although the segregation of pain and tactile processing has been proposed in prior research \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, how this distinction is reflected in the microstructural organization of the human STT remains poor characterized. PLSC analysis identified a distinct multivariate pattern linked to tactile sensitivity (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). Notably, the spatial distribution of tactile-related regions showed minimal overlap with the pain-related regions identified earlier (pain: Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB; tactile: Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB), as quantified by a Dice coefficient of 0.28. This association remained consistent after controlling for sex and age (Figure S4).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eMultivariate STT Microstructural Pattern Predicts Individual Pain Sensitivity\u003c/h3\u003e\n\u003cp\u003eA multivariate pattern analysis (MVPA) framework was developed based on STT microstructural properties, and leave-one-out cross-validation (LOOCV) was employed to simultaneously predict individual pain sensitivity across all ten pain measurements. For Dataset 1, data were iteratively partitioned into a discovery set (n\u0026thinsp;=\u0026thinsp;50) and a holdout test set (n\u0026thinsp;=\u0026thinsp;1), such that each participant served as the holdout test case once. Initially, the model utilized PLSC to identify the multivariate STT microstructural pattern jointly associated with all pain measurements, and then used PLSR to simultaneously predict each pain measurement in the holdout participant (PLSC-PLSR; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). Across holdout tests, the PLSC-PLSR model simultaneously and significantly predicted all ten pain measurements, as shown by LOOCV performance across participants (n\u0026thinsp;=\u0026thinsp;51; p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo derive a unified behavioral target reflecting shared variance across pain modalities, principal component regression (PCR) was applied. Principal component analysis (PCA) was performed on the ten pain measurements to extract a latent pain-sensitivity score that captured their common variance (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). The first principal component was used as the outcome variable in a regression model with STT microstructural features as predictors. The resulting PCR model showed robust predictive performance under LOOCV (n\u0026thinsp;=\u0026thinsp;51; r\u0026thinsp;=\u0026thinsp;0.406; p\u0026thinsp;=\u0026thinsp;0.003; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB), supporting the existence of a shared STT-based substrate underlying individual differences in pain sensitivity. Finally, to assess modality specificity, we applied the same model to tactile sensitivity measurements. Consistent with earlier results, the pain-derived STT pattern did not predict tactile sensitivity (p\u0026thinsp;\u0026gt;\u0026thinsp;0.05; Figure S5), supporting its selectivity for nociceptive rather than somatosensory-general processing.\u003c/p\u003e\n\u003ch3\u003eSTT-Based Pain Sensitivity Pattern Generalizes to Clinical Pain in AZP and IBS\u003c/h3\u003e\n\u003cp\u003eHaving established in Dataset 1 that STT microstructure predicts individual pain sensitivity, encompassing both modality-specific measures and a unified latent pain-sensitivity score, we applied the PLSC-PCR pain-sensitivity model to two independent clinical cohorts. In patients with AZP (Dataset 2), the model significantly predicted the intensity of acute zoster pain intensity (n\u0026thinsp;=\u0026thinsp;18; r\u0026thinsp;=\u0026thinsp;0.710; p\u0026thinsp;=\u0026thinsp;0.003; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC). Similarly, in patients with IBS (Dataset 3), the model significantly predicted the severity of abdominal pain severity (n\u0026thinsp;=\u0026thinsp;76; r\u0026thinsp;=\u0026thinsp;0.333; p\u0026thinsp;=\u0026thinsp;0.003; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eUsing combined brain-spinal cord diffusion MRI, this study provides a tract-level characterization of STT microstructure across the central nervous system and demonstrates that inter-individual variability along this pathway systematically correlates with pain sensitivity across multiple experimental modalities. Three key findings emerged. First, strong associations were identified between STT microstructural properties and pain sensitivity across ten experimentally derived measures, spanning both thermal and mechanical pain modalities, supporting the existence of a shared structural substrate underlying generalized nociceptive responsiveness. Second, although STT microstructure was associated with tactile sensitivity, pain and tactile measures exhibited largely distinct multivariate patterns, indicating a modality-differentiated organization within the tract. Third, an STT-derived multivariate pattern predicted individual differences in experimental pain sensitivity in healthy participants and generalized to predict clinical pain severity in independent cohorts of patients with AZP and IBS. Collectively, these findings identify the STT microstructure as a common neuroanatomical substrate of pain sensitivity and establish a structural framework linking individual pain vulnerability to ascending nociceptive pathways in humans.\u003c/p\u003e\n\u003ch3\u003eSTT Microstructure as a Shared Neuroanatomical Substrate for Multimodal Pain Sensitivity\u003c/h3\u003e\n\u003cp\u003eThe STT serves as the principal ascending pathway responsible for transmitting nociceptive information from the spinal dorsal horn to supraspinal structures, particularly the brainstem and thalamus \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Our results demonstrate that inter-individual variability in STT white-matter microstructure is strongly associated with pain sensitivity across multiple experimental modalities, including heat pain threshold, heat pain tolerance, and mechanical pain sensitivity. Importantly, these associations were not confined to any single spinal segment but instead reflected a tract-level pattern distributed along the STT, spanning cervical spinal levels, and extending into brainstem and thalamic regions. This spatially extended and modality-convergent pattern supports the interpretation that STT microstructure constitutes a unified tract-level substrate underlying generalized pain sensitivity in humans, rather than modalities or segment-specific phenomena.\u003c/p\u003e \u003cp\u003eHeat and mechanical pain rely on distinct peripheral transduction mechanisms and afferent fibers \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, yet their sensitivities converge in STT microstructural features. Further evidence indicates that high-threshold nociceptive (NOX) neurons in the spinal cord's anterolateral tract (including the STT) are broadly responsive to both thermal and mechanical stimuli and can be activated by low-intensity warmth following peripheral sensitization \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. This convergence suggests that STT microstructure does not primarily encode modality-specific pain representations \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e, but instead reflects a higher-order property of the nociceptive system, such as general nociceptive gain or amplification capacity. Individuals with specific STT microstructural profiles may therefore exhibit heightened responsiveness across various pain modalities, providing a structural basis for generalized pain susceptibility. Extensive animal electrophysiological studies have established that spinothalamic neurons encode nociceptive input in a graded, intensity-dependent manner across thermal and mechanical modalities, providing a cellular and circuit-level account of how pain signals are centrally transmitted \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. However, these mechanistic studies were not designed to explain why human pain sensitivity varies so markedly across individuals. By integrating tract-level STT microstructural analysis with multivariate modeling in humans, the present study complements animal findings by demonstrating that inter-individual differences in STT organization underlie variability in pain sensitivity, thereby bridging the gap between cellular nociceptive mechanisms and observed behavioral differences.\u003c/p\u003e\n\u003ch3\u003eModality-differentiated organization of pain and tactile sensitivity within the STT\u003c/h3\u003e\n\u003cp\u003eAlthough the spinothalamic tract is traditionally regarded as a major nociceptive pathway, it is also known to convey non-noxious somatosensory signals such as touch and itch \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Anatomical and physiological studies have shown that STT fibers originate from distinct spinal laminae and follow partially differentiated trajectories within the tract \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, which reflects heterogeneity in the types of sensory information transmitted, rather than an absolute segregation between painful and non-painful modalities \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Consistent with this background, our results indicate that pain and tactile sensitivity are underpinned by partially distinct microstructural features within the STT. While white-matter microstructural properties within the tract were significantly mapped onto both pain and tactile sensitivity, these associations were associated with largely distinct multivariate patterns. The spatial overlap between pain-related and tactile-related STT features was limited (Dice = 0.28), suggesting only modest convergence at the tract level. Crucially, this dissociation was further bolstered by the finding that the STT-based pattern derived from pain sensitivity failed to predict individual differences in tactile sensitivity, despite its robust predictive utility for pain across multiple modalities. This pattern of partial segregation aligns with prior experimental evidence demonstrating functional heterogeneity within the STT. For instance, electrophysiological recordings have revealed that individual spinothalamic neurons can exhibit distinct response profiles to graded mechanical and thermal stimulation, including separate classes of nociception-specific and polymodal units, each characterized by distinct stimulus-response curves \u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. In primates, spinothalamic neurons responsive to pruritic versus other somatosensory inputs project to partially distinct thalamic nuclei, further illustrating functional heterogeneity within the pathway \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. By demonstrating that such functional distinctions are reflected in separable multivariate microstructural patterns in humans, our findings provide tract-level evidence that STT organization cannot be reduced to a single, modality-independent encoding of sensory intensity. Instead, they suggest that alterations in STT microstructure may selectively influence pain vulnerability without uniformly affecting other somatosensory functions.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eFrom Experimental Pain to Clinical Pain: Generalizability of STT-Based Patterns\u003c/h2\u003e \u003cp\u003eBeyond correlational findings in healthy participants, the multivariate STT pattern derived from Dataset 1 was applied to two independent patient cohorts characterized by distinct etiologies and symptom profiles. Despite heterogeneity in pain type, clinical context, rating scales, and MRI acquisition parameters, the model significantly predicted clinical pain severity in both AZP and IBS. These findings converge with accumulating neuroimaging evidence that white matter microstructural alterations within the ascending nociceptive pathway are clinically meaningful. For example, patients with chronic neck and shoulder pain exhibit localized reductions in STT FA and increases in MD relative to healthy controls, with these changes correlating with both pain intensity and duration \u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. In central post-stroke pain, diffusion tensor tractography has reported decreased FA and tract volume in the affected STT, suggesting a role for STT injury in the pathogenesis of persistent pain following cerebral infarction \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Additionally, patients with chronic back pain show reduced fiber density in white-matter tracts including the spinothalamic tracts, with microstructural differences related to clinical symptom measures \u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. These clinical examples illustrate that white-matter microstructural features similar to those identified in our predictive model are systematically associated with pain severity across diverse chronic pain conditions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eLimitation\u003c/h2\u003e \u003cp\u003eSeveral limitations of this study should be acknowledged. First, although DTI provides a unique, noninvasive window into in-vivo white-matter microstructure, the biological specificity of conventional DTI metrics remains limited. Metrics such as FA, MD, RD, and AD represent a composite of microstructural features, including axonal density, myelination, extracellular water content, and fiber geometry, particularly in regions with complex fiber crossings or fanning. Thus, these findings should not be attributed to a single cellular property of the STT. Instead, the observed associations likely reflect an aggregate microstructural signature pertaining to the conduction efficiency or structural integrity of the STT. Future studies using advanced diffusion models with improved biological specificity may help clarify the underlying microstructural substrates. Second, although our analyses identified spatially distributed STT segments spanning spinal, brainstem, and thalamic levels that were associated with pain sensitivity, the spatial resolution of spinal diffusion imaging remains limited. Confounds such as partial-volume effects, physiological motion, and susceptibility artifacts in spinal cord and lower brainstem imaging substantially limit precise localization of microstructural effects at specific laminar or sub-tract levels. Therefore, while our results support a tract-level interpretation, caution is warranted when inferring fine-scale anatomical subdivisions within the STT. Third, although we evaluated the generalizability of the STT-based pattern in independent clinical cohorts, the number of external datasets and the diversity of chronic pain conditions examined remain limited. Additionally, sample sizes were modest. Future studies with larger, well-characterized cohorts spanning a broader spectrum of chronic pain conditions are essential to validate the robustness and clinical applicability of STT-based microstructural biomarkers.\u003c/p\u003e \u003c/div\u003e "},{"header":"Methods","content":"\u003ch2\u003eParticipants\u003c/h2\u003e\u003cp\u003eThree DTI datasets acquired in China were analyzed in this study (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Dataset 1 was collected by the Institute of Psychology, Chinese Academy of Sciences. Participants were excluded if they had physical illnesses (e.g., brain tumors, hepatitis, epilepsy), chronic pain conditions (e.g., tension-type headache, fibromyalgia), neurological or psychiatric disorders, pregnancy, use of prescription medication within the past month, substance abuse (alcohol, nicotine, illicit drug), or claustrophobia precluding MRI. Based on these criteria, 53 individuals were initially recruited. Two participants were excluded due to poor DTI data quality, yielding a final cohort of 51 healthy participants (29 females and 22 males; mean age ± SD = 21.45 ± 3.21 years). Dataset 2 comprised patients with AZP recruited at Xuanwu Hospital from a longitudinal study \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. Patients who fulfilled the AZP classification were eligible if they presented with the acute phase, defined as pain occurring within 30 days after the onset of the herpes zoster rash. A total of 26 patients were initially recruited. After excluding participants with missing DTI data (n = 2), poor image quality (n = 5), or incomplete clinical assessments (n = 1), the final sample comprised 18 patients (7 females and 11 males; mean age ± SD = 54.72 ± 10.61 years). Dataset 3 included patients aged 20–40 years with diarrhea-predominant IBS who met Rome IV criteria \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Eligibility required meeting predefined stool-form and abdominal pain thresholds during a two-week screening period. Detailed exclusion criteria and definitions of alarm features are provided in the Supplementary Information. Of the 80 patients initially recruited, four were excluded due to poor DTI data quality, resulting in a final sample of 76 patients with IBS (26 females and 50 males; mean age ± SD = 26.71 ± 6.73 years). Basic demographic and imaging information for all three datasets is summarized in Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e (Supporting Information).\u003c/p\u003e\u003cp\u003e All participants gave written informed consent prior to the experiments. Local ethics committees approved the experimental procedures for the original studies (Dataset 1: Ethics Committee of the Institute of Psychology at the Chinese Academy of Science H19023; Dataset 2: Ethics Committee of the Beijing Xuanwu hospital No. 2024-005-001; Dataset 3: Ethics Committees of the Shaanxi Provincial Hospital of Traditional Chinese Medicine No. 2023-027).\u003c/p\u003e\u003ch2\u003eSensory Stimulation and Pain Assessment\u003c/h2\u003e\u003cp\u003ePain and somatosensory measurements included experimental paradigms in healthy participants (Dataset 1) and clinical pain ratings in patient cohorts (Datasets 2 and 3). Dataset 1 included both painful and non-painful stimuli, whereas Datasets 2 and 3 used standardized clinical pain intensity ratings.\u003c/p\u003e\u003cp\u003eIn Dataset 1, heat-pain threshold and tolerance were measured using a contact heat-evoked potential stimulator (CHEPS; Medoc Ltd., Ramat Yishai, Israel). A thermode (contact area: 573 mm2; diameter: 27 mm) was placed on the ventrolateral forearm, targeting the C5-C6 dermatomes. Starting from a baseline temperature of 32°C, thermal stimuli were applied using the method of limits with a ramp rate of 0.5°C/s. Participants were instructed to terminate the stimulus by pressing a button at initial perception of pain (threshold) or when the sensation became intolerable (tolerance). Each measurement was repeated four times, with termination temperatures recorded for subsequent analysis. Mechanical tactile and pain sensitivity were assessed using Semmes-Weinstein monofilaments (Touch-Test 20-Piece Kit; North Coast Medical, Inc., USA), with applied forces ranging from 0.008 to 300 g. Filaments were applied perpendicularly to the ipsilateral forearm for approximately one second per trial, with each application repeated three times. A stimulus was considered perceived if it was reported at least once across the three applications. Tactile and pain thresholds were determined using a standardized ascending-descending protocol, with averages calculated across four sequences. In Dataset 2, acute zoster pain intensity was assessed using a 0–10 Numeric Rating Scale (NRS), in accordance with the NeuPSIG grading system at the time of AZP diagnosis. In Dataset 3, abdominal pain severity was quantified using a 0–10 NRS, collected during the two-week screening phase prior to baseline assessment. For both clinical datasets, pain ratings served as the primary indices of clinical pain severity.\u003c/p\u003e\u003ch2\u003eImaging Acquisition\u003c/h2\u003e\u003cp\u003eAll three datasets underwent MRI scanning using identical parameters. Imaging was performed on a 3T MRI scanner (Siemens Prisma, Erlangen, Germany) equipped with a 64-channel head and neck coil and a 3D-MPRAGE sequence \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. Participants were positioned with foam pads beneath and alongside the head, together with neck braces, to elevate and stabilize the head and to align the cervical spinal cord, thereby reducing motion and physiological noise. High-resolution anatomical images covering the whole brain and cervical spinal cord (C1-C7) were subsequently acquired in the sagittal orientation with the following parameters: phase-encoding direction: anterior-to-posterior (A→P), repetition time (TR) = 2300 ms, echo time (TE) = 2.28 ms, inversion time (TI) = 900 ms, field of view (FOV) = 192 × 288 × 288 mm3, phase FOV = 100%, voxel size = 1 × 1 × 1 mm3, and flip angle = 8°.\u003c/p\u003e\u003cp\u003eDTI images were then acquired using a dedicated corticospinal protocol with simultaneous multi-slice (SMS) acquisition and the following parameters: TR = 7500 ms, TE = 87 ms, FOV = 230 × 230 mm2, number of slices = 70, voxel size = 1.2 × 1.2 × 4.0 mm3, base resolution = 192 × 192, diffusion directions = 32, b-value = 1500 s/mm2, GRAPPA factor = 2, acceleration factor = 2, and slice order = interleaved. Data were collected with phase-encoding direction = A→P, and an additional reversed phase-encoding b0 image (P→A, b = 0 s/mm2) was acquired for susceptibility-induced distortion correction.\u003c/p\u003e\u003ch2\u003eT1-weighted image preprocessing\u003c/h2\u003e\u003cp\u003eThree-dimensional T1-weighted images were corrected for bias fields using Advanced Normalization Tools with N4BiasFieldCorrection to reduce intensity inhomogeneity (Figure S6A) \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Images were then divided at the caudal end of the brainstem into brain and spinal-cord sections. Brain images were skull-stripped using deepbet to facilitate registration and alignment with DTI and template images \u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Each skull-stripped image was visually inspected, and manual adjustments were made when automatic extraction was inadequate. Spinal-cord centerline detection and binary mask generation were performed using sct_deepseg spinalcord in the Spinal Cord Toolbox (SCT) \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e, with manual adjustments as needed to ensure segmentation quality. To identify spinal cord levels, seven intervertebral disc points (C1-C7) were manually labeled on T1-weighted images using SCT sct_label_utils. Spatial registration was performed with SCT sct_register_to_template, aligning individual spinal-cord images to the PAM50 template using the binary mask and intervertebral labels to guide the nonlinear registration \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. This procedure generated a warp field that enabled bidirectional transformations between the template and individual structural images, ensuring precise alignment for subsequent DTI registration.\u003c/p\u003e\u003ch2\u003eDTI preprocessing\u003c/h2\u003e\u003cp\u003eDTI images were preprocessed using the FMRIB Software Library (FSL, v6.0.7) \u003csup\u003e39\u003c/sup\u003e. Two experienced neuroradiologists visually inspected the corticospinal diffusion-weighted images for noise artifacts (Figure S6B). Marchenko-Pastur principal component denoising was applied with MRtrix3 dwidenoise \u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e, and Gibbs ringing artifacts were removed with MRtrix3 degibbs \u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Susceptibility-induced distortions were corrected using paired b0 images with opposite phase-encoding directions (A→P and P→A) processed with FSL topup \u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. Eddy current distortions and head motion were corrected using FSL eddy \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Finally, FSL dtifit was applied to fit the preprocessed images to a diffusion tensor model, generating diffusion metrics including FA, MD, RD, and AD.\u003c/p\u003e\u003cp\u003eThe preprocessed images were divided at the caudal end of the brainstem for separate processing of the brain and spinal cord sections. For the brain section, skull stripping was performed with the FSL Brain Extraction Tool (bet), followed by alignment to the T1-weighted images using FSL flirt \u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. For the spinal cord section, DTI images were first separated into b0 and diffusion-weighted (DWI) volumes using SCT sct_dmri_separate_b0_and_dwi. Spinal cord centerline detection and binary mask generation from the mean DWI image were then performed with SCT sct_propseg \u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e, with manual corrections applied when necessary. Finally, mean DWIs were aligned directly to the T1-weighted images using SCT sct_register_multimodal to ensure anatomical correspondence and to obtain registration parameters for subsequent tractography (Figure S6B).\u003c/p\u003e\u003ch2\u003eDTI Quality Check\u003c/h2\u003e\u003cp\u003eAn automated DTI quality-control framework based on the FSL eddy (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://fsl.fmrib.ox.ac.uk/fsl/docs/diffusion/eddyqc.html\u003c/span\u003e\u003cspan address=\"https://fsl.fmrib.ox.ac.uk/fsl/docs/diffusion/eddyqc.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was employed 18. The voxel-wise tSNR was calculated as the mean signal intensity divided by its standard deviation \u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003ch2\u003eAnalysis of White-Matter Microstructural Properties Within the STT in Native Space\u003c/h2\u003e\u003cp\u003eThe analysis was performed in two primary steps (Figure S7). First, for the brain, the thalamus mask from the Automated Anatomical Labeling 3 atlas was transformed into each participant's native DTI space by applying concatenated registration parameters from DTI to T1-weighted and from T1-wighted to template space (Figure S7), generating ROI A \u003csup\u003e50\u003c/sup\u003e. For the spinal cord, geometry was segmented and cervical vertebral levels were labeled on the T1-weighted images using SCT sct_label_vertebrae (Figure S7). Anatomical label maps derived from T1-weighted images were subsequently warped onto DTI images using the DTI-to-T1 registration parameters (Figure S7), producing ROI B (spinal segment C6) and ROI C (spinal segment C1) in native space.\u003c/p\u003e\u003cp\u003eSecond, spinothalamic tractography was conducted separately for the left and right hemispheres using MRtrix3 (v3.0.6) \u003csup\u003e43\u003c/sup\u003e. Tracking parameters included a seed point resolution of 1.2 × 1.2 × 4.0 mm3, a step size of 1 mm, an angle threshold of 15° or 30°, a FA threshold of 0.1, and a maximum streamline length of 250 mm. Tractography used ROI A (thalamus) as the seed region, with ROI B (spinal segment C6) and ROI C (spinal segment C1) as inclusion regions to extract the corresponding streamlines (Figure S7). Extracted streamlines were cropped such that the inferior boundary of the thalamus serving as the upper limit and the inferior boundary of spinal segment C6 serving as the lower limit. Based on the number of points along each streamline, the 70th percentile was used as the center, a window of ± 5 points was applied, and streamlines that were excessively long or short were excluded. Each streamline was divided into 185 points, as determined by the tracking step size. A representative streamline was selected and modeled using a continuous arc-length coordinate, which was subsequently transformed to align with all other streamlines for consistency. Given the slice thickness of the data, each spinothalamic streamline was resampled to 60 equally spaced points, and diffusion measures were obtained at each point along the streamlines, with mean values calculated across participants at each point (Figure S7).\u003c/p\u003e\u003ch2\u003eMultivariate Correlation Analysis Between White-Matter Microstructural Properties Within the STT and Pain Measurements\u003c/h2\u003e\u003cp\u003ePLSC analysis was performed to identify structural patterns within spinothalamic fiber streamlines and their multivariate relationships with pain measurements. The analysis was implemented using the myPLS toolbox (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/MIPLabCH/myPLS\u003c/span\u003e\u003cspan address=\"https://github.com/MIPLabCH/myPLS\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). PLSC extracts latent variables (LVs) that maximize covariance between two data matrices: an image matrix (X) and a behavioral matrix (Y) (Figure S8). Each LV comprises a behavioral salience vector, reflecting the contribution weights of behavioral measures to the image-behavior relationship, and an image-salience vector, representing the contribution weights of individual points along fiber streamlines. For each participant, an image score was computed by projecting image features onto the corresponding image-salience vector. High absolute image scores indicated strong contributions of specific image features to the image-behavior correlation, whereas scores near zero indicated negligible contribution. Correlation analyses between image scores and individual behavioral measures in matrix Y were performed to visualize the associations captured by the behavioral-salience vectors. To assess the reliability of salience estimates, bootstrap resampling with 5000 iterations was performed, and a BSR was computed as the observed salience divided by its standard error.\u003c/p\u003e\u003cp\u003eTo examine the pain selectivity of the relationship between white-matter microstructural properties within the STT and sensory measurements, we performed correlational analysis for tactile sensitivity in Dataset 1. To evaluate overlap between regions encoding pain and non-pain measurements, we calculated the Dice similarity coefficient between maps of spatial patterns.\u003c/p\u003e\u003ch2\u003eMultivariate Pattern Analysis for Predicting Individual Pain Measurements\u003c/h2\u003e\u003cp\u003eTwo multivariate pattern analyses (MVPA) using LOOCV were performed, referred to as PLSC-PLSR and PLSC-PCR. The PLSC-PLSR analysis aimed to simultaneously predict ten pain measurements across individuals (Figure S9A), while the PLSC-PCR constructed a univariate prediction model across individuals (Figure S9B). LOOCV provides a robust estimate of model generalizability, particularly for studies with limited sample sizes. In each iteration, n-1 participants formed the training set, and the remaining participant constituted the test set. For the first model, PLSC was performed on each training set, generating BSR values through 5000 bootstrap resamples. Features with absolute BSR values consistently exceeding predetermined thresholds (1.96, 2.58, 3.29) were selected as candidate predictors. PLSR was subsequently applied using these features to build a multivariate prediction model, which was then tested on the held-out participant.\u003c/p\u003e\u003cp\u003eFor the PLSC-PCR model, the PLSC procedure was identical to that described above. PCR was performed by first extracting the first principal component (PC1) of pain measurements to capture the primary variance, followed by PLSR using PC1 and the selected diffusion features. Model performance was assessed across different combinations of BSR thresholds and numbers of PLSR components. Predicted pain scores were calculated as the dot product of model weights and STT microstructural features. The models were trained using heat and mechanical pain data from Dataset 1. To evaluate generalizability, the PLSC-PCR model was applied to Datasets 2 (AZP) and 3 (IBS). Because pain measures differed across datasets, predictions were aligned to the target behavioral space using PLSR to correct for scale discrepancies.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eAuthor Contributions Statement\u003c/h2\u003e \u003cp\u003eL-M W, B-L L, Z-X W, X-M L, and B N contributed to investigation and data curation; L-M W, Z-X W, X-M L, and B N contributed to methodology development; L-M W contributed to conceptualization, formal analysis, original draft writing, and visualization. Y-Z K and J-X L contributed to funding acquisition, supervision, methodology, resources, project administration, and review \u0026amp; editing; J-X L also contributed to conceptualization. All authors had access to a summary of all data. Y-Z K and J-X reviewed and verified the data in the study. All authors were responsible for the final decision to submit the manuscript.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eCompeting Interests Statement\u003c/h2\u003e \u003cp\u003eAll authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eThis work was supported by the National Key R\u0026amp;D Program of China (No.2022YFC3500603), National Natural Science Foundation of China under Grant (82572197), Fundamental Research Funds for the Central Universities (QTZX25103), and Xidian University Specially Funded Project for Interdisciplinary Exploration (TZJH2024016). Thanks to Zhaoxing Wei, Yupu Zhu, Yunyun Duan, Xianchang Zhang, Yunyun Duan, Xiaomin Lin, and others for their support in data collection and methodology.\u003c/p\u003e\u003ch2\u003eData Availability Statement\u003c/h2\u003e \u003cp\u003eThe corresponding author takes full responsibility for the integrity of the data and the accuracy of the analyses. The anonymized dataset used in this study is available upon reasonable request from the corresponding author.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eNahman-Averbuch H et al (2019) Increased pain sensitivity but normal pain modulation in adolescents with migraine. Pain 160:1019\u0026ndash;1028\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeints SM et al (2019) The relationship between catastrophizing and altered pain sensitivity in patients with chronic low back pain. Pain 160:833\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee YC, Nassikas NJ, Clauw DJ (2011) The role of the central nervous system in the generation and maintenance of chronic pain in rheumatoid arthritis, osteoarthritis and fibromyalgia. Arthritis Res Therapy 13:211\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFields H (2004) State-dependent opioid control of pain. Nat Rev Neurosci 5:565\u0026ndash;575\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLu Y et al (2013) A feed-forward spinal cord glycinergic neural circuit gates mechanical allodynia. J Clin Invest 123:4050\u0026ndash;4062\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Y et al (2018) Touch and tactile neuropathic pain sensitivity are set by corticospinal projections. Nature 561:547\u0026ndash;550\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGier\u0026eacute; C, Melchior M, Dufour A, Poisbeau P (2021) Spinal integration of hot and cold nociceptive stimuli by wide-dynamic-range neurons in anesthetized adult rats. Pain Rep 6:e983\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeo JP, Jang SH (2013) Traumatic thalamic injury demonstrated by diffusion tensor tractography of the spinothalamic pathway. Brain Inj 27:749\u0026ndash;753\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJang SH, Lee J, Yeo SS (2017) Central post-stroke pain due to injury of the spinothalamic tract in patients with cerebral infarction: a diffusion tensor tractography imaging study. Neural Regen Res 12:2021\u0026ndash;2024\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAllam AK et al (2025) The human spinothalamic tract: lessons from cordotomy. Brain Commun 7:fcaf237\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNathan PW (1990) Touch and surgical division of the anterior quadrant of the spinal cord. J Neurol Neurosurg Psychiatry 53:935\u0026ndash;939\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen H et al (2024) The functional and anatomical characterization of three spinal output pathways of the anterolateral tract. Cell Rep 43\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Z, Zheng H, Yu Q, Jing X (2024) Understanding of Spinal Wide Dynamic Range Neurons and Their Modulation on Pathological Pain. J Pain Res 17:441\u0026ndash;457\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDauleac C, Boukhari A, Jacquesson T, Criton G, Cotton F, Frindel C (2025) Stitching brain and spinal cord DTI using cross-correlation registration: toward an atlas of spinal tracts. Eur Radiol Experimental 9:112\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTinnermann A, B\u0026uuml;chel C, Cohen-Adad J (2021) Cortico-spinal imaging to study pain. NeuroImage 224:117439\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWei Z et al (2025) CoSpine open access simultaneous cortico-spinal fMRI database of thermal pain and motor tasks. Sci Data 12:1696\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM (2012) FSL Neuroimage 62:782\u0026ndash;790\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndersson JLR, Sotiropoulos SN (2016) An integrated approach to correction for off-resonance effects and subject movement in diffusion MR imaging. NeuroImage 125:1063\u0026ndash;1078\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYam MF, Loh YC, Tan CS, Khadijah Adam S, Abdul Manan N, Basir R (2018) General Pathways of Pain Sensation and the Major Neurotransmitters Involved in Pain Regulation. Int J Mol Sci 19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAb Aziz CB, Ahmad AH (2006) The role of the thalamus in modulating pain. Malaysian J Med sciences: MJMS 13:11\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSciacca S, Lynch J, Davagnanam I, Barker R (2019) Midbrain, pons, and medulla: anatomy and syndromes. Radiographics 39:1110\u0026ndash;1125\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBasbaum AI, Bautista DM, Scherrer G, Julius D (2009) Cellular and Molecular Mechanisms of Pain. Cell 139:267\u0026ndash;284\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYarmolinsky DA et al (2025) Differential modification of ascending spinal outputs in acute and chronic pain states. Neuron 113:1223\u0026ndash;1239e1225\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLenz FA, Ohara S, Gracely RH, Dougherty PM, Patel SH (2004) Pain encoding in the human forebrain: binary and analog exteroceptive channels. J Neurosci 24:6540\u0026ndash;6544\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun YG (2025) Central neural circuits underlying itch sensation. Nat Rev Neurosci 26:765\u0026ndash;777\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarshall AG, Sharma ML, Marley K, Olausson H, McGlone FP (2019) Spinal signalling of C-fiber mediated pleasant touch in humans. Elife 8\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChisholm KI, Lo Re L, Polg\u0026aacute;r E, Gutierrez-Mecinas M, Todd AJ, McMahon SB (2021) Encoding of cutaneous stimuli by lamina I projection neurons. Pain 162:2405\u0026ndash;2417\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndrew D, Craig AD (2002) Quantitative responses of spinothalamic lamina I neurones to graded mechanical stimulation in the cat. J Physiol 545:913\u0026ndash;931\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavidson S et al (2012) Pruriceptive spinothalamic tract neurons: physiological properties and projection targets in the primate. J Neurophysiol 108:1711\u0026ndash;1723\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu Z, Liu T, Zeng C, Yang M, Xu X (2025) Local abnormal white matter microstructure in the spinothalamic tract in people with chronic neck and shoulder pain. Front NeuroSci 18\u0026ndash;2024\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobertson JW, Aristi G, Hashmi JA (2023) White matter microstructure predicts measures of clinical symptoms in chronic back pain patients. Neuroimage Clin 37:103309\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin X-M et al (2025) A corticospinal signature for interindividual pain sensitivity. Nat Commun\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang C et al (2025) The role of pain catastrophizing in modulating irritable bowel syndrome symptoms through brain-spinal cord integration. Meta-Radiology, 100187\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMugler IIIJP, Brookeman JR (1990) Three-dimensional magnetization-prepared rapid gradient-echo imaging (3D MP RAGE). Magn Reson Med 15:152\u0026ndash;157\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTustison NJ et al (2010) N4ITK: improved N3 bias correction. IEEE Trans Med Imaging 29:1310\u0026ndash;1320\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFisch L et al (2024) deepbet: Fast brain extraction of T1-weighted MRI using Convolutional Neural Networks. Comput Biol Med 179:108845\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Leener B et al (2017) Spinal Cord Toolbox, an open-source software for processing spinal cord MRI data. NeuroImage 145:24\u0026ndash;43\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Leener B, Fonov VS, Collins DL, Callot V, Stikov N, Cohen-Adad J (2018) PAM50: Unbiased multimodal template of the brainstem and spinal cord aligned with the ICBM152 space. NeuroImage 165:170\u0026ndash;179\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJenkinson M, Beckmann CF, Behrens TEJ, Woolrich MW, Smith SM (2012) FSL NeuroImage 62:782\u0026ndash;790\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVeraart J, Novikov DS, Christiaens D, Ades-Aron B, Sijbers J, Fieremans E (2016) Denoising of diffusion MRI using random matrix theory. NeuroImage 142:394\u0026ndash;406\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCordero-Grande L, Christiaens D, Hutter J, Price AN, Hajnal JV (2019) Complex diffusion-weighted image estimation via matrix recovery under general noise models. \u003cem\u003eNeuroImage\u003c/em\u003e 200, 391\u0026ndash;404\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVeraart J, Fieremans E, Novikov DS (2016) Diffusion MRI noise mapping using random matrix theory. Magn Reson Med 76:1582\u0026ndash;1593\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTournier JD et al (2019) MRtrix3: A fast, flexible and open software framework for medical image processing and visualisation. NeuroImage 202:116137\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKellner E, Dhital B, Kiselev VG, Reisert M (2016) Gibbs-ringing artifact removal based on local subvoxel-shifts. Magn Reson Med 76:1574\u0026ndash;1581\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndersson JL, Skare S, Ashburner J (2003) How to correct susceptibility distortions in spin-echo echo-planar images: application to diffusion tensor imaging. NeuroImage 20:870\u0026ndash;888\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith SM et al (2004) Advances in functional and structural MR image analysis and implementation as FSL. NeuroImage 23(Suppl 1):S208\u0026ndash;219\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreve DN, Fischl B (2009) Accurate and robust brain image alignment using boundary-based registration. NeuroImage 48:63\u0026ndash;72\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eB\u0026eacute;dard S et al (2025) Towards contrast-agnostic soft segmentation of the spinal cord. Med Image Anal 101:103473\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWelvaert M, Rosseel Y (2013) On the definition of signal-to-noise ratio and contrast-to-noise ratio for FMRI data. PLoS ONE 8:e77089\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRolls ET, Huang C-C, Lin C-P, Feng J, Joliot M (2020) Automated anatomical labelling atlas 3. \u003cem\u003eNeuroImage\u003c/em\u003e 206, 116189\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"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":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8640197/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8640197/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe spinothalamic tract (STT) is the principal ascending pathway for nociceptive transmission and a key integrative hub for pain. However, whether inter‑individual microstructural variability constitutes a shared neuroanatomical basis for cross‑modal pain sensitivity remains unclear, partly due to limitations in tract‑level imaging along the brain-spinal axis. We applied a combined brain-spinal cord diffusion tensor imaging framework to characterize STT microstructure across the central nervous system. Multivariate analyses identified a distributed STT microstructural pattern robustly associated with individual sensitivity to heat and mechanical pain across ten experimental measures. Although STT microstructure was also related to tactile sensitivity, pain and tactile measures were linked to largely distinct multivariate patterns, indicating a modality-differentiated organization within the tract. Importantly, an STT-derived microstructural pattern predicted individual differences in experimental pain sensitivity and generalized to clinical pain severity in cohorts with zoster and irritable bowel syndrome. Together, these findings provide that variability in STT microstructure shapes individual pain vulnerability and establishes a structural framework linking ascending nociceptive pathways to experimental and clinical pain in humans.\u003c/p\u003e","manuscriptTitle":"Spinothalamic Tract Microstructure as a Common Neural Substrate of Pain Sensitivity Across Modalities: A Combined Brain-Spinal Cord Diffusion Imaging Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-27 06:51:12","doi":"10.21203/rs.3.rs-8640197/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"3da1946a-c146-430c-83a6-aafdd5c5dd93","owner":[],"postedDate":"January 27th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":61649533,"name":"Biological sciences/Neuroscience/Neural circuits"},{"id":61649534,"name":"Health sciences/Neurology"}],"tags":[],"updatedAt":"2026-04-28T16:45:40+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-27 06:51:12","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8640197","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8640197","identity":"rs-8640197","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}