Assessment of Diagnostic Accuracy and Measurement Reliability of Low-keV Virtual Monoenergetic Dual-energy CT in The Liver Metastases of Colorectal Cancer: A Prospective, Imaging-Reference Study

Assessment of Diagnostic Accuracy and Measurement Reliability of Low-keV Virtual Monoenergetic Dual-energy CT in The Liver Metastases of Colorectal Cancer: A Prospective, Imaging-Reference 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 Research Article Assessment of Diagnostic Accuracy and Measurement Reliability of Low-keV Virtual Monoenergetic Dual-energy CT in The Liver Metastases of Colorectal Cancer: A Prospective, Imaging-Reference Study Zhaoyang Zheng, Jianyang Yang, Hong Zhu, Xi Feng, Jiayu Sun, Lie Yang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8359303/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Objective: To assess the value of low-keV virtual monoenergetic imaging (VMI) from dual-energy CT (DECT) for detecting liver metastases of colorectal cancer (CRC) and for evaluating treatment response. Materials and Methods: Patients diagnosed with CRC who had liver metastases (CRLM) were prospectively enrolled. All underwent DECT, followed by magnetic resonance imaging (MRI) used as the gold standard. DECT datasets were reconstructed as standard linearly-blended (M_0.6) images to simulate contrast-enhanced CT (CECT) and as VMI at 10-keV intervals (40–70 keV). Signal-to-noise (SNR) and contrast-to-noise (CNR) ratios were measured. Radiologists independently assessed image quality, lesion delineation, and image noise using a 5-point Likert scale. Per-lesion sensitivity, specificity, and detection rates were calculated.Treatment response was evaluated using RECIST 1.1. Results: Thirty-five patients (125 liver metastases) were enrolled. 40-keV VMI provided the best CNR (6.2 ± 4.3, p < 0.01) and SNR (15.5 ± 9.5, p < 0.01), with AUCs of 0.974 for SNR and 0.966 for attenuation. It outperformed M_0.6 images for detecting lesions ≤10 mm (86.7% vs. 60.0%, p 0.05). Lesion delineation was optimal with 40-keV and 50-keV VMI (median 5). Image noise was lowest with 60-keV VMI (median 4, p < 0.01). Lesion size reduction was comparable between MRI and DECT (5.2 ± 5.5 mm vs. 4.6 ± 6.2 mm, p = 0.41). Conclusion: Low-keV VMI improves the diagnostic accuracy of liver metastases of CRC compared to CECT, while maintaining measurement reliability in treatment-response assessment. Colorectal liver metastases Dual-energy CT Virtual monoenergetic imaging Diagnostic accuracy Measurement reliability Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Colorectal cancer (CRC) is the third most common cancer and the second leading cause of cancer-related death worldwide [ 1 ]. The liver remains the predominant site of hematogenous metastasis, with 15%–25% of patients presenting synchronous metastases at initial diagnosis and 15%–25% developing metachronous metastases after resection of the primary colorectal tumor [ 2 , 3 ]. Liver metastases has been the most common cause of CRC-related death, and patients with unresectable metastatic lesions have a five-year survival rate of less than 5% [ 2 , 4 ]. The length of survival is significantly influenced by the extent of surgical removal of the hepatic metastases [ 5 , 6 ]. Achieving complete resection or no-evidence-of-disease (NED) status for liver metastases of CRC has been shown to increase five-year survival to 30%–57% [ 2 ]; however, this benefit depends largely on precise pre-operative tumor imaging. There have been significant advancements in the imaging diagnosis of colorectal cancer with liver metastases (CRLM) in recent years. Although contrast-enhanced computed tomography (CECT) remains the most commonly used modality for screening and staging [ 7 ], its ability to detect sub-centimeter lesions is limited. The sensitivity of CECT for detecting liver metastases smaller than 10 mm ranges from 22% to 68%, which directly affects the assessment of surgical resectability and patient prognosis [ 8 – 10 ]. Nevertheless, CT remains crucial for initial diagnosis. Its capacity to detect small lesions during first-line examination usually prompts referral to magnetic resonance imaging (MRI) for definitive assessment. This helps prevent missed diagnosis, particularly for sub-centimeter metastases. Gadoxetic acid disodium-enhanced MRI has demonstrated superior performance in the diagnosis of liver metastases of CRC, achieving a sensitivity from 90.0% to 96.0% and a specificity of approximately 90.0%, particularly in identifying sub-centimeter lesions [ 11 – 14 ]. However, MRI has inherent limitations, including higher cost, longer examination times, and greater patient cooperation. Moreover, its use is restricted in patients with metallic implants or severe renal dysfunction. Consequently, more efficient and widely accessible alternatives need to be explored. Dual-energy CT (DECT) has emerged as an innovative and promising imaging modality. This technology enables the reconstruction of iodine concentration maps, virtual non-contrast images, and Z-effective maps, which are invaluable for quantifying iodine uptake and for differentiating hypovascular metastases from normal tissue, particularly when lesions are small or poorly attenuated [ 15 , 16 ]. Virtual monoenergetic imaging (VMI), a crucial post-processing technique of DECT, improves the visibility of iodine uptake. In particular, VMI at low kiloelectron volts (keV) levels increases contrast-to-noise ratio (CNR) between lesions and liver tissue without the need for additional radiation or contrast agent. This reconstruction approach decreases potential health risks and promotes the detectability of both hypo- and hyper-vascular hepatic lesions [ 17 – 19 ]. Prior single-center retrospective studies have shown that 40-keV and 60-keV VMI improves image quality and the detection of liver metastases of CRC [ 17 , 20 ]; however, prospective reference data are lacking, particularly for small metastases (< 10 mm). We therefore conducted a prospective imaging-reference study to assess the value of low-keV (40–70 keV) virtual monoenergetic DECT. The objectives were 1) to improve both quantitative and qualitative image quality in comparison with CECT, 2) to increase diagnostic accuracy in the detection of liver metastases of CRC, especially sub-centimeter ones, and 3) to maintain measurement reliability in treatment-response evaluation of liver lesions according to RECIST 1.1 criteria[ 21 ]. Materials and Methods Patient Selection Between January 2022 and December 2023, consecutive patients were enrolled if they had histologically confirmed CRC on colonoscopy and were highly suspected of having CRLM on abdominal DECT. MRI was performed within one week after DECT examination. Inclusion criteria: 1) Age ≥ 18 years; 2) No history of radical hepatic resection for malignancy. Exclusion criteria: 1) Severe image artefacts; 2) Contraindications to MRI or contrast agents; 3) Protocol deviations during DECT or contrast administration; 4) Confirmed or suspected pregnancy; 5) Estimated glomerular filtration rate < 45 mL/min. Sample-size Calculation On the basis of previous per-patient analyses that reported 91% sensitivity and 94–97% specificity for DECT in the detection of liver metastases [ 22 ], we conservatively estimated both parameters [p and (1-p)] as 90% in the present study. This estimation used a self-paired design (metastatic versus normal liver tissue) with a margin of error ( \(\:{\delta\:}\) ) of 0.10 and an alpha (α) of 0.05. The sample-size formula was as follows: $$\:\text{n}=\left(\frac{{\text{z}}_{{\alpha\:}}}{{\delta\:}}\right)²\left(1-\text{p}\right)\text{p}$$ The minimum required sample size was 35 patients. DECT Protocol Patients fasted for 4–6 hours before CT and ingested 200 ml of water prior to the scan. Non-enhanced and double-phase-enhanced scans were performed with a DECT scanner (Spectral CT 7500, Philips Healthcare, Netherlands). Patients were positioned supine and scanned from the top of the diaphragm to the inferior border of the pubic symphysis. A total of 1.2 mL/kg body weight of contrast medium was injected into the middle elbow vein using a double-barrelled high-pressure syringe at a flow rate of 3 mL/s. High-dose tracking was initiated with a threshold set at 150 Hounsfield units in the abdominal aorta. The arterial phase scan was automatically triggered upon reaching peak contrast concentration, and the portal phase scan was initiated 70 seconds post-contrast injection. The scanning parameters were as follows: tube voltage, 120 kVp; tube current, modulated by automatic exposure control; pitch, 1.000; rotation time, 0.5 seconds; collimation, 128 × 0.625 mm; field of view, 350 mm × 350 mm; matrix, 512 × 512. DECT Image Reconstruction Standard linearly-blended (M_0.6) images were automatically generated with a blending factor of 0.6 to simulate CECT (conventional 120-kV single-energy) images [ 23 ]. Spectral based images were reconstructed with the spectral level-4 algorithm. These spectral datasets were transferred to a dedicated spectral post-processing workstation (IntelliSpace Portal, Version 12.0; Philips Healthcare, Amsterdam, The Netherlands). Monoenergetic reconstructions were generated with the VMI algorithm. We solely reconstructed four VMI image series (40, 50, 60, 70 keV) because lesion conspicuity and CNR would be considered too low beyond 70 keV, as reported previously [ 17 , 24 – 26 ]. Quantitative Analysis of DECT Datasets Quantitative analysis was conducted by two radiologists with 3 years of experience in oncological abdominal DECT imaging. Signal attenuation (mean Hounsfield units, HU) and standard deviation (SD) were measured by placing circular regions of interest (ROI) within the liver metastases and adjacent, unaffected liver parenchyma (100 mm²). Each ROI was placed on the largest regions of the lesion with the most significant enhancement in the portal venous phase, avoiding areas of focal heterogeneity and tumor necrosis. The size of the tumor ROI was adjusted to the tumor’s overall dimensions for precise depiction. Additionally, homogeneous ROIs were placed within the psoas muscle to evaluate image contrast (100 mm 2 ). All measurements were performed three times in three consecutive slices, and average values were calculated. The signal-to-noise ratio (SNR) and CNR calculations were as follows: SNR = HU ROI /SD ROI CNR=|HU ROI −HU muscle |/SD muscle Qualitative Analysis of DECT Datasets Qualitative analysis was performed by two radiologists with 15 years of abdominal CT interpretation experience. The readers were blinded to the reconstruction type being assessed. Standard linearly-blended (M_0.6) images and VMI reconstructions were rated in random order, with only one single image series rated during each reading session. Readers were permitted to adjust standard window settings manually, as prior studies suggest different window configurations for the VMI algorithm [ 27 ]. Overall image quality (ranging from 1 = poor overall image quality to 5 = excellent overall image quality), lesion delineation (ranging from 1 = lesion cannot be ruled out to 5 = excellent lesion margin delineation), and image noise (ranging from 1 = enormous image noise to 5 = no relevant image noise perceivable) were scored on 5-point Likert scales. Reference Standard Gadoxetic acid disodium-enhanced MRI combined with ≥ 6 months of imaging follow-up was used as the reference standard for confirming CRLM. Although histopathology is the ideal gold standard, it was not universally available due to the prospective nature of this study. MRI was considered the first-line modality and the “gold standard” for pre-operative imaging because of its higher sensitivity and specificity for CRLM detection compared with CECT and fluoro-2-deoxyglucose positron emission tomography (FDG-PET) [ 8 , 28 – 30 ]. All indeterminate lesions were independently reviewed by two senior abdominal radiologists and confirmed through follow-up imaging. Treatment-response Assessment The treatment modality was determined at initial evaluation. For resectable or potentially resectable metastases, surgical resection or neoadjuvant chemotherapy followed by surgery was recommended. Patients were reassessed with DECT and MRI after 4–6 cycles of systemic chemotherapy. Target lesions were selected according to RECIST 1.1 [ 21 ]. Four response categories were recorded: complete response, partial response, progressive disease, and stable disease. The sum of the longest diameters (SLD) was compared between baseline and follow-up. Statistical Analysis Statistical analyses were conducted using SPSS 20.0 (IBM, Corp., Armonk, NY, USA) and MedCalc (version 20.218). Normally distributed data are expressed as mean ± SD, whereas non-normally distributed data are expressed as median and interquartile range (IQR). The Kolmogorov-Smirnov test was used to assess Gaussian distribution. Analysis of variance (ANOVA) was applied for multiple comparisons of normally distributed variables, while the Wilcoxon matched pairs test was used for non-Gaussian data distribution. Inter-reader agreement was assessed with the intraclass correlation coefficient (ICC) using a two-way mixed model and 95% confidence intervals (CIs). ICC values < 0.40 were considered poor, values ranging from 0.40–0.59 as fair, 0.60–0.74 as good, and 0.75–1.0 as excellent agreement [ 31 ]. Receiver operating characteristic (ROC) curves were constructed, and the area under the curve (AUC) was calculated. Optimal sensitivity and specificity were determined with the maximum Youden index. The four VMI series (40–70 keV) were entered into a multivariable logistic regression model to assess the diagnostic value of the combined approach. The Chi-square test was used to compare the detection rates of liver metastases of CRC between groups. A p -value๤0.05 was considered statistically significant. Results Patient Characteristics Thirty-five patients (22 male, 13 female; mean age 61.0 ± 12.2 years) were enrolled. A total of 125 liver metastases of CRC were confirmed: ≥ 5 lesions in 11 patients, < 5 lesions in 11 patients, and a single lesion in 13 patients. Mean lesion size was 27.4 ± 17.9 mm. Tumor characteristics and patient demographics are summarized in Table 1 . Table 1 Patient Characteristics Parameter Gender Male 22 (60.5%) Female 13 (61.9%) Age (years) 61.0 ± 12.2 No. of liver metastases ≥ 5 11 (31.4%) ๤5 11 (31.4%) Singular 13 (37.1%) Total number 125 Mean tumor size (mm) 27.4 ± 17.9 Radiation dose CTDIvol (mGy) 41.6 ± 12.6 DLP (mGy·cm) 2108.1 ± 463.5 CTDIvol, volume CT dose index; DLP, dose-length product. Quantitative Image Quality Quantitative analysis of the DECT datasets revealed that both SNR (average) and CNR (liver metastases) peaked with 40-keV VMI (15.5 ± 9.5 and 6.2 ± 4.3, respectively) and were significantly higher than those with M_0.6 (7.0 ± 3.3 and 1.5 ± 1.2; p < 0.01) and all other VMI series. Table 2 details mean attenuation, SNR, and CNR values in patients with CRLM. Figure 1 shows box-and-whisker plots of SNR and CNR distributions. Table 2 Comparison of mean attenuation, SNR and CNR of patients with CRLM. Quantitative image analysis included standard linearly-blended (M_0.6) image series and VMI series at energy levels ranging from 40 keV to 70 keV. M_0.6 VMI 40 VMI 50 VMI 60 VMI 70 Attenuation Liver parenchyma 118.1 ± 17.2 275.8 ± 44.1 195.1 ± 30.0 147.7 ± 22.7 119.5 ± 18.5 Liver metastases 70.75 ± 20.4 157.0 ± 51.3 * 112.6 ± 36.1 85.5 ± 27.6 69.4 ± 22.9 p value 0.000 0.000 0.000 0.000 0.000 SNR Liver parenchyma 9.2 ± 2.8 22.9 ± 6.8 17.2 ± 5.1 13.4 ± 4.0 11.1 ± 3.4 Liver metastases 4.9 ± 2.2 7.9 ± 4.0 * 7.0 ± 3.3 6.1 ± 2.8 5.4 ± 2.5 p value 0.000 0.000 0.000 0.000 0.000 Average 7.0 ± 3.3 15.5 ± 9.5 * 12.2 ± 7.0 9.9 ± 5.9 8.5 ± 5.6 CNR Liver metastases 1.5 ± 1.2 6.2 ± 4.3 * 4.0 ± 3.0 2.5 ± 2.0 1.8 ± 1.4 SNR, signal-to-noise ratio; CNR, contrast-to-noise ratio; CRLM, colorectal cancer with liver metastases. * indicates significance at p ๤0.05 (VMI 40 vs. other image series) CRC, colorectal cancer. ROC analysis revealed no significant differences ( p > 0.05) in AUCs among M_0.6 and 40-, 50-, 60-, and 70-keV VMI images for mean attenuation in diagnosing CRLM (AUCs: 0.955, 0.966, 0.959, 0.962, and 0.958, respectively). The combined 40–70 keV VMI approach yielded an AUC of 0.969 (Fig. 2 ). AUCs for SNR were 0.893 (M_0.6), 0.974 (40 keV), 0.958 (50 keV), 0.939 (60 keV), and 0.914 (70 keV), with statistically significant differences ( p < 0.01). M_0.6 images showed the lowest AUC, whereas 40-keV VMI achieved the highest. The combined 40–70 keV VMI approach achieved an AUC of 0.987 (Fig. 3 ). Supplementary Table S1 lists the thresholds, specificity, and sensitivity derived from the ROC curves. A representative case is shown in Fig. 4 . Qualitative Image Quality Table 3 summarizes the qualitative image-analysis parameters. Overall image quality was comparable between 60-keV VMI and M_0.6 images (median 5 vs 5, p ๥0.05), both reached the highest scores, with excellent inter-reader agreement (60-keV VMI: ICC, 0.82; 95% CI, 0.75–0.87; M_0.6: ICC, 0.84; 95% CI, 0.78–0.89). Table 3 Qualitative image quality parameters are given as medians and ranges in parentheses. Standard linearly-blended (M_0.6) images and VMI series at energy levels ranging from 40 keV to 70 keV were subjectively evaluated in terms of overall image quality, lesion delineation, and image noise by two blinded radiologists. M_0.6 VMI 40 VMI 50 VMI 60 VMI 70 Overall image quality 5 (3–5) 4 (3–4) 4 (3–5) 5 (3–5) 4 (3–5) Lesion delineation 3 (1–5) 5 (1–5) 5 (1–5) 4 (1–5) 4 (1–5) Image noise 4 (2–5) 3 (2–5) 4 (2–4) 4 (3–5) 3 (2–4) Lesion delineation scores were highest with 40-keV and 50-keV VMI (median 5 vs 5; p ๥0.05) and were significantly higher than those for M_0.6 images (median 3, range 1–5; p ๤0.01), with excellent inter-reader agreement (40-keV VMI: ICC, 0.96; 95% CI, 0.94–0.97; 50-keV VMI: ICC, 0.92; 95% CI, 0.89–0.94). The lowest image-noise scores (i.e., greatest noise) were observed with 70-keV VMI (median 3, range 2–4; ICC 0.71; 95% CI 0.59–0.79). By contrast, 60-keV VMI achieved the highest image-noise scores (median 4, range 3–5; p ๤0.01) with excellent inter-reader agreement (ICC, 0.98; 95% CI, 0.98–0.99). Overall inter-reader agreement for image quality, lesion delineation and image noise was excellent (ICC, 0.86; 95% CI, 0.84–0.87). Lesion Detection A total of 125 liver metastases of CRC were detected and included in the analysis. No significant differences were observed for lesions > 10 mm between M_0.6 and any VMI series ( p ๥0.05). For lesions ≤ 10 mm (n = 15), per-lesion sensitivity was 60.0% with M_0.6, 86.7% with 40-keV VMI, 80.0% with 50-keV VMI, 80.0% with 60-keV VMI, and 73.3% with 70-keV VMI. 40-keV VMI was significantly superior to M_0.6 images ( p ๤0.05). Detailed data are provided in Table 4 . Lesion conspicuity across image series is compared in Fig. 5 . Table 4 Comparison of M_0.6 images, VMI series at energy levels ranging from 40 keV to 70 keV in the detection rates of liver metastases of CRC. ≤ 10mm (n = 15) 11–19mm (n = 33) ≥ 20mm (n = 77) VMI 40 13 (86.7%) a 30 (90.9%) 77 (100.0%) VMI 50 12 (80.0%) a 31 (93.9%) 77 (100.0%) VMI 60 12 (80.0%) a 31 (93.9%) 77 (100.0%) VMI 70 11 (73.3%) a,b 29 (87.9%) 77 (100.0%) M_0.6 9 (60.0%) b 27 (81.8%) 74 (96.1%) p value ๤0.05 ๥0.05 ๥0.05 CRC, colorectal cancer. The same superscript letters indicate no significant difference ( p ๥0.05), and different superscript letters indicate significant differences ( p ๤0.05). CRC, colorectal cancer. Treatment-Response Assessment Table 5 lists the treatments administered to all 35 patients. Three patients underwent radical surgery without receiving chemotherapy. Thirty-two patients received systemic chemotherapy with or without targeted therapy; 17 of them subsequently underwent radical surgery, and 1 underwent transcatheter arterial chemoembolization. Table 5 Treatment and tumor response assessment in patients with CRLM. Thirty-five CRLM patients Data Initial resectability assessment Resectable 3 (8.6%) Unresectable 15 (42.8%) Potential-resectable 17 (48.6%) Distribution Solitary 9 (25.7%) Multiple in a single lobe 8 (22.9%) Multiple in both lobes 18 (51.4%) Treatment process Radical surgery 3 (8.6%) Chemotherapy → radical surgery 17 (48.6%) Chemotherapy ± targeted therapy 14 (40.0%) Chemotherapy → TACE 1 (2.9%) SLD before treatment (mm) MRI 32.2 ± 22.8 DECT 33.5 ± 23.3 p value >0.05 SLD after treatment (mm) MRI 27.0 ± 22.4 DECT 28.8 ± 22.5 p value >0.05 SLD difference before and after treatment (mm) MRI 5.2 ± 5.5 DECT 4.6 ± 6.2 p value 0.41 Tumor response assessment Stable disease MRI 25 (71.4%) DECT 27 (77.1%) Partial response MRI 7 (20.0%) DECT No detection of metastases DECT 5 (14.3%) 1 (2.9%) CRLM, colorectal cancer with liver metastases; TACE, transcatheter arterial chemoembolization; SLD, sum of longest diameters. Mean SLD reduction was 5.2 ± 5.5 mm with MRI and 4.6 ± 6.2 mm with 40-keV VMI ( p = 0.41). According to tumor response assessment, MRI classified 7 patients as having partial response and 25 as stable disease, whereas DECT (40-keV VMI) classified 5 as partial response and 27 as stable disease; lesions were undetected in one patient. Response classification differed between DECT (40-keV VMI) and MRI in 3 of the 35 patients (8.57%), all with a measurement difference of less than 5 mm. Discussion CRLM remains a key determinant of survival; however CECT may miss up to 40% of sub-centimeter liver metastases. As a result, curable tumors may be neglected at a stage when radical resection is still possible. We demonstrates that low-keV VMI (40–70 keV) reconstructed from DECT significantly increases conspicuity between metastatic lesions and background liver parenchyma. This improvement leads to a maximum 20% increase in per-lesion sensitivity for detecting liver metastases, without loss of specificity. Moreover, the detection of sub-centimeter metastases rises by 13–26%, without additional radiation or contrast medium. Importantly, the technique also allows for accurate measurement of chemotherapeutic response. By transforming a routine scan into a high-definition “digital biopsy”, low-keV VMI provides superior surgical guidance. This enables more patients to become eligible for complete (R0) hepatectomy and reduces inter-observer variability during post-chemotherapy monitoring. Thus, we consider the reconstruction approach to become a practical and cost-effective strategy for improving oncological outcomes. The present study further explores the value of low-keV VMI in optimizing image quality for liver metastases of CRC. Lesion visibility on CT images is primarily affected by image noise, tissue contrast, and CNR [ 32 ]. DECT enhances diagnostic confidence for the detection of sub-centimeter liver metastases by employing multi-energy reconstruction to improve CNR and suppress image noise [ 22 , 33 , 34 ]. Although lower-keV VMI affords higher tissue contrast compared to higher energy levels, it also amplifies image noise, which may obscure the margins of small or hypovascular lesions. Conversely, as keV increases, image noise diminishes but contrast is sacrificed [ 35 – 37 ]. Therefore, identification of an optimal monoenergetic level that balances these competing demands is critical for maximizing lesion conspicuity. This balance requires a reconstruction strategy that integrates tumor biology (e.g., perfusion pattern, degree of necrosis) and patient characteristics (e.g., body shapes, radiation-dose constraints). Prior studies have suggested that 50–60 keV represents the ideal range for abdominal oncological imaging, offering the best compromise between contrast preservation and noise suppression [ 38 , 39 ]. Emerging evidence, including the study by Lenga et al. [ 17 ], indicates that 40-keV VMI may offer superior diagnostic value for oncologic status and tumor response assessment. Our findings extend this paradigm by demonstrating that 40-keV VMI yields peak CNR and SNR, whereas 60-keV VMI delivers the best overall image quality. The most accurate lesion delineation was occurred at 40 keV and 50 keV, which is crucial for precise tumor-volume calculation and treatment planning. Furthermore, the combined 40–70 keV approach achieved significantly better diagnostic accuracy than M_0.6 images. By optimizing sensitivity while preserving measurement reliability, this strategy establishes a reliable and reproducible framework for CRLM evaluation. Sub-centimeter liver metastases are of great importance in clinical diagnosis. Compared to larger lesions, smaller ones are more likely to be missed during pre-operative imaging and intra-operative inspection. However, once successfully detected, they are usually more suitable for radical surgical resection or local ablation [ 10 , 40 , 41 ]. Missed detection might lead to serious implications. Ko et al. [ 10 ] reported that false-negative pre-operative CT nodules were associated with a signigicantly higher 1-year recurrence rate after liver resection in patients with CRLM ( p = 0.048), indicating that undetected lesions may ultimately contribute to recurrence. The investigators thus concluded that improving the sensitivity could improve prognosis. Nevertheless, the sensitivity of CECT for the detection of sub-centimeter liver metastases is variable and limited, ranging from 22% to 68% [ 8 ]. This variability primarily results from methodological differences across studies and insufficient statistical power in small lesion cohorts. Accordingly, the average sensitivity of CECT for metastases smaller than 10 mm remains modest. Our findings demonstrate that 40-keV VMI achieves the highest per-lesion detection rate for sub-centimeter liver metastases of CRC, whereas M_0.6 images yield the lowest (86.7% versus 60.0%). This 26% incremental gain highlights the clinical value of low-keV VMI in tumor staging, as it identifies metastatic foci that would otherwise escape CECT detection. By facilitating more precise surgical planning toward R0 resection, the reconstruction approach increases the likelihood of achieving NED status. However, the association between improved detection and long-term outcomes remains to be proven. While low-keV VMI improves diagnostic accuracy in detecting sub-centimeter metastases, prospective studies are required to verify its impact on surgical outcomes and recurrence rates. Accurate pre-operative imaging and staging are critical for the management of CRLM, as they determine whether treatment should consist of surgical resection, loco-regional therapy, or systemic chemotherapy [ 42 , 43 ]. Low-keV VMI enhances this process by amplifying SNR without additional contrast or radiation. This technique transforms CECT images into high-definition “digital biopsies” that enable precise staging and visualization of the sub-millimeter margin between tumor and portal radicle, a key factor for R0 resection [ 44 ]. Furthermore, the same reconstruction algorithm provides RECIST-based [ 45 ] tumor measurements with minimal intra-reader variability, thereby reducing inter-observer error and the risk of incorrectly assessing disease progression. Our findings demonstrate that integrating low-keV VMI reconstructions into clinical protocol for post-chemotherapy and surgical monitoring transforms patient management from subjective experience toward objective, reproducible parameters. This helps avoid futile regimens and may ultimately improve survival. Several limitations should be acknowledged. First, the generalizability of our findings is limited by the single-center design, relatively small sample size, and absence of external validation. Second, despite a standardized ROI protocol, sampling variance was inevitable because of the variable shapes, necrotic cores, and irregular margins of liver metastases of CRC. Third, evaluation of low-keV VMI for sub-centimeter metastases was constrained by the small sample (n = 15), which may diminish statistical power. Finally, results obtained on a single platform may not be fully generalizable to other DECT devices in different institutions. In conclusion, our study demonstrates that low-keV VMI reconstructions significantly improve diagnostic precision for detecting liver metastases of CRC while preserving measurement reliability for treatment-response evaluation. Notably, this technology delivers superior image quality and higher sensitivity for sub-centimeter metastases than CECT, without additional radiation or contrast. By minimizing missed diagnosis, low-keV VMI provides a reliable tool for initial screening. Integrating this reconstruction strategy into routine abdominal imaging protocols can improve staging accuracy, guide personalized surgical planning, and facilitate timely detection of recurrence, thereby offering a practical and cost-effective approach of improving patient outcomes. Abbreviations AUC, area under the curve (a statistical measure used to evaluate the performance of a diagnostic test, higher value indicates better diagnostic performance [value ranges from 0 to 1]) CRC, colorectal cancer (a malignant tumor arising from the epithelial lining of the colon or rectum) CRLM, colorectal cancer with liver metastases (advanced-stage colorectal cancer characterized by the spread of malignant cells from the primary colorectal tumor to the liver) CECT, contrast-enhanced computed tomography (a standard imaging technique that utilizes intravenous iodinated contrast material to improve visualization of vascular structures and enhance the detection and characterization of pathological conditions) CNR, contrast-to-noise ratio (a quantitative metric in medical imaging that compares the difference in signal intensity between a region of interest and adjacent background tissue to the image noise) DECT, dual-energy CT (an advanced computed tomography technique that acquires data at two distinct X-ray energy spectra) HU, hounsfield units (a standardized quantitative scale used in computed tomography to express the relative attenuation of X-rays by tissue) ICC, intraclass correlation coefficient (a statistical measure of reliability that quantifies the degree of agreement among multiple observers or repeated measurements) keV, kiloelectron volts (refers to the energy levels in virtual monoenergetic imaging) MRI, magnetic resonance imaging (an imaging modality that utilizes strong magnetic fields and radiofrequency pulses to generate detailed anatomical and functional images without ionizing radiation) NED, no-evidence-of-disease (a clinical status indicating complete remission where comprehensive imaging and clinical evaluation reveal no detectable residual malignancy following treatment) ROC, receiver operating characteristic (an analytical curve plotting the relationship between sensitivity and specificity across all possible thresholds of a diagnostic test) ROI, regions of interest (user-defined areas placed on medical images for quantitative analysis of tissue characteristics) SNR, signal-to-noise (a fundamental image quality metric calculated as the ratio of signal intensity from a tissue region to the background noise) SLD, sum of longest diameters (the arithmetic sum of the longest diameters of all target lesions identified at baseline) SD, standard deviation (a measure of statistical dispersion representing the variation or spread of a set of values from their mean, commonly used to quantify image noise in radiological measurements) VMI, virtual monoenergetic imaging (a post-processing technique derived from dual-energy CT data that reconstructs images as if they were acquired at a single, selectable X-ray energy level) Declarations Conflict of interest The authors declare no competing interests. Ethical approval and Informed consent This prospective study was conducted in accordance with the ethical standards and national research committee and with the principles of the Helsinki Declaration and its later amendments or comparable ethical standards. Approval was granted by the Biomedical Ethics Review Committee of West China Hospital of Sichuan University (2022-1904). And the study was registered with the Chinese Clinical Trial Registry (ChiCTR2500100120). Data availability All data generated or analyzed during this study are available on request from the corresponding author. Funding This work was funded by the National Key Research and Development Program of China—Industrial Software Key Special Project [grant number 2024YFB3311700]. Acknowledgements The authors would like to express their sincere gratitude. We are deeply grateful to the Department of Radiology, West China Hospital of Sichuan University for their professional technical support and assistance in image analysis throughout this research. We also extend our thanks to all our colleagues in the Division of Gastrointestinal Surgery, Department of General Surgery, West China Hospital of Sichuan University for their valuable suggestions and insightful discussions during the course of this project. Finally, we acknowledge the anonymous reviewers and editors for their diligent work and constructive comments. Author Contributions Conceptualization: Lie Yang, Jiayu Sun; Methodology: Zhaoyang Zheng, Jianyang Yang; Formal analysis and investigation: Zhaoyang Zheng, Jianyang Yang; Writing - original draft preparation: Zhaoyang Zheng; Writing - review and editing: Lie Yang, Hong Zhu, Xi Feng; Supervision: Lie Yang. References Bray F, Laversanne M, Sung H, et al (2024) Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 74(3):229-263 Chinese College of Surgeons, Chinese Society of Gastrointestinal Surgery, Chinese Society of Colorectal Surgery et al (2025) Chinese Guidelines for Diagnosis and Comprehensive Management of Liver Metastases from Colorectal Cancer (2025). Chinese Journal of Practical Surgery 45(8):841-854 Reboux N, Jooste V, Goungounga J, Robaszkiewicz M, Nousbaum JB, Bouvier AM (2022) Incidence and Survival in Synchronous and Metachronous Liver Metastases From Colorectal Cancer. JAMA Netw Open 5(10):e2236666 Stewart CL, Warner S, Ito K, et al (2018) Cytoreduction for colorectal metastases: liver, lung, peritoneum, lymph nodes, bone, brain. When does it palliate, prolong survival, and potentially cure?. Curr Probl Surg 55(9):330-379 Akgül Ö, Çetinkaya E, Ersöz Ş, Tez M (2014) Role of surgery in colorectal cancer liver metastases. World J Gastroenterol 20(20):6113-6122 Margonis GA, Buettner S, Andreatos N, et al (2019) Prognostic Factors Change Over Time After Hepatectomy for Colorectal Liver Metastases: A Multi-institutional, International Analysis of 1099 Patients. Ann Surg 269(6):1129-1137 Morris VK, Kennedy EB, Baxter NN, et al (2023) Treatment of Metastatic Colorectal Cancer: ASCO Guideline. J Clin Oncol 41(3):678-700 Niekel MC, Bipat S, Stoker J (2010) Diagnostic imaging of colorectal liver metastases with CT, MR imaging, FDG PET, and/or FDG PET/CT: a meta-analysis of prospective studies including patients who have not previously undergone treatment. Radiology 257(3):674-684 Norén A, Eriksson HG, Olsson LI (2016) Selection for surgery and survival of synchronous colorectal liver metastases; a nationwide study. Eur J Cancer 53:105-114 Ko Y, Kim J, Park JK, et al (2017) Limited detection of small (≤ 10 mm) colorectal liver metastasis at preoperative CT in patients undergoing liver resection. PLoS One 12(12):e0189797 Haimerl M, Wächtler M, Platzek I, et al. (2013) Added value of Gd-EOB-DTPA-enhanced Hepatobiliary phase MR imaging in evaluation of focal solid hepatic lesions. BMC Med Imaging 13:41 Granata V, Fusco R, Maio F, et al. (2019) Qualitative assessment of EOB-GD-DTPA and Gd-BT-DO3A MR contrast studies in HCC patients and colorectal liver metastases. Infect Agent Cancer 14:40 Choi SH, Kim SY, Park SH, et al (2018) Diagnostic performance of CT, gadoxetate disodium-enhanced MRI, and PET/CT for the diagnosis of colorectal liver metastasis: Systematic review and meta-analysis. J Magn Reson Imaging 47(5):1237-1250 Mao Y, Chen B, Wang H, et al. (2020) performance of magnetic resonance imaging for colorectal liver metastasis: A systematic review and meta-analysis. Sci Rep 10(1):1969 Rizzo S, Radice D, Femia M, et al. (2018) Metastatic and non-metastatic lymph nodes: quantification and different distribution of iodine uptake assessed by dual-energy CT. Eur Radiol 28(2):760-769 Liu H, Yan F, Pan Z, et al. (2015) Evaluation of dual energy spectral CT in differentiating metastatic from non-metastatic lymph nodes in rectal cancer: Initial experience. Eur J Radiol 84(2):228-234 Lenga L, Czwikla R, Wichmann JL, et al (2018) Dual-energy CT in patients with colorectal cancer: Improved assessment of hypoattenuating liver metastases using noise-optimized virtual monoenergetic imaging. Eur J Radiol 106:184-191 Parakh A, Lennartz S, An C, et al (2021) Dual-Energy CT Images: Pearls and Pitfalls. Radiographics 41(1):98-119 Husarik DB, Gordic S, Desbiolles L, et al (2015) Advanced virtual monoenergetic computed tomography of hyperattenuating and hypoattenuating liver lesions: ex-vivo and patient experience in various body sizes. Invest Radiol 50(10):695-702 Lenga L, Lange M, Arendt CT, et al (2020) Measurement Reliability and Diagnostic Accuracy of Virtual Monoenergetic Dual-Energy CT in Patients with Colorectal Liver Metastases. Acad Radiol 27(7):e168-e175 Eisenhauer EA, Therasse P, Bogaerts J, et al (2009) New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 45(2):228-247 Lee T, Lee JM, Yoon JH, et al (2022) Deep learning-based image reconstruction of 40-keV virtual monoenergetic images of dual-energy CT for the assessment of hypoenhancing hepatic metastasis. Eur Radiol 32(9):6407-6417 Martin SS, Wichmann JL, Weyer H, et al (2017) Dual-energy computed tomography in patients with cutaneous malignant melanoma: Comparison of noise-optimized and traditional virtual monoenergetic imaging. Eur J Radiol 95:1-8 Martin SS, Wichmann JL, Weyer H, et al (2017) Dual-energy computed tomography in patients with cutaneous malignant melanoma: Comparison of noise-optimized and traditional virtual monoenergetic imaging. Eur J Radiol 95:1-8 Sudarski S, Apfaltrer P, Nance JW Jr, et al (2014) Objective and subjective image quality of liver parenchyma and hepatic metastases with virtual monoenergetic dual-source dual-energy CT reconstructions: an analysis in patients with gastrointestinal stromal tumor. Acad Radiol 21(4):514-522 Wichmann JL, Gillott MR, De Cecco CN, et al (2016) Dual-Energy Computed Tomography Angiography of the Lower Extremity Runoff: Impact of Noise-Optimized Virtual Monochromatic Imaging on Image Quality and Diagnostic Accuracy. Invest Radiol 51(2):139-146 De Cecco CN, Caruso D, Schoepf UJ, et al (2016) Optimization of window settings for virtual monoenergetic imaging in dual-energy CT of the liver: A multi-reader evaluation of standard monoenergetic and advanced imaged-based monoenergetic datasets. Eur J Radiol 85(4):695-699 Blyth S, Blakeborough A, Peterson M, Cameron IC, Majeed AW (2008) Sensitivity of magnetic resonance imaging in the detection of colorectal liver metastases. Ann R Coll Surg Engl 90(1):25-28 Maino C, Vernuccio F, Cannella R, et al (2023) Liver metastases: The role of magnetic resonance imaging. World J Gastroenterol 29(36):5180-5197 Yu MH, Lee JM, Hur BY, et al (2015) Gadoxetic acid-enhanced MRI and diffusion-weighted imaging for the detection of colorectal liver metastases after neoadjuvant chemotherapy. Eur Radiol 25(8):2428-2436 D.V. Cicchetti (1994) Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology, Psychol. Assess 6 (4) 284 Peijie Lv, Nana Liu, Luotong Wang, et al (2023) Deep learning image reconstruction for optimizing image quality of low-energy spectral monochromatic CT and detecting liver small low-contrast lesions. Chinese Journal of Medical Imaging Technology 39(1):104-108 Greffier J, Si-Mohamed S, Guiu B, et al (2022) Comparison of virtual monoenergetic imaging between a rapid kilovoltage switching dual-energy computed tomography with deep-learning and four dual-energy CTs with iterative reconstruction. Quant Imaging Med Surg 12(2):1149-1162 Lyu P, Liu N, Harrawood B, et al (2023) Is it possible to use low-dose deep learning reconstruction for the detection of liver metastases on CT routinely?. Eur Radiol 33(3):1629-1640 McCollough CH, Leng S, Yu L, Fletcher JG (2015) Dual- and Multi-Energy CT: Principles, Technical Approaches, and Clinical Applications. Radiology 276(3):637-653 Sun X, Niwa T, Nomura T, Yoshida R, Koyanagi K, Hashimoto J (2025) Low-keV Virtual Monoenergetic Imaging for Bronchial Artery Visualization on Photon-Counting Detector Computed Tomography. Diagnostics (Basel) 15(11):1354 Greffier J, Faby S, Pastor M, et al (2024) Comparison of low-energy virtual monoenergetic images between photon-counting CT and energy-integrating detectors CT: A phantom study. Diagn Interv Imaging 105(9):311-318 Dunning CAS, Rajendran K, Inoue A, et al (2023) Optimal Virtual Monoenergetic Photon Energy (keV) for Photon-Counting-Detector Computed Tomography Angiography. J Comput Assist Tomogr 47(4):569-575 Zhou W, Michalak GJ, Weaver JM, et al (2020) A Universal Protocol for Abdominal CT Examinations Performed on a Photon-Counting Detector CT System: A Feasibility Study. Invest Radiol 55(4):226-232 Rajeev R, Hussain HK (2025) Thermal Ablation versus Surgical Resection for Management of Small Colorectal Liver Metastases: Insights from the COLLISION Trial. Radiol Imaging Cancer 7(2):e259008 Akhan O, Akçalar S, Ünal E, Metin Y, Çiftçi T, Akıncı D (2023) Radiofrequency Ablation for Colorectal Cancer Liver Metastases: Outcomes and Prognostic Factors Associated with Survival. Turk J Gastroenterol 34(6):645-651 Charnsangavej C, Clary B, Fong Y, Grothey A, Pawlik TM, Choti MA (2006) Selection of patients for resection of hepatic colorectal metastases: expert consensus statement. Ann Surg Oncol 13(10):1261-1268 Dempsey PJ, Farrelly C, Cronin CG, Fenlon HM (2024) Preoperative imaging of colorectal liver metastases: what the radiologist and the multidisciplinary team need to know. Br J Radiol 97(1162):1602-1618 Hamady ZZ, Lodge JP, Welsh FK, et al (2014) One-millimeter cancer-free margin is curative for colorectal liver metastases: a propensity score case-match approach. Ann Surg 259(3):543-548 Schwartz LH, Litière S, de Vries E, et al (2016) RECIST 1.1-Update and clarification: From the RECIST committee. Eur J Cancer 62:132-137 Additional Declarations No competing interests reported. Supplementary Files TableS1.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8359303","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":561978128,"identity":"078e64da-9cb4-4ec7-b7fe-21d268f17382","order_by":0,"name":"Zhaoyang Zheng","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Zhaoyang","middleName":"","lastName":"Zheng","suffix":""},{"id":561978129,"identity":"3bc7098b-9717-4a27-83c8-a7c7531b3810","order_by":1,"name":"Jianyang Yang","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Jianyang","middleName":"","lastName":"Yang","suffix":""},{"id":561978130,"identity":"5c5d735c-35b0-475f-8eab-b2633593158e","order_by":2,"name":"Hong Zhu","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Hong","middleName":"","lastName":"Zhu","suffix":""},{"id":561978131,"identity":"90c292b7-f6ce-433e-8033-3d61ab53b9e3","order_by":3,"name":"Xi Feng","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Xi","middleName":"","lastName":"Feng","suffix":""},{"id":561978132,"identity":"6b07a525-9416-42e4-b0c1-c2efd8a840c6","order_by":4,"name":"Jiayu Sun","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Jiayu","middleName":"","lastName":"Sun","suffix":""},{"id":561978133,"identity":"174188eb-eb17-40fb-8888-cae15c52022d","order_by":5,"name":"Lie Yang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtklEQVRIie3QrQ7CMBDA8Wua3EzZbBMIvMIRLIJHOQxT85NgOsMDoHgM9C2zewPMHmFyho8aNIdD9Je0OdF/0hYgSf5QbuM21tslZo3oEoyJufSHTe56ViZx2Vno9le/I2WSWQGHtgweGKb6prkYMniHVZifxJz7uyZxBORdFRbC1gRVUozA5Ev0TNrEAQgT/5IgmaPwOsRPblVvKYpusI/na7VqmnaYakUSZdNnEtX5JEmS5Ls37QwxjoN2JKAAAAAASUVORK5CYII=","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":true,"prefix":"","firstName":"Lie","middleName":"","lastName":"Yang","suffix":""}],"badges":[],"createdAt":"2025-12-14 16:23:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8359303/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8359303/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":98779991,"identity":"2d88d66b-ac7d-4e9d-82f6-c202f611a8c4","added_by":"auto","created_at":"2025-12-22 12:30:58","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5119022,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/913d280c42ed07c5952ae747.docx"},{"id":98766110,"identity":"9471ed19-9a50-4896-9247-ce8d8bdb9b40","added_by":"auto","created_at":"2025-12-22 10:14:25","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8142,"visible":true,"origin":"","legend":"","description":"","filename":"aed412d7a723407fb0a447c4fdd36fee.json","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/6fb1c79cd8e06ba8635e51bc.json"},{"id":98766112,"identity":"0664a1cb-8921-4d5f-a03a-b8a61da7365d","added_by":"auto","created_at":"2025-12-22 10:14:25","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13525,"visible":true,"origin":"","legend":"","description":"","filename":"TableS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/665a6eb69187344de0d4df4b.docx"},{"id":98766115,"identity":"ea780b74-7482-441f-8f83-767742663794","added_by":"auto","created_at":"2025-12-22 10:14:25","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":129917,"visible":true,"origin":"","legend":"","description":"","filename":"aed412d7a723407fb0a447c4fdd36fee1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/6d14b13fd8af193ac16e9138.xml"},{"id":98766119,"identity":"f9e1115a-a090-466e-9448-1b36c38ed463","added_by":"auto","created_at":"2025-12-22 10:14:26","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1576654,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/141b873b5723da1326be7eeb.jpeg"},{"id":98766120,"identity":"7b5d9110-7328-4795-b7a8-cb1467f2a4e3","added_by":"auto","created_at":"2025-12-22 10:14:26","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":247625,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/9744db8cb44566210091854e.jpeg"},{"id":98779793,"identity":"b08bc00d-0b10-4141-a868-34b90cb24e6a","added_by":"auto","created_at":"2025-12-22 12:30:46","extension":"emf","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":102240,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.emf","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/a22982d24e8878fb64cc3c43.emf"},{"id":98778976,"identity":"24e4cea8-cd8c-4794-bdcc-fe3dc6004369","added_by":"auto","created_at":"2025-12-22 12:29:51","extension":"emf","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9814116,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.emf","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/e8cbb72e70a2a4247d64955a.emf"},{"id":98766123,"identity":"c242a925-7c19-4e01-8734-ec53f5a52f7b","added_by":"auto","created_at":"2025-12-22 10:14:26","extension":"emf","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":502044,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.emf","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/f31d2bc27adedd111696d940.emf"},{"id":98778960,"identity":"a8b3843b-6e3b-4b73-bb76-e3e9d96dab6e","added_by":"auto","created_at":"2025-12-22 12:29:51","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1328844,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/cd373e857bab2bf6f4c54847.png"},{"id":98780702,"identity":"547fa21f-fefa-461e-9468-c8954d5ef4b4","added_by":"auto","created_at":"2025-12-22 12:31:34","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36070,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/f122c69f5b618b91dd1f537d.png"},{"id":98766126,"identity":"39627eeb-3c7c-4b8b-9af3-9fd38a3e31ca","added_by":"auto","created_at":"2025-12-22 10:14:26","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":121256,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/040b7f1db214881af3d1a780.png"},{"id":98780390,"identity":"d4d5cb28-b1c5-4644-bef4-763e8d72d381","added_by":"auto","created_at":"2025-12-22 12:31:17","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1087847,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/9cff54db615e046fc2f3662c.png"},{"id":98766128,"identity":"16d7e991-37ea-4fc9-a6d2-ceb118dbc9f3","added_by":"auto","created_at":"2025-12-22 10:14:26","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":999508,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/24e8e622c48137f3463a6b07.png"},{"id":98779686,"identity":"a89a6574-1e6e-48af-a30a-00b8a2797ebc","added_by":"auto","created_at":"2025-12-22 12:30:36","extension":"xml","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":130333,"visible":true,"origin":"","legend":"","description":"","filename":"aed412d7a723407fb0a447c4fdd36fee1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/95b2a3b5be6f7508622e278b.xml"},{"id":98766130,"identity":"63d3ba3d-8994-4cf3-a283-f1dbfdc4386c","added_by":"auto","created_at":"2025-12-22 10:14:26","extension":"html","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":143610,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/51713fbf6548e9c3f698d8e2.html"},{"id":98780363,"identity":"4c963929-466f-4fed-b135-9da91a2ca223","added_by":"auto","created_at":"2025-12-22 12:31:15","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2159281,"visible":true,"origin":"","legend":"\u003cp\u003eBox-and-whisker plots show average SNR and CNR in patients with CRLM. Boxes represent 25th and 75th percentile, horizontal lines 50th percentile (median), and whiskers minimum and maximum values. Quantitative image quality parameters were compared between standard linearly-blended (M_0.6) image series and VMI series at energy levels ranging from 40 keV to 70 keV.\u003c/p\u003e\n\u003cp\u003eSNR, signal-to-noise ratio; CNR, contrast-to-noise ratio; CRLM, colorectal cancer with liver metastases.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/50ae1ad16e95499427a27836.png"},{"id":98780015,"identity":"6f38049e-6d27-46d2-b492-e233f5d6a8fe","added_by":"auto","created_at":"2025-12-22 12:30:58","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":371301,"visible":true,"origin":"","legend":"\u003cp\u003eROC curves of M_0.6 images, VMI series at energy levels ranging from 40 keV to 70 keV, and combined approach (VMI 40–70) for mean attenuation in diagnosing liver metastases of CRC. \u003cstrong\u003ea\u003c/strong\u003eStandard linearly-blended (M_0.6) image series and four VMI images. \u003cstrong\u003eb\u003c/strong\u003eCombined approach.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eCRC, colorectal cancer.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/7ae410f8b9ba250558d20729.png"},{"id":98766116,"identity":"339833cb-aa88-41b9-8796-7c459e267633","added_by":"auto","created_at":"2025-12-22 10:14:25","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":3729229,"visible":true,"origin":"","legend":"\u003cp\u003eROC curves of M_0.6 images, VMI series at energy levels ranging from 40 keV to 70 keV, and combined approach (VMI 40–70) for SNR in diagnosing liver metastases of CRC. \u003cstrong\u003ea\u003c/strong\u003e Standard linearly-blended (M_0.6) image series and four VMI images. \u003cstrong\u003eb\u003c/strong\u003e Combined approach.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eSNR, signal-to-noise ratio; CRC, colorectal cancer.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/71d3181ac04a082ec880d2c9.png"},{"id":98780384,"identity":"fdb8b875-dd5d-4ddc-bb9d-831cb4926071","added_by":"auto","created_at":"2025-12-22 12:31:16","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":4306582,"visible":true,"origin":"","legend":"\u003cp\u003eTypical case of a 62-year-old male patient with left-lobe liver metastases of CRC. \u003cstrong\u003ea\u003c/strong\u003e M_0.6 image. \u003cstrong\u003eb\u003c/strong\u003e VMI 40 image. \u003cstrong\u003ec\u003c/strong\u003e VMI 50 image. \u003cstrong\u003ed\u003c/strong\u003e VMI 60 image. \u003cstrong\u003ee\u003c/strong\u003eVMI 70 image.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eCRC, colorectal cancer.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/f9f7030cb015cb11557144d3.png"},{"id":98779632,"identity":"efb77083-34cd-4c75-84d4-b9a81d76870e","added_by":"auto","created_at":"2025-12-22 12:30:34","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":5005623,"visible":true,"origin":"","legend":"\u003cp\u003eA 57-year-old male patient with CRC and multiple liver metastases. Images in the portal venous phase are shown. Representative metastatic lesions (n=3) are selected and indicated by arrows. The lesions are better delineated on VMI 40 image (\u003cstrong\u003eb\u003c/strong\u003e) than on M_0.6 image (\u003cstrong\u003ea\u003c/strong\u003e), though with less clarity compared to reference MRI image (\u003cstrong\u003ec\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eCRC, colorectal cancer.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/ccdfe6b85e28f0b5e5a204d0.png"},{"id":101751885,"identity":"7a58ed03-51e6-4daf-82b8-2dddc2b08003","added_by":"auto","created_at":"2026-02-03 10:24:13","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":16496608,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/a286bcc8-1231-4691-bf0f-789ef7fd1ad8.pdf"},{"id":98766114,"identity":"e812d7f3-568d-4eac-b731-63f29421f830","added_by":"auto","created_at":"2025-12-22 10:14:25","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":13525,"visible":true,"origin":"","legend":"","description":"","filename":"TableS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8359303/v1/9687600eff044e4abfa05392.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Assessment of Diagnostic Accuracy and Measurement Reliability of Low-keV Virtual Monoenergetic Dual-energy CT in The Liver Metastases of Colorectal Cancer: A Prospective, Imaging-Reference Study","fulltext":[{"header":"Introduction","content":"\u003cp\u003eColorectal cancer (CRC) is the third most common cancer and the second leading cause of cancer-related death worldwide [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. The liver remains the predominant site of hematogenous metastasis, with 15%\u0026ndash;25% of patients presenting synchronous metastases at initial diagnosis and 15%\u0026ndash;25% developing metachronous metastases after resection of the primary colorectal tumor [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Liver metastases has been the most common cause of CRC-related death, and patients with unresectable metastatic lesions have a five-year survival rate of less than 5% [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. The length of survival is significantly influenced by the extent of surgical removal of the hepatic metastases [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Achieving complete resection or no-evidence-of-disease (NED) status for liver metastases of CRC has been shown to increase five-year survival to 30%\u0026ndash;57% [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]; however, this benefit depends largely on precise pre-operative tumor imaging.\u003c/p\u003e \u003cp\u003eThere have been significant advancements in the imaging diagnosis of colorectal cancer with liver metastases (CRLM) in recent years. Although contrast-enhanced computed tomography (CECT) remains the most commonly used modality for screening and staging [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], its ability to detect sub-centimeter lesions is limited. The sensitivity of CECT for detecting liver metastases smaller than 10 mm ranges from 22% to 68%, which directly affects the assessment of surgical resectability and patient prognosis [\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Nevertheless, CT remains crucial for initial diagnosis. Its capacity to detect small lesions during first-line examination usually prompts referral to magnetic resonance imaging (MRI) for definitive assessment. This helps prevent missed diagnosis, particularly for sub-centimeter metastases. Gadoxetic acid disodium-enhanced MRI has demonstrated superior performance in the diagnosis of liver metastases of CRC, achieving a sensitivity from 90.0% to 96.0% and a specificity of approximately 90.0%, particularly in identifying sub-centimeter lesions [\u003cspan additionalcitationids=\"CR12 CR13\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. However, MRI has inherent limitations, including higher cost, longer examination times, and greater patient cooperation. Moreover, its use is restricted in patients with metallic implants or severe renal dysfunction. Consequently, more efficient and widely accessible alternatives need to be explored.\u003c/p\u003e \u003cp\u003eDual-energy CT (DECT) has emerged as an innovative and promising imaging modality. This technology enables the reconstruction of iodine concentration maps, virtual non-contrast images, and Z-effective maps, which are invaluable for quantifying iodine uptake and for differentiating hypovascular metastases from normal tissue, particularly when lesions are small or poorly attenuated [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Virtual monoenergetic imaging (VMI), a crucial post-processing technique of DECT, improves the visibility of iodine uptake. In particular, VMI at low kiloelectron volts (keV) levels increases contrast-to-noise ratio (CNR) between lesions and liver tissue without the need for additional radiation or contrast agent. This reconstruction approach decreases potential health risks and promotes the detectability of both hypo- and hyper-vascular hepatic lesions [\u003cspan additionalcitationids=\"CR18\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Prior single-center retrospective studies have shown that 40-keV and 60-keV VMI improves image quality and the detection of liver metastases of CRC [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]; however, prospective reference data are lacking, particularly for small metastases (\u0026lt;\u0026thinsp;10 mm).\u003c/p\u003e \u003cp\u003eWe therefore conducted a prospective imaging-reference study to assess the value of low-keV (40\u0026ndash;70 keV) virtual monoenergetic DECT. The objectives were 1) to improve both quantitative and qualitative image quality in comparison with CECT, 2) to increase diagnostic accuracy in the detection of liver metastases of CRC, especially sub-centimeter ones, and 3) to maintain measurement reliability in treatment-response evaluation of liver lesions according to RECIST 1.1 criteria[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatient Selection\u003c/h2\u003e \u003cp\u003eBetween January 2022 and December 2023, consecutive patients were enrolled if they had histologically confirmed CRC on colonoscopy and were highly suspected of having CRLM on abdominal DECT. MRI was performed within one week after DECT examination.\u003c/p\u003e \u003cp\u003eInclusion criteria: 1) Age\u0026thinsp;\u0026ge;\u0026thinsp;18 years; 2) No history of radical hepatic resection for malignancy.\u003c/p\u003e \u003cp\u003eExclusion criteria: 1) Severe image artefacts; 2) Contraindications to MRI or contrast agents; 3) Protocol deviations during DECT or contrast administration; 4) Confirmed or suspected pregnancy; 5) Estimated glomerular filtration rate \u0026lt; 45 mL/min.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSample-size Calculation\u003c/h3\u003e\n\u003cp\u003eOn the basis of previous per-patient analyses that reported 91% sensitivity and 94\u0026ndash;97% specificity for DECT in the detection of liver metastases [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], we conservatively estimated both parameters [p and (1-p)] as 90% in the present study. This estimation used a self-paired design (metastatic versus normal liver tissue) with a margin of error (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\delta\\:}\\)\u003c/span\u003e\u003c/span\u003e) of 0.10 and an alpha (α) of 0.05. The sample-size formula was as follows:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\text{n}=\\left(\\frac{{\\text{z}}_{{\\alpha\\:}}}{{\\delta\\:}}\\right)\u0026sup2;\\left(1-\\text{p}\\right)\\text{p}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe minimum required sample size was 35 patients.\u003c/p\u003e\n\u003ch3\u003eDECT Protocol\u003c/h3\u003e\n\u003cp\u003ePatients fasted for 4\u0026ndash;6 hours before CT and ingested 200 ml of water prior to the scan. Non-enhanced and double-phase-enhanced scans were performed with a DECT scanner (Spectral CT 7500, Philips Healthcare, Netherlands). Patients were positioned supine and scanned from the top of the diaphragm to the inferior border of the pubic symphysis. A total of 1.2 mL/kg body weight of contrast medium was injected into the middle elbow vein using a double-barrelled high-pressure syringe at a flow rate of 3 mL/s. High-dose tracking was initiated with a threshold set at 150 Hounsfield units in the abdominal aorta. The arterial phase scan was automatically triggered upon reaching peak contrast concentration, and the portal phase scan was initiated 70 seconds post-contrast injection. The scanning parameters were as follows: tube voltage, 120 kVp; tube current, modulated by automatic exposure control; pitch, 1.000; rotation time, 0.5 seconds; collimation, 128 \u0026times; 0.625 mm; field of view, 350 mm \u0026times; 350 mm; matrix, 512 \u0026times; 512.\u003c/p\u003e\n\u003ch3\u003eDECT Image Reconstruction\u003c/h3\u003e\n\u003cp\u003eStandard linearly-blended (M_0.6) images were automatically generated with a blending factor of 0.6 to simulate CECT (conventional 120-kV single-energy) images [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Spectral based images were reconstructed with the spectral level-4 algorithm. These spectral datasets were transferred to a dedicated spectral post-processing workstation (IntelliSpace Portal, Version 12.0; Philips Healthcare, Amsterdam, The Netherlands). Monoenergetic reconstructions were generated with the VMI algorithm. We solely reconstructed four VMI image series (40, 50, 60, 70 keV) because lesion conspicuity and CNR would be considered too low beyond 70 keV, as reported previously [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan additionalcitationids=\"CR25\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e\n\u003ch3\u003eQuantitative Analysis of DECT Datasets\u003c/h3\u003e\n\u003cp\u003eQuantitative analysis was conducted by two radiologists with 3 years of experience in oncological abdominal DECT imaging. Signal attenuation (mean Hounsfield units, HU) and standard deviation (SD) were measured by placing circular regions of interest (ROI) within the liver metastases and adjacent, unaffected liver parenchyma (100 mm\u0026sup2;). Each ROI was placed on the largest regions of the lesion with the most significant enhancement in the portal venous phase, avoiding areas of focal heterogeneity and tumor necrosis. The size of the tumor ROI was adjusted to the tumor\u0026rsquo;s overall dimensions for precise depiction. Additionally, homogeneous ROIs were placed within the psoas muscle to evaluate image contrast (100 mm\u003csup\u003e2\u003c/sup\u003e). All measurements were performed three times in three consecutive slices, and average values were calculated. The signal-to-noise ratio (SNR) and CNR calculations were as follows:\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eSNR\u0026thinsp;=\u0026thinsp;HU \u003csub\u003eROI\u003c/sub\u003e/SD \u003csub\u003eROI\u003c/sub\u003e\u003c/h2\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003eCNR=|HU \u003csub\u003eROI\u003c/sub\u003e\u0026minus;HU \u003csub\u003emuscle\u003c/sub\u003e|/SD \u003csub\u003emuscle\u003c/sub\u003e\u003c/h2\u003e \u003cdiv id=\"Sec10\" class=\"Section4\"\u003e \u003ch2\u003eQualitative Analysis of DECT Datasets\u003c/h2\u003e \u003cp\u003eQualitative analysis was performed by two radiologists with 15 years of abdominal CT interpretation experience. The readers were blinded to the reconstruction type being assessed. Standard linearly-blended (M_0.6) images and VMI reconstructions were rated in random order, with only one single image series rated during each reading session. Readers were permitted to adjust standard window settings manually, as prior studies suggest different window configurations for the VMI algorithm [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Overall image quality (ranging from 1\u0026thinsp;=\u0026thinsp;poor overall image quality to 5\u0026thinsp;=\u0026thinsp;excellent overall image quality), lesion delineation (ranging from 1\u0026thinsp;=\u0026thinsp;lesion cannot be ruled out to 5\u0026thinsp;=\u0026thinsp;excellent lesion margin delineation), and image noise (ranging from 1\u0026thinsp;=\u0026thinsp;enormous image noise to 5\u0026thinsp;=\u0026thinsp;no relevant image noise perceivable) were scored on 5-point Likert scales.\u003c/p\u003e \u003cp\u003e \u003cb\u003eReference Standard\u003c/b\u003e \u003c/p\u003e \u003cp\u003eGadoxetic acid disodium-enhanced MRI combined with \u0026ge;\u0026thinsp;6 months of imaging follow-up was used as the reference standard for confirming CRLM. Although histopathology is the ideal gold standard, it was not universally available due to the prospective nature of this study. MRI was considered the first-line modality and the \u0026ldquo;gold standard\u0026rdquo; for pre-operative imaging because of its higher sensitivity and specificity for CRLM detection compared with CECT and fluoro-2-deoxyglucose positron emission tomography (FDG-PET) [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. All indeterminate lesions were independently reviewed by two senior abdominal radiologists and confirmed through follow-up imaging.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eTreatment-response Assessment\u003c/h2\u003e \u003cp\u003eThe treatment modality was determined at initial evaluation. For resectable or potentially resectable metastases, surgical resection or neoadjuvant chemotherapy followed by surgery was recommended. Patients were reassessed with DECT and MRI after 4\u0026ndash;6 cycles of systemic chemotherapy. Target lesions were selected according to RECIST 1.1 [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Four response categories were recorded: complete response, partial response, progressive disease, and stable disease. The sum of the longest diameters (SLD) was compared between baseline and follow-up.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eStatistical analyses were conducted using SPSS 20.0 (IBM, Corp., Armonk, NY, USA) and MedCalc (version 20.218). Normally distributed data are expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD, whereas non-normally distributed data are expressed as median and interquartile range (IQR). The Kolmogorov-Smirnov test was used to assess Gaussian distribution. Analysis of variance (ANOVA) was applied for multiple comparisons of normally distributed variables, while the Wilcoxon matched pairs test was used for non-Gaussian data distribution. Inter-reader agreement was assessed with the intraclass correlation coefficient (ICC) using a two-way mixed model and 95% confidence intervals (CIs). ICC values\u0026thinsp;\u0026lt;\u0026thinsp;0.40 were considered poor, values ranging from 0.40\u0026ndash;0.59 as fair, 0.60\u0026ndash;0.74 as good, and 0.75\u0026ndash;1.0 as excellent agreement [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Receiver operating characteristic (ROC) curves were constructed, and the area under the curve (AUC) was calculated. Optimal sensitivity and specificity were determined with the maximum Youden index. The four VMI series (40\u0026ndash;70 keV) were entered into a multivariable logistic regression model to assess the diagnostic value of the combined approach. The Chi-square test was used to compare the detection rates of liver metastases of CRC between groups.\u003c/p\u003e \u003cp\u003eA \u003cem\u003ep\u003c/em\u003e-value๤0.05 was considered statistically significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003ePatient Characteristics\u003c/h2\u003e\n \u003cp\u003eThirty-five patients (22 male, 13 female; mean age 61.0\u0026thinsp;\u0026plusmn;\u0026thinsp;12.2 years) were enrolled. A total of 125 liver metastases of CRC were confirmed: \u0026ge; 5 lesions in 11 patients, \u0026lt; 5 lesions in 11 patients, and a single lesion in 13 patients. Mean lesion size was 27.4\u0026thinsp;\u0026plusmn;\u0026thinsp;17.9 mm. Tumor characteristics and patient demographics are summarized in Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\n \u003ctable id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003ePatient Characteristics\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eParameter\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e22 (60.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13 (61.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAge (years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e61.0\u0026thinsp;\u0026plusmn;\u0026thinsp;12.2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo. of liver metastases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026ge; 5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11 (31.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e๤5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11 (31.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSingular\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13 (37.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTotal number\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e125\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMean tumor size (mm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e27.4\u0026thinsp;\u0026plusmn;\u0026thinsp;17.9\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRadiation dose\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCTDIvol (mGy)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e41.6\u0026thinsp;\u0026plusmn;\u0026thinsp;12.6\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDLP (mGy\u0026middot;cm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2108.1\u0026thinsp;\u0026plusmn;\u0026thinsp;463.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\"\u003eCTDIvol, volume CT dose index; DLP, dose-length product.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003eQuantitative Image Quality\u003c/h2\u003e\n \u003cp\u003eQuantitative analysis of the DECT datasets revealed that both SNR (average) and CNR (liver metastases) peaked with 40-keV VMI (15.5\u0026thinsp;\u0026plusmn;\u0026thinsp;9.5 and 6.2\u0026thinsp;\u0026plusmn;\u0026thinsp;4.3, respectively) and were significantly higher than those with M_0.6 (7.0\u0026thinsp;\u0026plusmn;\u0026thinsp;3.3 and 1.5\u0026thinsp;\u0026plusmn;\u0026thinsp;1.2; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) and all other VMI series. Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e details mean attenuation, SNR, and CNR values in patients with CRLM. Figure \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e shows box-and-whisker plots of SNR and CNR distributions.\u003c/p\u003e\n \u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eComparison of mean attenuation, SNR and CNR of patients with CRLM. Quantitative image analysis included standard linearly-blended (M_0.6) image series and VMI series at energy levels ranging from 40 keV to 70 keV.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eM_0.6\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVMI 40\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVMI 50\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVMI 60\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVMI 70\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAttenuation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLiver parenchyma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e118.1\u0026thinsp;\u0026plusmn;\u0026thinsp;17.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e275.8\u0026thinsp;\u0026plusmn;\u0026thinsp;44.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e195.1\u0026thinsp;\u0026plusmn;\u0026thinsp;30.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e147.7\u0026thinsp;\u0026plusmn;\u0026thinsp;22.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e119.5\u0026thinsp;\u0026plusmn;\u0026thinsp;18.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLiver metastases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e70.75\u0026thinsp;\u0026plusmn;\u0026thinsp;20.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e157.0\u0026thinsp;\u0026plusmn;\u0026thinsp;51.3\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e112.6\u0026thinsp;\u0026plusmn;\u0026thinsp;36.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e85.5\u0026thinsp;\u0026plusmn;\u0026thinsp;27.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e69.4\u0026thinsp;\u0026plusmn;\u0026thinsp;22.9\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSNR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLiver parenchyma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9.2\u0026thinsp;\u0026plusmn;\u0026thinsp;2.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e22.9\u0026thinsp;\u0026plusmn;\u0026thinsp;6.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17.2\u0026thinsp;\u0026plusmn;\u0026thinsp;5.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13.4\u0026thinsp;\u0026plusmn;\u0026thinsp;4.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11.1\u0026thinsp;\u0026plusmn;\u0026thinsp;3.4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLiver metastases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.9\u0026thinsp;\u0026plusmn;\u0026thinsp;2.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7.9\u0026thinsp;\u0026plusmn;\u0026thinsp;4.0\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7.0\u0026thinsp;\u0026plusmn;\u0026thinsp;3.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6.1\u0026thinsp;\u0026plusmn;\u0026thinsp;2.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5.4\u0026thinsp;\u0026plusmn;\u0026thinsp;2.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAverage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7.0\u0026thinsp;\u0026plusmn;\u0026thinsp;3.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e15.5\u0026thinsp;\u0026plusmn;\u0026thinsp;9.5\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12.2\u0026thinsp;\u0026plusmn;\u0026thinsp;7.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9.9\u0026thinsp;\u0026plusmn;\u0026thinsp;5.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e8.5\u0026thinsp;\u0026plusmn;\u0026thinsp;5.6\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCNR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLiver metastases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.5\u0026thinsp;\u0026plusmn;\u0026thinsp;1.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6.2\u0026thinsp;\u0026plusmn;\u0026thinsp;4.3\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.0\u0026thinsp;\u0026plusmn;\u0026thinsp;3.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.5\u0026thinsp;\u0026plusmn;\u0026thinsp;2.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.8\u0026thinsp;\u0026plusmn;\u0026thinsp;1.4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"6\"\u003eSNR, signal-to-noise ratio; CNR, contrast-to-noise ratio; CRLM, colorectal cancer with liver metastases.\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"6\"\u003e\u003csup\u003e*\u003c/sup\u003e indicates significance at \u003cem\u003ep\u003c/em\u003e๤0.05 (VMI 40 vs. other image series)\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"6\"\u003eCRC, colorectal cancer.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n \u003cp\u003eROC analysis revealed no significant differences (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05) in AUCs among M_0.6 and 40-, 50-, 60-, and 70-keV VMI images for mean attenuation in diagnosing CRLM (AUCs: 0.955, 0.966, 0.959, 0.962, and 0.958, respectively). The combined 40\u0026ndash;70 keV VMI approach yielded an AUC of 0.969 (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). AUCs for SNR were 0.893 (M_0.6), 0.974 (40 keV), 0.958 (50 keV), 0.939 (60 keV), and 0.914 (70 keV), with statistically significant differences (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01). M_0.6 images showed the lowest AUC, whereas 40-keV VMI achieved the highest. The combined 40\u0026ndash;70 keV VMI approach achieved an AUC of 0.987 (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). Supplementary Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e lists the thresholds, specificity, and sensitivity derived from the ROC curves. A representative case is shown in Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003eQualitative Image Quality\u003c/h2\u003e\n \u003cp\u003eTable \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e summarizes the qualitative image-analysis parameters. Overall image quality was comparable between 60-keV VMI and M_0.6 images (median 5 vs 5, \u003cem\u003ep\u003c/em\u003e๥0.05), both reached the highest scores, with excellent inter-reader agreement (60-keV VMI: ICC, 0.82; 95% CI, 0.75\u0026ndash;0.87; M_0.6: ICC, 0.84; 95% CI, 0.78\u0026ndash;0.89).\u003c/p\u003e\n \u003ctable id=\"Tab3\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eQualitative image quality parameters are given as medians and ranges in parentheses. Standard linearly-blended (M_0.6) images and VMI series at energy levels ranging from 40 keV to 70 keV were subjectively evaluated in terms of overall image quality, lesion delineation, and image noise by two blinded radiologists.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eM_0.6\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVMI 40\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVMI 50\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVMI 60\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVMI 70\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eOverall image quality\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5 (3\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4 (3\u0026ndash;4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4 (3\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5 (3\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4 (3\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLesion delineation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3 (1\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5 (1\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5 (1\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4 (1\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4 (1\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eImage noise\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4 (2\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3 (2\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4 (2\u0026ndash;4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4 (3\u0026ndash;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3 (2\u0026ndash;4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003eLesion delineation scores were highest with 40-keV and 50-keV VMI (median 5 vs 5; \u003cem\u003ep\u003c/em\u003e๥0.05) and were significantly higher than those for M_0.6 images (median 3, range 1\u0026ndash;5; \u003cem\u003ep\u003c/em\u003e๤0.01), with excellent inter-reader agreement (40-keV VMI: ICC, 0.96; 95% CI, 0.94\u0026ndash;0.97; 50-keV VMI: ICC, 0.92; 95% CI, 0.89\u0026ndash;0.94).\u003c/p\u003e\n \u003cp\u003eThe lowest image-noise scores (i.e., greatest noise) were observed with 70-keV VMI (median 3, range 2\u0026ndash;4; ICC 0.71; 95% CI 0.59\u0026ndash;0.79). By contrast, 60-keV VMI achieved the highest image-noise scores (median 4, range 3\u0026ndash;5; \u003cem\u003ep\u003c/em\u003e๤0.01) with excellent inter-reader agreement (ICC, 0.98; 95% CI, 0.98\u0026ndash;0.99). Overall inter-reader agreement for image quality, lesion delineation and image noise was excellent (ICC, 0.86; 95% CI, 0.84\u0026ndash;0.87).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003eLesion Detection\u003c/h2\u003e\n \u003cp\u003eA total of 125 liver metastases of CRC were detected and included in the analysis. No significant differences were observed for lesions\u0026thinsp;\u0026gt;\u0026thinsp;10 mm between M_0.6 and any VMI series (\u003cem\u003ep\u003c/em\u003e๥0.05). For lesions\u0026thinsp;\u0026le;\u0026thinsp;10 mm (n\u0026thinsp;=\u0026thinsp;15), per-lesion sensitivity was 60.0% with M_0.6, 86.7% with 40-keV VMI, 80.0% with 50-keV VMI, 80.0% with 60-keV VMI, and 73.3% with 70-keV VMI. 40-keV VMI was significantly superior to M_0.6 images (\u003cem\u003ep\u003c/em\u003e๤0.05). Detailed data are provided in Table \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e. Lesion conspicuity across image series is compared in Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e.\u003c/p\u003e\n \u003ctable id=\"Tab4\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eComparison of M_0.6 images, VMI series at energy levels ranging from 40 keV to 70 keV in the detection rates of liver metastases of CRC.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u0026le;\u0026thinsp;10mm\u003c/p\u003e\n \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;15)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e11\u0026ndash;19mm\u003c/p\u003e\n \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;33)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u0026ge;\u0026thinsp;20mm\u003c/p\u003e\n \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;77)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVMI 40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13 (86.7%) \u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e30 (90.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e77 (100.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVMI 50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12 (80.0%) \u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e31 (93.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e77 (100.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVMI 60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12 (80.0%) \u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e31 (93.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e77 (100.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVMI 70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11 (73.3%) \u003csup\u003ea,b\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e29 (87.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e77 (100.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eM_0.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9 (60.0%) \u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e27 (81.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e74 (96.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e๤0.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e๥0.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e๥0.05\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\"\u003eCRC, colorectal cancer.\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\"\u003eThe same superscript letters indicate no significant difference (\u003cem\u003ep\u003c/em\u003e๥0.05), and different superscript letters indicate significant differences (\u003cem\u003ep\u003c/em\u003e๤0.05).\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\"\u003eCRC, colorectal cancer.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n \u003ch2\u003eTreatment-Response Assessment\u003c/h2\u003e\n \u003cp\u003eTable \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e lists the treatments administered to all 35 patients. Three patients underwent radical surgery without receiving chemotherapy. Thirty-two patients received systemic chemotherapy with or without targeted therapy; 17 of them subsequently underwent radical surgery, and 1 underwent transcatheter arterial chemoembolization.\u003c/p\u003e\n \u003ctable id=\"Tab5\" border=\"1\" class=\"fr-table-selection-hover\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eTreatment and tumor response assessment in patients with CRLM.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eThirty-five CRLM patients\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eData\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInitial resectability assessment\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eResectable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3 (8.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUnresectable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e15 (42.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePotential-resectable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e17 (48.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDistribution\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSolitary\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9 (25.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMultiple in a single lobe\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e8 (22.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMultiple in both lobes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e18 (51.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTreatment process\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRadical surgery\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3 (8.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eChemotherapy \u0026rarr; radical surgery\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e17 (48.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eChemotherapy\u0026thinsp;\u0026plusmn;\u0026thinsp;targeted therapy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e14 (40.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eChemotherapy \u0026rarr; TACE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1 (2.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSLD before treatment (mm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMRI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e32.2\u0026thinsp;\u0026plusmn;\u0026thinsp;22.8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDECT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e33.5\u0026thinsp;\u0026plusmn;\u0026thinsp;23.3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026gt;0.05\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSLD after treatment (mm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMRI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e27.0\u0026thinsp;\u0026plusmn;\u0026thinsp;22.4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDECT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e28.8\u0026thinsp;\u0026plusmn;\u0026thinsp;22.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026gt;0.05\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSLD difference before and after treatment (mm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMRI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.2\u0026thinsp;\u0026plusmn;\u0026thinsp;5.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDECT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e4.6\u0026thinsp;\u0026plusmn;\u0026thinsp;6.2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.41\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTumor response assessment\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eStable disease\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMRI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e25 (71.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDECT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e27 (77.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePartial response\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMRI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e7 (20.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDECT\u003c/p\u003e\n \u003cp\u003eNo detection of metastases\u003c/p\u003e\n \u003cp\u003eDECT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5 (14.3%)\u003c/p\u003e\n \u003cp\u003e1 (2.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\"\u003eCRLM, colorectal cancer with liver metastases; TACE, transcatheter arterial chemoembolization; SLD, sum of longest diameters.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n \u003cp\u003eMean SLD reduction was 5.2\u0026thinsp;\u0026plusmn;\u0026thinsp;5.5 mm with MRI and 4.6\u0026thinsp;\u0026plusmn;\u0026thinsp;6.2 mm with 40-keV VMI (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.41). According to tumor response assessment, MRI classified 7 patients as having partial response and 25 as stable disease, whereas DECT (40-keV VMI) classified 5 as partial response and 27 as stable disease; lesions were undetected in one patient. Response classification differed between DECT (40-keV VMI) and MRI in 3 of the 35 patients (8.57%), all with a measurement difference of less than 5 mm.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eCRLM remains a key determinant of survival; however CECT may miss up to 40% of sub-centimeter liver metastases. As a result, curable tumors may be neglected at a stage when radical resection is still possible. We demonstrates that low-keV VMI (40\u0026ndash;70 keV) reconstructed from DECT significantly increases conspicuity between metastatic lesions and background liver parenchyma. This improvement leads to a maximum 20% increase in per-lesion sensitivity for detecting liver metastases, without loss of specificity. Moreover, the detection of sub-centimeter metastases rises by 13\u0026ndash;26%, without additional radiation or contrast medium. Importantly, the technique also allows for accurate measurement of chemotherapeutic response. By transforming a routine scan into a high-definition \u0026ldquo;digital biopsy\u0026rdquo;, low-keV VMI provides superior surgical guidance. This enables more patients to become eligible for complete (R0) hepatectomy and reduces inter-observer variability during post-chemotherapy monitoring. Thus, we consider the reconstruction approach to become a practical and cost-effective strategy for improving oncological outcomes.\u003c/p\u003e \u003cp\u003eThe present study further explores the value of low-keV VMI in optimizing image quality for liver metastases of CRC. Lesion visibility on CT images is primarily affected by image noise, tissue contrast, and CNR [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. DECT enhances diagnostic confidence for the detection of sub-centimeter liver metastases by employing multi-energy reconstruction to improve CNR and suppress image noise [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Although lower-keV VMI affords higher tissue contrast compared to higher energy levels, it also amplifies image noise, which may obscure the margins of small or hypovascular lesions. Conversely, as keV increases, image noise diminishes but contrast is sacrificed [\u003cspan additionalcitationids=\"CR36\" citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. Therefore, identification of an optimal monoenergetic level that balances these competing demands is critical for maximizing lesion conspicuity. This balance requires a reconstruction strategy that integrates tumor biology (e.g., perfusion pattern, degree of necrosis) and patient characteristics (e.g., body shapes, radiation-dose constraints). Prior studies have suggested that 50\u0026ndash;60 keV represents the ideal range for abdominal oncological imaging, offering the best compromise between contrast preservation and noise suppression [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. Emerging evidence, including the study by Lenga et al. [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], indicates that 40-keV VMI may offer superior diagnostic value for oncologic status and tumor response assessment. Our findings extend this paradigm by demonstrating that 40-keV VMI yields peak CNR and SNR, whereas 60-keV VMI delivers the best overall image quality. The most accurate lesion delineation was occurred at 40 keV and 50 keV, which is crucial for precise tumor-volume calculation and treatment planning. Furthermore, the combined 40\u0026ndash;70 keV approach achieved significantly better diagnostic accuracy than M_0.6 images. By optimizing sensitivity while preserving measurement reliability, this strategy establishes a reliable and reproducible framework for CRLM evaluation.\u003c/p\u003e \u003cp\u003eSub-centimeter liver metastases are of great importance in clinical diagnosis. Compared to larger lesions, smaller ones are more likely to be missed during pre-operative imaging and intra-operative inspection. However, once successfully detected, they are usually more suitable for radical surgical resection or local ablation [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. Missed detection might lead to serious implications. Ko et al. [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e] reported that false-negative pre-operative CT nodules were associated with a signigicantly higher 1-year recurrence rate after liver resection in patients with CRLM (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.048), indicating that undetected lesions may ultimately contribute to recurrence. The investigators thus concluded that improving the sensitivity could improve prognosis. Nevertheless, the sensitivity of CECT for the detection of sub-centimeter liver metastases is variable and limited, ranging from 22% to 68% [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. This variability primarily results from methodological differences across studies and insufficient statistical power in small lesion cohorts. Accordingly, the average sensitivity of CECT for metastases smaller than 10 mm remains modest. Our findings demonstrate that 40-keV VMI achieves the highest per-lesion detection rate for sub-centimeter liver metastases of CRC, whereas M_0.6 images yield the lowest (86.7% versus 60.0%). This 26% incremental gain highlights the clinical value of low-keV VMI in tumor staging, as it identifies metastatic foci that would otherwise escape CECT detection. By facilitating more precise surgical planning toward R0 resection, the reconstruction approach increases the likelihood of achieving NED status. However, the association between improved detection and long-term outcomes remains to be proven. While low-keV VMI improves diagnostic accuracy in detecting sub-centimeter metastases, prospective studies are required to verify its impact on surgical outcomes and recurrence rates.\u003c/p\u003e \u003cp\u003eAccurate pre-operative imaging and staging are critical for the management of CRLM, as they determine whether treatment should consist of surgical resection, loco-regional therapy, or systemic chemotherapy [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. Low-keV VMI enhances this process by amplifying SNR without additional contrast or radiation. This technique transforms CECT images into high-definition \u0026ldquo;digital biopsies\u0026rdquo; that enable precise staging and visualization of the sub-millimeter margin between tumor and portal radicle, a key factor for R0 resection [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. Furthermore, the same reconstruction algorithm provides RECIST-based [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e] tumor measurements with minimal intra-reader variability, thereby reducing inter-observer error and the risk of incorrectly assessing disease progression. Our findings demonstrate that integrating low-keV VMI reconstructions into clinical protocol for post-chemotherapy and surgical monitoring transforms patient management from subjective experience toward objective, reproducible parameters. This helps avoid futile regimens and may ultimately improve survival.\u003c/p\u003e \u003cp\u003eSeveral limitations should be acknowledged. First, the generalizability of our findings is limited by the single-center design, relatively small sample size, and absence of external validation. Second, despite a standardized ROI protocol, sampling variance was inevitable because of the variable shapes, necrotic cores, and irregular margins of liver metastases of CRC. Third, evaluation of low-keV VMI for sub-centimeter metastases was constrained by the small sample (n\u0026thinsp;=\u0026thinsp;15), which may diminish statistical power. Finally, results obtained on a single platform may not be fully generalizable to other DECT devices in different institutions.\u003c/p\u003e \u003cp\u003eIn conclusion, our study demonstrates that low-keV VMI reconstructions significantly improve diagnostic precision for detecting liver metastases of CRC while preserving measurement reliability for treatment-response evaluation. Notably, this technology delivers superior image quality and higher sensitivity for sub-centimeter metastases than CECT, without additional radiation or contrast. By minimizing missed diagnosis, low-keV VMI provides a reliable tool for initial screening. Integrating this reconstruction strategy into routine abdominal imaging protocols can improve staging accuracy, guide personalized surgical planning, and facilitate timely detection of recurrence, thereby offering a practical and cost-effective approach of improving patient outcomes.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eAUC, area under the curve (a statistical measure used to evaluate the performance of a diagnostic test, higher value indicates better diagnostic performance [value ranges from 0 to 1])\u003c/p\u003e\n\u003cp\u003eCRC, colorectal cancer (a malignant tumor arising from the epithelial lining of the colon or rectum)\u003c/p\u003e\n\u003cp\u003eCRLM, colorectal cancer with liver metastases (advanced-stage colorectal cancer characterized by the spread of malignant cells from the primary colorectal tumor to the liver)\u003c/p\u003e\n\u003cp\u003eCECT, contrast-enhanced computed tomography (a standard imaging technique that utilizes intravenous iodinated contrast material to improve visualization of vascular structures and enhance the detection and characterization of pathological conditions)\u003c/p\u003e\n\u003cp\u003eCNR, contrast-to-noise ratio (a quantitative metric in medical imaging that compares the difference in signal intensity between a region of interest and adjacent background tissue to the image noise)\u003c/p\u003e\n\u003cp\u003eDECT, dual-energy CT\u0026nbsp;(an advanced computed tomography technique that acquires data at two distinct X-ray energy spectra)\u003c/p\u003e\n\u003cp\u003eHU, hounsfield units (a standardized quantitative scale used in computed tomography to express the relative attenuation of X-rays by tissue)\u003c/p\u003e\n\u003cp\u003eICC, intraclass correlation coefficient (a statistical measure of reliability that quantifies the degree of agreement among multiple observers or repeated measurements)\u003c/p\u003e\n\u003cp\u003ekeV, kiloelectron volts (refers to the energy levels in virtual monoenergetic imaging)\u003c/p\u003e\n\u003cp\u003eMRI, magnetic resonance imaging (an imaging modality that utilizes strong magnetic fields and radiofrequency pulses to generate detailed anatomical and functional images without ionizing radiation)\u003c/p\u003e\n\u003cp\u003eNED, no-evidence-of-disease (a clinical status indicating complete remission where comprehensive imaging and clinical evaluation reveal no detectable residual malignancy following treatment)\u003c/p\u003e\n\u003cp\u003eROC, receiver operating characteristic (an analytical curve plotting the relationship between sensitivity and specificity across all possible thresholds of a diagnostic test)\u003c/p\u003e\n\u003cp\u003eROI, regions of interest (user-defined areas placed on medical images for quantitative analysis of tissue characteristics)\u003c/p\u003e\n\u003cp\u003eSNR, signal-to-noise (a fundamental image quality metric calculated as the ratio of signal intensity from a tissue region to the background noise)\u003c/p\u003e\n\u003cp\u003eSLD, sum of longest diameters (the arithmetic sum of the longest diameters of all target lesions identified at baseline)\u003c/p\u003e\n\u003cp\u003eSD, standard deviation (a measure of statistical dispersion representing the variation or spread of a set of values from their mean, commonly used to quantify image noise in radiological measurements)\u003c/p\u003e\n\u003cp\u003eVMI, virtual monoenergetic imaging (a post-processing technique derived from dual-energy CT data that reconstructs images as if they were acquired at a single, selectable X-ray energy level)\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eConflict of interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical approval and Informed consent\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis prospective study was conducted in accordance with the ethical standards and national research committee and with the principles of the Helsinki Declaration and its later amendments or comparable ethical standards. Approval was granted by the Biomedical Ethics Review Committee of West China Hospital of Sichuan University (2022-1904). And the study was registered with the Chinese Clinical Trial Registry (ChiCTR2500100120).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data generated or analyzed during this study are available on request from the corresponding author.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was funded by the National Key Research and Development Program of China\u0026mdash;Industrial Software Key Special Project [grant number 2024YFB3311700].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to express their sincere gratitude. We are deeply grateful to the Department of Radiology, West China Hospital of Sichuan University for their professional technical support and assistance in image analysis throughout this research. We also extend our thanks to all our colleagues in the Division\u0026nbsp;of Gastrointestinal Surgery, Department\u0026nbsp;of General Surgery, West China Hospital of Sichuan University for their valuable suggestions and insightful discussions during the course of this project. Finally, we acknowledge the anonymous reviewers and editors for their diligent work and constructive comments.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization: Lie Yang, Jiayu Sun; Methodology: Zhaoyang Zheng, Jianyang Yang; Formal analysis and investigation: Zhaoyang Zheng, Jianyang Yang; Writing - original draft preparation: Zhaoyang Zheng; Writing - review and editing: Lie Yang, Hong Zhu, Xi Feng; Supervision: Lie Yang.\u003c/p\u003e"},{"header":"References","content":"\u003col class=\"decimal_type\"\u003e\n \u003cli\u003eBray F, Laversanne M, Sung H, et al (2024) Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 74(3):229-263\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eChinese College of Surgeons, Chinese Society of Gastrointestinal Surgery, Chinese Society of Colorectal Surgery et al (2025) Chinese Guidelines for Diagnosis and Comprehensive Management of Liver Metastases from Colorectal Cancer (2025). Chinese Journal of Practical Surgery 45(8):841-854\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eReboux N, Jooste V, Goungounga J, Robaszkiewicz M, Nousbaum JB, Bouvier AM (2022) Incidence and Survival in Synchronous and Metachronous Liver Metastases From Colorectal Cancer. JAMA Netw Open 5(10):e2236666\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eStewart CL, Warner S, Ito K, et al (2018) Cytoreduction for colorectal metastases: liver, lung, peritoneum, lymph nodes, bone, brain. When does it palliate, prolong survival, and potentially cure?. Curr Probl Surg 55(9):330-379\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eAkg\u0026uuml;l \u0026Ouml;, \u0026Ccedil;etinkaya E, Ers\u0026ouml;z Ş, Tez M (2014) Role of surgery in colorectal cancer liver metastases. World J Gastroenterol 20(20):6113-6122\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eMargonis GA, Buettner S, Andreatos N, et al (2019) Prognostic Factors Change Over Time After Hepatectomy for Colorectal Liver Metastases: A Multi-institutional, International Analysis of 1099 Patients.\u0026nbsp;Ann Surg 269(6):1129-1137\u003c/li\u003e\n \u003cli\u003eMorris VK, Kennedy EB, Baxter NN, et al (2023) Treatment of Metastatic Colorectal Cancer: ASCO Guideline. J Clin Oncol 41(3):678-700\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eNiekel MC, Bipat S, Stoker J (2010) Diagnostic imaging of colorectal liver metastases with CT, MR imaging, FDG PET, and/or FDG PET/CT: a meta-analysis of prospective studies including patients who have not previously undergone treatment.\u0026nbsp;Radiology 257(3):674-684\u003c/li\u003e\n \u003cli\u003eNor\u0026eacute;n A, Eriksson HG, Olsson LI (2016) Selection for surgery and survival of synchronous colorectal liver metastases; a nationwide study. Eur J Cancer 53:105-114\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eKo Y, Kim J, Park JK, et al (2017) Limited detection of small (\u0026le; 10 mm) colorectal liver metastasis at preoperative CT in patients undergoing liver resection. PLoS One 12(12):e0189797\u003c/li\u003e\n \u003cli\u003eHaimerl M, W\u0026auml;chtler M, Platzek I, et al. (2013) Added value of Gd-EOB-DTPA-enhanced Hepatobiliary phase MR imaging in evaluation of focal solid hepatic lesions. BMC Med Imaging 13:41\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eGranata V, Fusco R, Maio F, et al. (2019) Qualitative assessment of EOB-GD-DTPA and Gd-BT-DO3A MR contrast studies in HCC patients and colorectal liver metastases. Infect Agent Cancer 14:40\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eChoi SH, Kim SY, Park SH, et al (2018) Diagnostic performance of CT, gadoxetate disodium-enhanced MRI, and PET/CT for the diagnosis of colorectal liver metastasis: Systematic review and meta-analysis.\u0026nbsp;J Magn Reson Imaging 47(5):1237-1250\u003c/li\u003e\n \u003cli\u003eMao Y, Chen B, Wang H, et al. (2020) performance of magnetic resonance imaging for colorectal liver metastasis: A systematic review and meta-analysis. Sci Rep 10(1):1969\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eRizzo S, Radice D, Femia M, et al. (2018) Metastatic and non-metastatic lymph nodes: quantification and different distribution of iodine uptake assessed by dual-energy CT. Eur Radiol 28(2):760-769\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eLiu H, Yan F, Pan Z, et al. (2015) Evaluation of dual energy spectral CT in differentiating metastatic from non-metastatic lymph nodes in rectal cancer: Initial experience.\u0026nbsp;Eur J Radiol 84(2):228-234\u003c/li\u003e\n \u003cli\u003eLenga L, Czwikla R, Wichmann JL, et al (2018) Dual-energy CT in patients with colorectal cancer: Improved assessment of hypoattenuating liver metastases using noise-optimized virtual monoenergetic imaging.\u0026nbsp;Eur J Radiol 106:184-191\u003c/li\u003e\n \u003cli\u003eParakh A, Lennartz S, An C, et al (2021) Dual-Energy CT Images: Pearls and Pitfalls.\u0026nbsp;Radiographics 41(1):98-119\u003c/li\u003e\n \u003cli\u003eHusarik DB, Gordic S, Desbiolles L, et al (2015) Advanced virtual monoenergetic computed tomography of hyperattenuating and hypoattenuating liver lesions: ex-vivo and patient experience in various body sizes.\u0026nbsp;Invest Radiol 50(10):695-702\u003c/li\u003e\n \u003cli\u003eLenga L, Lange M, Arendt CT, et al (2020) Measurement Reliability and Diagnostic Accuracy of Virtual Monoenergetic Dual-Energy CT in Patients with Colorectal Liver Metastases.\u0026nbsp;Acad Radiol 27(7):e168-e175\u003c/li\u003e\n \u003cli\u003eEisenhauer EA, Therasse P, Bogaerts J, et al (2009) New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1).\u0026nbsp;Eur J Cancer 45(2):228-247\u003c/li\u003e\n \u003cli\u003eLee T, Lee JM, Yoon JH, et al (2022) Deep learning-based image reconstruction of 40-keV virtual monoenergetic images of dual-energy CT for the assessment of hypoenhancing hepatic metastasis.\u0026nbsp;Eur Radiol 32(9):6407-6417\u003c/li\u003e\n \u003cli\u003eMartin SS, Wichmann JL, Weyer H, et al (2017) Dual-energy computed tomography in patients with cutaneous malignant melanoma: Comparison of noise-optimized and traditional virtual monoenergetic imaging.\u0026nbsp;Eur J Radiol 95:1-8\u003c/li\u003e\n \u003cli\u003eMartin SS, Wichmann JL, Weyer H, et al (2017) Dual-energy computed tomography in patients with cutaneous malignant melanoma: Comparison of noise-optimized and traditional virtual monoenergetic imaging.\u0026nbsp;Eur J Radiol 95:1-8\u003c/li\u003e\n \u003cli\u003eSudarski S, Apfaltrer P, Nance JW Jr, et al (2014) Objective and subjective image quality of liver parenchyma and hepatic metastases with virtual monoenergetic dual-source dual-energy CT reconstructions: an analysis in patients with gastrointestinal stromal tumor.\u0026nbsp;Acad Radiol 21(4):514-522\u003c/li\u003e\n \u003cli\u003eWichmann JL, Gillott MR, De Cecco CN, et al (2016) Dual-Energy Computed Tomography Angiography of the Lower Extremity Runoff: Impact of Noise-Optimized Virtual Monochromatic Imaging on Image Quality and Diagnostic Accuracy.\u0026nbsp;Invest Radiol 51(2):139-146\u003c/li\u003e\n \u003cli\u003eDe Cecco CN, Caruso D, Schoepf UJ, et al (2016) Optimization of window settings for virtual monoenergetic imaging in dual-energy CT of the liver: A multi-reader evaluation of standard monoenergetic and advanced imaged-based monoenergetic datasets.\u0026nbsp;Eur J Radiol 85(4):695-699\u003c/li\u003e\n \u003cli\u003eBlyth S, Blakeborough A, Peterson M, Cameron IC, Majeed AW (2008) Sensitivity of magnetic resonance imaging in the detection of colorectal liver metastases. Ann R Coll Surg Engl 90(1):25-28\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eMaino C, Vernuccio F, Cannella R, et al (2023) Liver metastases: The role of magnetic resonance imaging. World J Gastroenterol 29(36):5180-5197\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eYu MH, Lee JM, Hur BY, et al (2015) Gadoxetic acid-enhanced MRI and diffusion-weighted imaging for the detection of colorectal liver metastases after neoadjuvant chemotherapy.\u0026nbsp;Eur Radiol 25(8):2428-2436\u003c/li\u003e\n \u003cli\u003eD.V. Cicchetti (1994) Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology, Psychol. Assess 6 (4) 284\u003c/li\u003e\n \u003cli\u003ePeijie Lv, Nana Liu, Luotong Wang, et al (2023) Deep learning image reconstruction for optimizing image quality of low-energy spectral monochromatic CT and detecting liver small low-contrast lesions. Chinese Journal of Medical Imaging Technology 39(1):104-108\u003c/li\u003e\n \u003cli\u003eGreffier J, Si-Mohamed S, Guiu B, et al (2022) Comparison of virtual monoenergetic imaging between a rapid kilovoltage switching dual-energy computed tomography with deep-learning and four dual-energy CTs with iterative reconstruction.\u0026nbsp;Quant Imaging Med Surg 12(2):1149-1162\u003c/li\u003e\n \u003cli\u003eLyu P, Liu N, Harrawood B, et al (2023) Is it possible to use low-dose deep learning reconstruction for the detection of liver metastases on CT routinely?.\u0026nbsp;Eur Radiol 33(3):1629-1640\u003c/li\u003e\n \u003cli\u003eMcCollough CH, Leng S, Yu L, Fletcher JG (2015) Dual- and Multi-Energy CT: Principles, Technical Approaches, and Clinical Applications.\u0026nbsp;Radiology 276(3):637-653\u003c/li\u003e\n \u003cli\u003eSun X, Niwa T, Nomura T, Yoshida R, Koyanagi K, Hashimoto J (2025) Low-keV Virtual Monoenergetic Imaging for Bronchial Artery Visualization on Photon-Counting Detector Computed Tomography.\u0026nbsp;Diagnostics (Basel) 15(11):1354\u003c/li\u003e\n \u003cli\u003eGreffier J, Faby S, Pastor M, et al (2024) Comparison of low-energy virtual monoenergetic images between photon-counting CT and energy-integrating detectors CT: A phantom study.\u0026nbsp;Diagn Interv Imaging 105(9):311-318\u003c/li\u003e\n \u003cli\u003eDunning CAS, Rajendran K, Inoue A, et al (2023) Optimal Virtual Monoenergetic Photon Energy (keV) for Photon-Counting-Detector Computed Tomography Angiography.\u0026nbsp;J Comput Assist Tomogr 47(4):569-575\u003c/li\u003e\n \u003cli\u003eZhou W, Michalak GJ, Weaver JM, et al (2020) A Universal Protocol for Abdominal CT Examinations Performed on a Photon-Counting Detector CT System: A Feasibility Study.\u0026nbsp;Invest Radiol 55(4):226-232\u003c/li\u003e\n \u003cli\u003eRajeev R, Hussain HK (2025) Thermal Ablation versus Surgical Resection for Management of Small Colorectal Liver Metastases: Insights from the COLLISION Trial.\u0026nbsp;Radiol Imaging Cancer 7(2):e259008\u003c/li\u003e\n \u003cli\u003eAkhan O, Ak\u0026ccedil;alar S, \u0026Uuml;nal E, Metin Y, \u0026Ccedil;ift\u0026ccedil;i T, Akıncı D (2023) Radiofrequency Ablation for Colorectal Cancer Liver Metastases: Outcomes and Prognostic Factors Associated with Survival.\u0026nbsp;Turk J Gastroenterol 34(6):645-651\u003c/li\u003e\n \u003cli\u003eCharnsangavej C, Clary B, Fong Y, Grothey A, Pawlik TM, Choti MA (2006) Selection of patients for resection of hepatic colorectal metastases: expert consensus statement.\u0026nbsp;Ann Surg Oncol 13(10):1261-1268\u003c/li\u003e\n \u003cli\u003eDempsey PJ, Farrelly C, Cronin CG, Fenlon HM (2024) Preoperative imaging of colorectal liver metastases: what the radiologist and the multidisciplinary team need to know.\u0026nbsp;Br J Radiol 97(1162):1602-1618\u003c/li\u003e\n \u003cli\u003eHamady ZZ, Lodge JP, Welsh FK, et al (2014) One-millimeter cancer-free margin is curative for colorectal liver metastases: a propensity score case-match approach.\u0026nbsp;Ann Surg 259(3):543-548\u003c/li\u003e\n \u003cli\u003eSchwartz LH, Liti\u0026egrave;re S, de Vries E, et al (2016) RECIST 1.1-Update and clarification: From the RECIST committee. Eur J Cancer 62:132-137\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Colorectal liver metastases, Dual-energy CT, Virtual monoenergetic imaging, Diagnostic accuracy, Measurement reliability","lastPublishedDoi":"10.21203/rs.3.rs-8359303/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8359303/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eObjective: \u003c/strong\u003eTo assess the value of low-keV virtual monoenergetic imaging (VMI) from dual-energy CT (DECT) for detecting liver metastases of colorectal cancer (CRC) and for evaluating treatment response.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMaterials and Methods:\u003c/strong\u003e Patients diagnosed with CRC who had liver metastases (CRLM) were prospectively enrolled. All underwent DECT, followed by magnetic resonance imaging (MRI) used as the gold standard. DECT datasets were reconstructed as standard linearly-blended (M_0.6) images to simulate contrast-enhanced CT (CECT) and as VMI at 10-keV intervals (40–70 keV). Signal-to-noise (SNR) and contrast-to-noise (CNR) ratios were measured. Radiologists independently assessed image quality, lesion delineation, and image noise using a 5-point Likert scale. Per-lesion sensitivity, specificity, and detection rates were calculated.Treatment response was evaluated using RECIST 1.1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003eThirty-five patients (125 liver metastases) were enrolled. 40-keV VMI provided the best CNR (6.2 ± 4.3, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01) and SNR (15.5 ± 9.5, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01), with AUCs of 0.974 for SNR and 0.966 for attenuation. It outperformed M_0.6 images for detecting lesions ≤10 mm (86.7% vs. 60.0%, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05). Image quality of 60-keV VMI was equivalent to M_0.6 (median 5, \u003cem\u003ep\u003c/em\u003e \u0026gt; 0.05). Lesion delineation was optimal with 40-keV and 50-keV VMI (median 5). Image noise was lowest with 60-keV VMI (median 4, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01). Lesion size reduction was comparable between MRI and DECT (5.2 ± 5.5 mm vs. 4.6 ± 6.2 mm, \u003cem\u003ep\u003c/em\u003e = 0.41).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion: \u003c/strong\u003eLow-keV VMI improves the diagnostic accuracy of liver metastases of CRC compared to CECT, while maintaining measurement reliability in treatment-response assessment.\u003c/p\u003e","manuscriptTitle":"Assessment of Diagnostic Accuracy and Measurement Reliability of Low-keV Virtual Monoenergetic Dual-energy CT in The Liver Metastases of Colorectal Cancer: A Prospective, Imaging-Reference Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-22 10:14:21","doi":"10.21203/rs.3.rs-8359303/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f90ab543-803f-4b5d-ab40-7218dd40f2ec","owner":[],"postedDate":"December 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-29T20:38:27+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-22 10:14:21","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8359303","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8359303","identity":"rs-8359303","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}