Tomography最新文献

{"title":"Repeatability of Semi-Quantitative and Volumetric Features from Artificial-Intelligence-Guided Lesion Segmentation on ¹⁸F-DCFPyL PSMA-PET/CT Images: Results from a Test-Retest Cohort.","authors":"Md Zobaer Islam, Timothy G Perk, Amy Weisman, Mark C Markowski, Kenneth J Pienta, Young E Whang, Matthew I Milowsky, Martin G Pomper, Nicholas Wisniewski, Ralph A Bundschuh, Rudolf A Werner, Michael A Gorin, Steven P Rowe","doi":"10.3390/tomography12030038","DOIUrl":"10.3390/tomography12030038","url":null,"abstract":"<p><p><b>Objectives:</b> This study evaluated the test-retest repeatability of semi-quantitative and volumetric features derived from artificial intelligence (AI)-assisted lesion segmentation on ¹⁸F-DCFPyL Prostate Specific Membrane Antigen (PSMA)-PET/CT imaging of patients with prostate cancer (PCa). Specifically, we assessed the reliability of maximum, minimum and total standardized uptake values (SUV_max, SUV_mean, SUV_total) and lesion volume measurements across varying lesion sizes and explored the implications of variability for clinical decision-making. <b>Methods:</b> We analyzed ¹⁸F-DCFPyL PSMA-PET/CT images from 22 patients with metastatic PCa. Lesion segmentation was performed using the AI-guided TRAQinform IQ technology, followed by a manual review to eliminate potential false-positive sites of uptake. Lesion-level test-retest repeatability was evaluated using 95% limits of agreement (LOA), intra-class correlation coefficient (ICC), within-subject coefficient of variation (wCOV) and Bland-Altman analysis for SUV and volumetric parameters. Lesions were stratified by size (>1 cm³ and >1.5 cm³) to assess the impact of lesion volume cut-offs on measurement variability. <b>Results:</b> A total of 297 lesions were analyzed, including 191 lesions > 1 cm³ and 161 lesions > 1.5 cm³. Test-retest variability was higher in smaller lesions, with narrower LOA and lower wCOV for larger lesions. SUV_max and SUV_mean exhibited lower variability than SUV_total and lesion volume. The 95% LOA for SUV_max ranged from -33.81% to +38.02% for all lesions, improving to -31.82% to +31.01% for lesions > 1.5 cm³. Similar trends were observed for SUV_mean, SUV_total, and volume. Bland-Altman plots confirmed reduced variability in larger lesions, with no significant systematic bias. <b>Conclusions:</b> The test-retest repeatability of AI-assisted PSMA-PET/CT features varies by feature type, with semi-quantitative features demonstrating improved repeatability relative to volumetric features. Additionally, repeatability is influenced by lesion size, with larger lesions exhibiting greater reliability. These findings highlight the importance of lesion size-dependent thresholds in response assessment and variability-aware feature selection in prognostic models. Current algorithms may be better optimized for larger lesions and higher volumes of disease, with limitations remaining in the robust detection and segmentation of smaller/more subtle lesions.</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13030691/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cerebral Accumulation of Gadolinium (Gd³⁺) and Related Cellular Stress Pathways in Rat Brain Tissue.","authors":"Göksel Tuzcu, Burak Çildağ, Songül Çildağ, Çiğdem Yenisey, Zahir Kızılay","doi":"10.3390/tomography12030037","DOIUrl":"10.3390/tomography12030037","url":null,"abstract":"<p><strong>Background/objectives: </strong>This study aimed to compare in vivo cerebral gadolinium (Gd³⁺) accumulation, associated unfolded protein response (UPR), and oxidative stress parameters in rats after exposure to gadolinium-based contrast agents (GBCAs).</p><p><strong>Methods: </strong>This study was designed as a controlled, experimental animal study to evaluate the accumulation of Gd³⁺ in the basal ganglia of rats following the administration of 0.6 mmol/kg gadopentetate dimeglumine (linear) and gadoterate meglumine (macrocyclic). Male Sprague-Dawley rats were exposed to the contrast agents for 24 and 72 h, and then the basal ganglia tissues were collected postmortem. The tissue levels of Gd³⁺ accumulation, activating transcription factor-6 (ATF6), inositol-requiring enzyme-1 (IRE-1), protein kinase RNA-like endoplasmic reticulum kinase (PERK), damage-inducible transcript-3 (DDIT3), total antioxidant status (TAS), and total oxidant status (TOS) were determined.</p><p><strong>Results: </strong>Linear GBCA-treated rats had persistent Gd³⁺ levels over time, whereas a significant reduction from 24 to 72 h was observed in macrocyclic GBCA-treated rats (<i>p</i> < 0.001). PERK, DDIT3, and ATF6 expressions were significantly elevated after linear GBCA exposure (<i>p</i> < 0.05), while no significant increase was observed in the macrocyclic GBCA-treated group. However, IRE-1, TAS, and TOS levels were not significantly different in either group.</p><p><strong>Conclusions: </strong>Linear and macrocyclic GBCAs demonstrated distinct patterns of cerebral Gd³⁺ accumulation and UPR levels in rats. Accordingly, GBCA administration should be reserved for instances where it is necessary, such as when contrast enhancement is clinically required.</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13030676/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534232","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of Radiation Dose and Image Quality in the Transition from Conventional Pelvimetry to Low-Dose Helical CT Pelvimetry.","authors":"K Shahgeldi, M Parenmark, L Claesson, T M Svahn","doi":"10.3390/tomography12030035","DOIUrl":"10.3390/tomography12030035","url":null,"abstract":"<p><strong>Purpose: </strong>The present study aimed to assess the radiation dose associated with low-dose (LD) CT pelvimetry compared with conventional radiography and to evaluate the adequacy of the resulting image quality.</p><p><strong>Methods: </strong>The absorbed dose was measured using thermoluminescent dosimeters positioned in an anthropomorphic female phantom, including uterine locations, to estimate the fetal dose. Conventional radiographic pelvimetry and LD-CT pelvimetry were performed using clinically implemented protocols. Effective dose was calculated using Monte Carlo-based modeling applying acquisition parameters and retrospective patient dose registry data. Image quality of LD-CT pelvimetry was independently evaluated in 14 consecutive clinical cases using a four-point ordinal scale.</p><p><strong>Results: </strong>LD-CT pelvimetry reduced the mean absorbed pelvic dose by approximately 50% compared with conventional pelvimetry (0.18 vs. 0.39 mGy) and decreased estimated fetal dose by 40% (0.21 vs. 0.37 mGy). These estimates were based on standardized single acquisitions and did not incorporate additional radiation from retakes commonly observed in conventional practice. CT demonstrated substantially more homogeneous dose distribution, whereas conventional pelvimetry exhibited marked heterogeneity with peak values up to 2.3 mGy. The maternal effective dose was lower for LD-CT (0.16 mSv) than for conventional pelvimetry (0.36 mSv); inclusion of retakes increased the conventional effective dose to 0.71 mSv. All CT examinations were diagnostically adequate, and no recalls were required.</p><p><strong>Conclusions: </strong>Optimized low-dose CT pelvimetry significantly reduces radiation dose compared with conventional radiographic pelvimetry while maintaining reliable diagnostic image quality. These results support the clinical adoption of CT-based pelvimetry as a dose-efficient and reproducible alternative to conventional techniques.</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13030744/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534168","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Eye Lens Radiation Exposure During TAVI: Current Evidence and Imaging-Based Strategies for Dose Reduction.","authors":"Chiara Zanon, Alessandro Fiocco, Vincenzo Tarzia, Emilio Quaia","doi":"10.3390/tomography12030036","DOIUrl":"10.3390/tomography12030036","url":null,"abstract":"<p><strong>Background: </strong>Transcatheter aortic valve implantation (TAVI) is increasingly performed in fluoroscopy-intensive environments, raising concerns about occupational eye lens dose (equivalent dose to the eye lens, Hp (3)) and the risk of radiation-induced cataract, particularly after the reduction of recommended annual eye lens dose limits to 20 mSv.</p><p><strong>Purpose: </strong>To summarize evidence on eye lens radiation exposure during TAVI, identify procedural and occupational determinants, and review strategies to reduce exposure with a focus on imaging optimization.</p><p><strong>Methods: </strong>We performed a narrative review of observational and prospective studies reporting direct eye-level dose measurements or validated surrogate eye lens dose estimates (head-level, chest-level, or DAP-normalized) during TAVI and related structural heart procedures. This approach was chosen to provide a qualitative synthesis of the available evidence rather than a formal systematic review.</p><p><strong>Results: </strong>Reported operator eye lens doses typically ranged from 30 to 110 µSv per procedure, with higher exposure during transapical/transaortal access and among staff working close to the patient (e.g., anesthesiologists and circulating nurses). Additional shielding and lead-free drapes reduced normalized eye dose by approximately 25-40%, and RADPAD^® use reduced operator eye-level dose from 24.3 to 14.8 µSv per procedure (<i>p</i> = 0.008). At these levels, cumulative exposure may approach recommended regulatory limits after approximately 150-300 procedures, depending on role, access route, and shielding practices.</p><p><strong>Conclusion: </strong>In conclusion, Occupational eye lens exposure during TAVI is clinically relevant and strongly influenced by access route, staff positioning, and imaging-system use. Dose reduction should combine routine eye protection and dedicated eye-level dosimetry with imaging optimization (low pulse-rate fluoroscopy, minimized Digital-Subtraction-Angiography (DSA)/cine acquisitions, tight collimation, avoidance of unnecessary magnification, and correct positioning of ceiling-suspended shields and table skirts).</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13030497/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534177","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Semi-Supervised Vertebra Segmentation and Identification in CT Images.","authors":"You Fu, Jiasen Feng, Hanlin Cheng","doi":"10.3390/tomography12030033","DOIUrl":"10.3390/tomography12030033","url":null,"abstract":"<p><p><b>Background/Objectives</b>: Automatic segmentation and identification of vertebrae in spinal CT are essential for assisting diagnosis of spinal disorders and for preoperative planning. The task is challenging due to the high structural similarity between adjacent vertebrae and the morphological variability of vertebrae. Most existing methods rely on fully supervised deep learning and, constrained by limited annotations, struggle to remain robust in complex scenarios. <b>Methods</b>: We propose a semi-supervised approach built on a dual-branch 3D U-Net. Mamba modules are inserted between the encoder and decoder to model long-range dependencies along the cranio-caudal axis. The identification branch employs a 3D convolutional block attention module (3D-CBAM) to enhance class discriminability. A unified semi-supervised objective is formulated via teacher-student consistency: for each unlabeled sample, weakly and strongly augmented views are generated, and cross-branch consistency is enforced, together with confidence-based filtering and class-frequency reweighting. In addition, a connected-component analysis is used to enforce anatomically plausible sequential continuity of vertebral indices in the outputs. <b>Results</b>: Experiments on VerSe 2019 and 2020 show that, on the public VerSe 2019 test set (with VerSe 2020 scans used as unlabeled training data), the supervised baseline achieved a Dice score of 89.8% and an identification accuracy of 92.3%. Incorporating unlabeled data improved performance to 91.6% Dice and 97.5% identification accuracy (relative gains of +1.8 and +5.2 percentage points). Compared with competing methods, the proposed semi-supervised model attains higher or comparable segmentation accuracy and the highest identification accuracy. <b>Conclusions</b>: Without additional annotation cost, the proposed method markedly improves the overall performance of vertebra segmentation and identification, offering more robust automated support for clinical workflows.</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13030516/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Direct Segmentation of Mammography and Tomosynthesis Sinograms for Lesion Localization.","authors":"Estefanía Ruíz Muñoz, Leopoldo Altamirano Robles, Raquel Díaz Hernández, Kelsey Alejandra Ramírez Gutiérrez, Saúl Zapotecas-Martínez, José de Jesús Velázquez Arreola","doi":"10.3390/tomography12030034","DOIUrl":"10.3390/tomography12030034","url":null,"abstract":"<p><p><b>Background</b>: The Detection and localization of breast lesions remain challenging in mammography and digital breast tomosynthesis (DBT) due to tissue overlap and information loss during volumetric reconstruction. Sinograms preserve the full angular projection data acquired during scanning, enabling analysis of tissue structure without reconstruction. <b>Methods</b>: This study proposes a direct segmentation approach for mammography and DBT sinograms using a U-Net architecture. Experiments were conducted on 1082 annotated mammography mass images from the CBIS-DDSM dataset (521 benign, 561 malignant) and 272 annotated DBT images from the Breast Cancer Screening DBT dataset (136 benign, 136 malignant). Dataset splitting was performed at the patient level to prevent data leakage, and all reported quantitative results correspond to the independent test set, with the validation set used solely for model selection and early stopping. Three input configurations were evaluated: mammography sinograms, DBT sinograms, and a combined model. <b>Results</b>: The mammography model achieved the highest performance (Dice: 0.94 training, 0.90 test), outperforming DBT alone (0.77 training, 0.70 test). Multimodal fusion improved DBT results (Dice: 0.84 test). Centroid analysis showed 99.11% correspondence with reference annotations (average distance: 2.83 pixels), and partial back-projection reconstructions confirmed anatomical consistency. Compared with YOLOv5x, the proposed approach provided superior lesion localization, particularly for small or multiple lesions. <b>Conclusions</b>: Direct sinogram segmentation is an efficient, clinically viable strategy for breast lesion detection and localization.</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13030243/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534263","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Diagnostic Test Accuracy and Semi-Quantitative Metrics of ¹⁸F-FDG PET in Assessing Treatment Response in Skull Base Osteomyelitis and Necrotising Otitis Externa: A Systematic Review and Meta-Analysis.","authors":"Mark Laidlaw, Maya Reid, Sukanya Rajiv, Jean-Marc Gerard","doi":"10.3390/tomography12030032","DOIUrl":"10.3390/tomography12030032","url":null,"abstract":"<p><strong>Background/objectives: </strong>Skull base osteomyelitis and necrotising otitis externa require prolonged antibiotic therapy, yet determining optimal treatment cessation timing remains challenging. Conventional imaging modalities demonstrate persistent abnormalities beyond infection resolution, confounding treatment decisions. This systematic review evaluated the diagnostic test accuracy of ¹⁸F-fluorodeoxyglucose positron emission tomography (¹⁸F-FDG PET) for treatment response monitoring in skull base osteomyelitis and necrotising otitis externa.</p><p><strong>Methods: </strong>We conducted a systematic review following PRISMA-DTA guidelines, searching MEDLINE, Embase, CENTRAL, CINAHL, Scopus, and Web of Science from inception to November 2025. Studies evaluating ¹⁸F-FDG PET diagnostic accuracy for treatment response assessment in confirmed skull base osteomyelitis or necrotising otitis externa were included. Two reviewers independently screened studies, extracted data, and assessed risk of bias using QUADAS-2. Bivariate random-effects meta-analysis was performed using MetaBayesDTA to obtain pooled sensitivity and specificity.</p><p><strong>Results: </strong>Eight studies comprising 154 lesions contributed to the primary analysis. Pooled sensitivity was 95.2% (95% credible interval 85.6-99.0%) and pooled specificity was 89.1% (95% credible interval 70.7-96.7%). The positive likelihood ratio was 8.7 (95% credible interval 3.2-28.4) and negative likelihood ratio was 0.05 (95% credible interval 0.01-0.17), with a diagnostic odds ratio of 172.0. Seven studies evaluating detection rate at initial presentation yielded a pooled rate of 96.1% (95% confidence interval 91.3-98.3%). SUVmax was the most frequently used metabolic parameter.</p><p><strong>Conclusions: </strong>¹⁸F-FDG PET, specifically using SUVmax, demonstrates high sensitivity and good specificity for treatment response monitoring, with excellent capacity to rule out persistent infection. However, evidence quality is limited by retrospective designs and substantial heterogeneity. Prospective studies with standardised thresholds are needed to validate clinical utility.</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-03-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13030614/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534192","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Lateralization of FDG-PET Hypometabolism Using Resting-State fMRI in Temporal Lobe Epilepsy: A Simultaneous PET-MRI Study.","authors":"Daniel Uher, Gerhard S Drenthen, Tineke van de Weijer, Jochem van der Pol, Christianne M Hoeberigs, Paul A M Hofman, Sam Springer, Rob P W Rouhl, Albert J Colon, Olaf E M G Schijns, Walter H Backes, Jacobus F A Jansen","doi":"10.3390/tomography12030030","DOIUrl":"10.3390/tomography12030030","url":null,"abstract":"<p><strong>Background: </strong>In temporal lobe epilepsy (TLE), locally reduced glucose metabolism (i.e., hypometabolism) is indicative of the epileptogenic onset zone (EZ). Here, we investigate the potential value of resting-state fMRI (rs-fMRI) for localizing the EZ with fluorodeoxyglucose positron emission tomography (FDG-PET) as ground truth.</p><p><strong>Methods: </strong>Twelve PET-positive patients (34.1 ± 13.1 y; 5 females) with unilateral drug-resistant TLE were included. FDG-PET and rs-fMRI were acquired simultaneously at a hybrid 3T PET-MR scanner. Hypometabolic regions were identified on the FDG-PET images by a nuclear medicine expert. The FDG-PET images were compared with a clinical FDG-PET control dataset with normal glucose uptake distribution. The output z-score maps were thresholded at z < -2 to produce a binary mask of the significantly hypometabolic regions. The hypometabolism masks were mirrored onto the contralateral hemisphere for the asymmetry comparison. Regional homogeneity (ReHo), amplitude of low-frequency fluctuations (ALFF), and fractional ALFF (fALFF) were calculated from the rs-fMRI in conventional (0.01-0.1 Hz) and slow-3 (0.073-0.198 Hz) frequency bands. Asymmetry indices (AIs) were calculated using the ipsilateral and contralateral hypometabolic masks in the PET-positive subjects and assessed via the one-sample Wilcoxon test and Spearman correlation coefficients.</p><p><strong>Results: </strong>The AIs of conventional fALFF were significantly lower in the hypometabolic zone (<i>p</i> < 0.05). A significant negative correlation was found between the AIs of FDG-PET and fALFF in the slow-3 band (r = -0.62; <i>p</i> < 0.05).</p><p><strong>Conclusions: </strong>Conventional and slow-3 band fALFF showed a potential to mimic the FDG-PET findings in terms of EZ localization. Further research with extended cohorts and histopathological validation is required to determine the clinical value.</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-03-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13029870/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534307","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"How to Deal with Paper Rejection.","authors":"Emilio Quaia","doi":"10.3390/tomography12030031","DOIUrl":"10.3390/tomography12030031","url":null,"abstract":"<p><p>This editorial provides insights into the common situation of paper rejection, which must be managed by the authors [...].</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-03-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13030062/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534200","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated Multi-Modal MRI Segmentation of Stroke Lesions and Corticospinal Tract Integrity for Functional Outcome Prediction.","authors":"Daniyal Iqbal, Domenec Puig, Muhammad Mursil, Hatem A Rashwan","doi":"10.3390/tomography12030029","DOIUrl":"10.3390/tomography12030029","url":null,"abstract":"<p><p><b>Background/Objectives:</b> Stroke is a leading cause of long-term disability, and predicting functional outcome at discharge, such as the modified Rankin Scale (mRS), is important for guiding treatment and rehabilitation. Many existing approaches depend on advanced imaging or complex corticospinal tract (CST) segmentation from multi-shell diffusion MRI, limiting clinical feasibility. Automated lesion segmentation is also challenging due to lesion heterogeneity and MRI variability. This study proposes a clinically feasible multimodal MRI pipeline based on routine imaging. <b>Methods:</b> Lesion segmentation models were trained and evaluated on the ISLES 2022 dataset (250 training, 150 test cases). Zero-shot external evaluation was performed on 149 cases from ISLES 2024 using standard MRI sequences only. An ensemble of deep learning models (SEALS, NVAUTO, FACTORIZER) was evaluated on ISLES 2022, while SEALS alone was used for external testing. CST segmentation was performed using TractSeg on single-shell diffusion-weighted imaging. Imaging biomarkers included lesion volume, shape, ADC-based texture features, CST integrity, and lesion-CST overlap. These features were used to train machine learning models for binary mRS prediction at discharge. <b>Results:</b> The ensemble achieved a Dice score of 0.82 on ISLES 2022, while zero-shot evaluation on ISLES 2024 achieved 0.57. In exploratory analysis, CatBoost achieved the highest point estimates (accuracy 0.88, F1-score 0.87, ROC-AUC 0.83). Key predictors included lesion-CST overlap, lesion volume, surface area, dissimilarity, and contrast. <b>Conclusions:</b> This exploratory study demonstrates the feasibility of combining automated lesion segmentation with anatomically informed biomarkers using routine clinical MRI, supporting interpretable stroke outcome modelling and motivating future large-scale validation.</p>","PeriodicalId":51330,"journal":{"name":"Tomography","volume":"12 3","pages":""},"PeriodicalIF":2.2,"publicationDate":"2026-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13030278/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147534014","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}