Jose Carlos Pariente
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jose Carlos Pariente has a total of 872 co-authors. Here are the top co-authors by number of shared publications:
- Núria Bargalló Alabart
Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS) Barcelona Spain
8 shared publications - Gisela Sugranyes
Child and Adolescent Psychiatry and Psychology Department, Hospital Clínic de Barcelona, Barcelona, Spain
6 shared publications - Daniel Joseph Stein
University of Cape Town
5 shared publications - Jun Soo Kwon
Department of Brain and Cognitive Sciences, Seoul National University College of Natural Sciences, Seoul, South Korea
5 shared publications - Minah Kim
Department of Neuropsychiatry, Seoul National University Hospital, Seoul, Korea
5 shared publications Neda Jahanshad
Imaging Genetics Center, Mark and Mary Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Marina del Rey, Los Angeles, USA
5 shared publications
Latest Publications
- Open Access
Structural Magnetic Resonance Imaging (MRI) in Differentiating Radionecrosis from Metastatic Recurrence in Post-Surgical Patients: A Tertiary Center Experience
Purpose: Distinguishing radionecrosis (RN) from recurrent brain metastasis (rM1) remains a major challenge in the follow-up of patients treated with surgery and radiotherapy. Although advanced imaging techniques have been proposed to improve diagnostic accuracy, they are not always available in routine clinical practice and may be limited by postoperative changes and artifacts. Histopathologically validated studies in this setting remain scarce. Methods: We retrospectively evaluated 16 patients with 17 histologically confirmed lesions who underwent repeat surgery for suspected progression after the resection of brain metastases and adjuvant radiotherapy. Preoperative MRI examinations were reviewed by two neuroradiologists using a standard MRI protocol. Qualitative features included signal characteristics, T1/T2 mismatch, enhancement patterns, and the presence of the “inflammatory cloud” and “incomplete ring enhancement” signs. Quantitative analysis included lesion, edema and necrosis volumes, apparent diffusion coefficient (ADC), intratumoral susceptibility signal (ITSS) grading, and volumetric lesion quotient (vLQ). A LASSO logistic regression model with leave-one-out cross-validation and bootstrap stability analysis was used to identify the most relevant predictors. Results: Histopathology demonstrated 11 RN lesions and six recurrent metastases. The LASSO model achieved an area under the curve of 0.71 and an overall accuracy of 76% (13/17 lesions) with a sensitivity of 82% for RN and a specificity of 67% for rM1. RN lesions were characterized by significantly higher ADC values and ITSS grades and by the exclusive presence of the inflammatory cloud and incomplete ring enhancement signs. Stability analysis identified incomplete ring enhancement (49%), ITSS grade (45%), ADC values (45%), and inflammatory cloud (44%) as the most robust predictors of RN, whereas volumetric measurements showed limited independent predictive value. Conclusions: In patients with histologically confirmed post-treatment lesions, conventional MRI provided several reproducible features that aided differentiation between radionecrosis and recurrent brain metastasis. The inflammatory cloud, incomplete ring enhancement, elevated ADC values, and higher ITSS grades emerged as the most robust imaging markers of RN. While predictive multivariable models remain exploratory, these readily available MRI features may improve diagnostic confidence in routine clinical practice and warrant validation in larger multicenter cohorts.
Open AccessStructural Magnetic Resonance Imaging (MRI) in Differentiating Radionecrosis from Metastatic Recurrence in Post-Surgical Patients: A Tertiary Center Experience
Purpose: Distinguishing radionecrosis (RN) from recurrent brain metastasis (rM1) remains a major challenge in the follow-up of patients treated with surgery and radiotherapy. Although advanced imaging techniques have been proposed to improve diagnostic accuracy, they are not always available in routine clinical practice and may be limited by postoperative changes and artifacts. Histopathologically validated studies in this setting remain scarce. Methods: We retrospectively evaluated 16 patients with 17 histologically confirmed lesions who underwent repeat surgery for suspected progression after the resection of brain metastases and adjuvant radiotherapy. Preoperative MRI examinations were reviewed by two neuroradiologists using a standard MRI protocol. Qualitative features included signal characteristics, T1/T2 mismatch, enhancement patterns, and the presence of the “inflammatory cloud” and “incomplete ring enhancement” signs. Quantitative analysis included lesion, edema and necrosis volumes, apparent diffusion coefficient (ADC), intratumoral susceptibility signal (ITSS) grading, and volumetric lesion quotient (vLQ). A LASSO logistic regression model with leave-one-out cross-validation and bootstrap stability analysis was used to identify the most relevant predictors. Results: Histopathology demonstrated 11 RN lesions and six recurrent metastases. The LASSO model achieved an area under the curve of 0.71 and an overall accuracy of 76% (13/17 lesions) with a sensitivity of 82% for RN and a specificity of 67% for rM1. RN lesions were characterized by significantly higher ADC values and ITSS grades and by the exclusive presence of the inflammatory cloud and incomplete ring enhancement signs. Stability analysis identified incomplete ring enhancement (49%), ITSS grade (45%), ADC values (45%), and inflammatory cloud (44%) as the most robust predictors of RN, whereas volumetric measurements showed limited independent predictive value. Conclusions: In patients with histologically confirmed post-treatment lesions, conventional MRI provided several reproducible features that aided differentiation between radionecrosis and recurrent brain metastasis. The inflammatory cloud, incomplete ring enhancement, elevated ADC values, and higher ITSS grades emerged as the most robust imaging markers of RN. While predictive multivariable models remain exploratory, these readily available MRI features may improve diagnostic confidence in routine clinical practice and warrant validation in larger multicenter cohorts.
- Open Access
Automated segmentation of postsurgical resection cavities on magnetic resonance imaging in focal epilepsy: A Multicentre Epilepsy Lesion Detection study
Objective: Quantitative assessment of extent of tissue resection following epilepsy surgery requires accurate delineation of the resection cavity on postoperative magnetic resonance imaging (MRI). Current methods for resection cavity masking are time‐consuming and labor‐intensive, and existing automated approaches exhibit variable segmentation accuracy, particularly on extratemporal resections. We developed MELD‐PostOp, a deep learning tool trained and evaluated on a large, heterogeneous cohort to automatically segment resection cavities. Methods: The study included 1.5‐ and 3T postoperative three‐dimensional T1‐weighted MRI images from the Multicentre Epilepsy Lesion Detection (MELD) project ( n subjects = 969, 27 centers) and from the EPISURG dataset ( n = 133). The cohort included children and adults, alongside a range of resection locations, pathologies, and MRI characteristics. Resection cavities were individually segmented in 285 subjects and used to train an nnU‐Net prototype model. The prototype model was used to generate an additional 680 resection masks, which were subsequently quality‐controlled, edited, and combined with the original 285 to train the final MELD‐PostOp model ( n = 965). A Stratified Test Cohort ( n = 50) and Independent Test Cohort ( n = 87) were withheld for model evaluation. Performance was evaluated using Dice similarity coefficient (DSC), 95th percentile Hausdorff distance (HD95), number of predicted clusters, and inference runtime, and compared against established tools (Epic‐CHOP, ResectVol, and RESSEG). Results: MELD‐PostOp achieved a median DSC of .85 and HD95 of 3.61 on the combined test cohort, outperforming Epic‐CHOP (DSC .69, HD95 9.67), ResectVol (DSC .66, HD95 15.05), and RESSEG (DSC .43, HD95 32.67), with significant improvements seen in both temporal and especially extratemporal resections. The model detected 98.5% (135/137) of resection cavities. MELD‐PostOp runtime was 17 s per MRI, compared to 612 s (ResectVol), 3205 s (Epic‐CHOP), and 4 s (RESSEG). MELD‐PostOp performance remained high across clinical and imaging subgroups (median DSC > .8). Significance: MELD‐PostOp is an open‐source research tool that provides an accurate, efficient, and generalizable solution for postoperative resection cavity segmentation using only postoperative MRI scans.
Open AccessAutomated segmentation of postsurgical resection cavities on magnetic resonance imaging in focal epilepsy: A Multicentre Epilepsy Lesion Detection study
Objective: Quantitative assessment of extent of tissue resection following epilepsy surgery requires accurate delineation of the resection cavity on postoperative magnetic resonance imaging (MRI). Current methods for resection cavity masking are time‐consuming and labor‐intensive, and existing automated approaches exhibit variable segmentation accuracy, particularly on extratemporal resections. We developed MELD‐PostOp, a deep learning tool trained and evaluated on a large, heterogeneous cohort to automatically segment resection cavities. Methods: The study included 1.5‐ and 3T postoperative three‐dimensional T1‐weighted MRI images from the Multicentre Epilepsy Lesion Detection (MELD) project ( n subjects = 969, 27 centers) and from the EPISURG dataset ( n = 133). The cohort included children and adults, alongside a range of resection locations, pathologies, and MRI characteristics. Resection cavities were individually segmented in 285 subjects and used to train an nnU‐Net prototype model. The prototype model was used to generate an additional 680 resection masks, which were subsequently quality‐controlled, edited, and combined with the original 285 to train the final MELD‐PostOp model ( n = 965). A Stratified Test Cohort ( n = 50) and Independent Test Cohort ( n = 87) were withheld for model evaluation. Performance was evaluated using Dice similarity coefficient (DSC), 95th percentile Hausdorff distance (HD95), number of predicted clusters, and inference runtime, and compared against established tools (Epic‐CHOP, ResectVol, and RESSEG). Results: MELD‐PostOp achieved a median DSC of .85 and HD95 of 3.61 on the combined test cohort, outperforming Epic‐CHOP (DSC .69, HD95 9.67), ResectVol (DSC .66, HD95 15.05), and RESSEG (DSC .43, HD95 32.67), with significant improvements seen in both temporal and especially extratemporal resections. The model detected 98.5% (135/137) of resection cavities. MELD‐PostOp runtime was 17 s per MRI, compared to 612 s (ResectVol), 3205 s (Epic‐CHOP), and 4 s (RESSEG). MELD‐PostOp performance remained high across clinical and imaging subgroups (median DSC > .8). Significance: MELD‐PostOp is an open‐source research tool that provides an accurate, efficient, and generalizable solution for postoperative resection cavity segmentation using only postoperative MRI scans.
- Open Access
- 2 July 2026
- Published by MDPI AG in Diagnostics
- Vol. 16 (13) , 2070
Fine Needle Aspiration Biopsy of Thyroid Nodules Using Aspiration vs. Capillary Technique: Our Experience
Background/Objectives: Although ultrasound-guided fine-needle aspiration biopsy with suction (FNA-S) is preferred for evaluating thyroid nodules, the capillary technique (FNA-C) is also used. However, the diagnostic performance of both techniques remains unclear. Our objective was to compare the diagnostic performance of FNA-S versus FNA-C for thyroid nodules. Methods: This retrospective, observational, exploratory, single-center pilot evaluation study was conducted at a tertiary hospital from January 2023 to June 2024. A total of 157 ultrasound-guided FNA biopsies were prospectively analyzed, comprising 71 FNA-S and 86 FNA-C procedures. Clinical, ultrasound, and cytological parameters were evaluated to compare the rates of non-diagnostic samples (Bethesda I) and the necessity for repeat punctures. Bivariate analyses and multivariate logistic regression were conducted. Results: Both techniques exhibited comparable rates of non-diagnostic samples (28.17% for FNA-S and 27.91% for FNA-C; p = 0.220) and repetition rates (38.03% for FNA-S and 43.02% for FNA-C; p = 0.637). The only factors significantly associated with sample adequacy, regardless of techniques employed, were the location and composition of the nodule. Conclusions: Ultrasound-guided FNA-S and FNA-C yield similar diagnostic performance for biopsy of thyroid nodules. The choice may depend on operator preference and nodule characteristics.
- 2 July 2026
- Published by MDPI AG in Diagnostics
- Vol. 16 (13) , 2070
Open AccessFine Needle Aspiration Biopsy of Thyroid Nodules Using Aspiration vs. Capillary Technique: Our Experience
Background/Objectives: Although ultrasound-guided fine-needle aspiration biopsy with suction (FNA-S) is preferred for evaluating thyroid nodules, the capillary technique (FNA-C) is also used. However, the diagnostic performance of both techniques remains unclear. Our objective was to compare the diagnostic performance of FNA-S versus FNA-C for thyroid nodules. Methods: This retrospective, observational, exploratory, single-center pilot evaluation study was conducted at a tertiary hospital from January 2023 to June 2024. A total of 157 ultrasound-guided FNA biopsies were prospectively analyzed, comprising 71 FNA-S and 86 FNA-C procedures. Clinical, ultrasound, and cytological parameters were evaluated to compare the rates of non-diagnostic samples (Bethesda I) and the necessity for repeat punctures. Bivariate analyses and multivariate logistic regression were conducted. Results: Both techniques exhibited comparable rates of non-diagnostic samples (28.17% for FNA-S and 27.91% for FNA-C; p = 0.220) and repetition rates (38.03% for FNA-S and 43.02% for FNA-C; p = 0.637). The only factors significantly associated with sample adequacy, regardless of techniques employed, were the location and composition of the nodule. Conclusions: Ultrasound-guided FNA-S and FNA-C yield similar diagnostic performance for biopsy of thyroid nodules. The choice may depend on operator preference and nodule characteristics.
- Open Access
Toward trustworthy clinical AI for obsessive-compulsive disorder: reliability, generalizability, and interpretability of a transformer model across the ENIGMA-OCD consortium
Background. Studies applying machine learning to obsessive-compulsive disorder (OCD) typically report accuracy in homogeneous samples but rarely assess model reliability, generalizability, and interpretability needed for clinical use. Methods. We applied a transformer-based deep learning model, the Multi-Band Brain Net, to the ENIGMA-OCD cohort - the largest available resting-state functional magnetic resonance imaging (rs-fMRI) dataset in OCD with 1,706 participants (869 cases with OCD, 837 controls) across 23 sites worldwide. We evaluated model reliability by calculating calibration - the model's ability to "know what it doesn't know". We assessed generalizability using leave-one-site-out validation to test performance on unseen sites with different scanners, acquisition protocols, and patient populations. Finally, we examined interpretability by analyzing model attention weights to identify the neural connectivity patterns that influence model predictions. Results. The model achieved modest but competitive classification performance (AUROC = .653, SD = .039). Crucially, while large-scale pretraining on the UK Biobank (N = 40,783) did not boost accuracy, it significantly enhanced model calibration by reducing overconfident predictions. Leave-one-site-out validation showed a generalization gap across sites (AUROC = .427-.819). Pretraining did not close this gap but removed scanner manufacturer bias. Finally, attention-based mapping identified biologically plausible patterns of widespread hypoconnectivity in OCD relative to healthy controls, particularly in low-frequency bands involving the default mode, salience, and somatomotor networks. These findings aligned with known OCD neurobiology. Conclusions. This study provides a framework for developing more reliable and trustworthy clinical artificial intelligence for OCD.
Open AccessToward trustworthy clinical AI for obsessive-compulsive disorder: reliability, generalizability, and interpretability of a transformer model across the ENIGMA-OCD consortium
Background. Studies applying machine learning to obsessive-compulsive disorder (OCD) typically report accuracy in homogeneous samples but rarely assess model reliability, generalizability, and interpretability needed for clinical use. Methods. We applied a transformer-based deep learning model, the Multi-Band Brain Net, to the ENIGMA-OCD cohort - the largest available resting-state functional magnetic resonance imaging (rs-fMRI) dataset in OCD with 1,706 participants (869 cases with OCD, 837 controls) across 23 sites worldwide. We evaluated model reliability by calculating calibration - the model's ability to "know what it doesn't know". We assessed generalizability using leave-one-site-out validation to test performance on unseen sites with different scanners, acquisition protocols, and patient populations. Finally, we examined interpretability by analyzing model attention weights to identify the neural connectivity patterns that influence model predictions. Results. The model achieved modest but competitive classification performance (AUROC = .653, SD = .039). Crucially, while large-scale pretraining on the UK Biobank (N = 40,783) did not boost accuracy, it significantly enhanced model calibration by reducing overconfident predictions. Leave-one-site-out validation showed a generalization gap across sites (AUROC = .427-.819). Pretraining did not close this gap but removed scanner manufacturer bias. Finally, attention-based mapping identified biologically plausible patterns of widespread hypoconnectivity in OCD relative to healthy controls, particularly in low-frequency bands involving the default mode, salience, and somatomotor networks. These findings aligned with known OCD neurobiology. Conclusions. This study provides a framework for developing more reliable and trustworthy clinical artificial intelligence for OCD.
- 13 April 2026
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- p. ajnr.A9348
Intraparotid Facial Nerve Segmentation: A Reproducibility Study of Manual Segmentation on Neurographic Sequences
Preoperative delineation of intraparotid facial nerve using advanced MR neurography may aid surgical planning of parotid tumors, reducing complications. However, the reliability of manual segmentation for depicting the nerve pathway has not been extensively investigated. This study aimed to assess the reproducibility of manual segmentation of the intraparotid facial nerve on neurographic MRI sequences. Neurographic MR datasets from 7 patients with unilateral benign parotid tumors were analyzed using 2 imaging sequences: double-echo steady-state (DESS) and T1 variable-flip-angle turbo spin-echo (VFA-TSE) with gadolinium. Two faculty head and neck radiologists and two radiology trainees independently segmented the intraparotid facial nerve in both tumorous and healthy glands using 3D Slicer. Given the absence of a gold standard, segmentation reliability was quantified through mean pairwise error, interobserver segmentation agreement for anatomical nerve branches, and reliability profiles along the nerve course. Comparisons by imaging sequence, gland status, and radiologist experience were performed using Wilcoxon signed-rank tests with Holm-Bonferroni correction. The mean pairwise error was 1.58 [95% CI: 1.38-2.04] mm for the trunk, 3.20 [95% CI: 2.94-4.50] mm for the inferior division, and 3.58 [95% CI: 2.29-5.99] mm for the superior division. The VFA-TSE sequence showed lower pairwise segmentation errors than the DESS sequence, particularly for the trunk (1.26 [95% CI: 0.84-2.10] mm vs 1.85 [95% CI: 1.55-2.37] mm), although no comparisons reached statistical significance after correction for multiple testing. Nerve segments near tumors exhibited increased segmentation errors and greater interobserver variability. No significant differences were observed between radiologists of differing experience. The trunk showed high interobserver agreement, whereas peripheral branches displayed lower consistency, more frequent non-segmentation and increased subdivision instances. Manual segmentation of the intraparotid facial nerve is reproducible for the trunk but remains challenging for peripheral branches. VFA-TSE may offer an advantage over DESS for trunk delineation, while radiologist experience level does not appear to limit segmentation reliability. Standardized segmentation protocols can mitigate the impact of prior experience, but imaging sequence choice may influence segmentation reproducibility. Further research should refine neurographic sequence and explore automated segmentation techniques to improve consistency in presurgical planning of parotid tumors.
- 13 April 2026
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- p. ajnr.A9348
Intraparotid Facial Nerve Segmentation: A Reproducibility Study of Manual Segmentation on Neurographic Sequences
Preoperative delineation of intraparotid facial nerve using advanced MR neurography may aid surgical planning of parotid tumors, reducing complications. However, the reliability of manual segmentation for depicting the nerve pathway has not been extensively investigated. This study aimed to assess the reproducibility of manual segmentation of the intraparotid facial nerve on neurographic MRI sequences. Neurographic MR datasets from 7 patients with unilateral benign parotid tumors were analyzed using 2 imaging sequences: double-echo steady-state (DESS) and T1 variable-flip-angle turbo spin-echo (VFA-TSE) with gadolinium. Two faculty head and neck radiologists and two radiology trainees independently segmented the intraparotid facial nerve in both tumorous and healthy glands using 3D Slicer. Given the absence of a gold standard, segmentation reliability was quantified through mean pairwise error, interobserver segmentation agreement for anatomical nerve branches, and reliability profiles along the nerve course. Comparisons by imaging sequence, gland status, and radiologist experience were performed using Wilcoxon signed-rank tests with Holm-Bonferroni correction. The mean pairwise error was 1.58 [95% CI: 1.38-2.04] mm for the trunk, 3.20 [95% CI: 2.94-4.50] mm for the inferior division, and 3.58 [95% CI: 2.29-5.99] mm for the superior division. The VFA-TSE sequence showed lower pairwise segmentation errors than the DESS sequence, particularly for the trunk (1.26 [95% CI: 0.84-2.10] mm vs 1.85 [95% CI: 1.55-2.37] mm), although no comparisons reached statistical significance after correction for multiple testing. Nerve segments near tumors exhibited increased segmentation errors and greater interobserver variability. No significant differences were observed between radiologists of differing experience. The trunk showed high interobserver agreement, whereas peripheral branches displayed lower consistency, more frequent non-segmentation and increased subdivision instances. Manual segmentation of the intraparotid facial nerve is reproducible for the trunk but remains challenging for peripheral branches. VFA-TSE may offer an advantage over DESS for trunk delineation, while radiologist experience level does not appear to limit segmentation reliability. Standardized segmentation protocols can mitigate the impact of prior experience, but imaging sequence choice may influence segmentation reproducibility. Further research should refine neurographic sequence and explore automated segmentation techniques to improve consistency in presurgical planning of parotid tumors.
- Open Access
Mapping Higher-Order Topology in OCD Brain Networks with Hodge Laplacian
Brain disorders are increasingly understood as disorders of distributed brain circuits, yet functional connectivity (FC), the dominant framework for mapping them, treats the brain as a collection of pairwise relationships between regions and cannot represent pathology distributed across coordinated sets of connections. We introduce a Hodge-Laplacian topological framework that localizes higher-order “loop” (1-cycle) organization within functional connectome, maps each loop to specific edges and networks, and yields a subject-level measure of loop expression. Applied to resting-state fMRI from the ENIGMA-OCD consortium (1,024 patients and 1,028 controls across 28 sites), the framework identified 93 loop-level abnormalities in obsessive–compulsive disorder (OCD), concentrated in frontoparietal and somatomotor systems. The edges forming these loops largely showed no significant differences between groups, indicating that the abnormalities were invisible to conventional FC analysis. The frontoparietal and somatomotor loop clusters recurred across the clinical subgroups, suggesting convergence on a shared higher-order phenotype. Robustness analyses showed the loop signal reflected higher-order organization rather than an artifact of individual edges, the network backbone, or any single site. These results indicate that coordinated, multi-edge pathology exists and can be localized even when pairwise analyses fail to detect it, positioning higher-order topology as a generalizable axis for mapping circuit pathology across psychiatric and neurological disorders.
Open AccessMapping Higher-Order Topology in OCD Brain Networks with Hodge Laplacian
Brain disorders are increasingly understood as disorders of distributed brain circuits, yet functional connectivity (FC), the dominant framework for mapping them, treats the brain as a collection of pairwise relationships between regions and cannot represent pathology distributed across coordinated sets of connections. We introduce a Hodge-Laplacian topological framework that localizes higher-order “loop” (1-cycle) organization within functional connectome, maps each loop to specific edges and networks, and yields a subject-level measure of loop expression. Applied to resting-state fMRI from the ENIGMA-OCD consortium (1,024 patients and 1,028 controls across 28 sites), the framework identified 93 loop-level abnormalities in obsessive–compulsive disorder (OCD), concentrated in frontoparietal and somatomotor systems. The edges forming these loops largely showed no significant differences between groups, indicating that the abnormalities were invisible to conventional FC analysis. The frontoparietal and somatomotor loop clusters recurred across the clinical subgroups, suggesting convergence on a shared higher-order phenotype. Robustness analyses showed the loop signal reflected higher-order organization rather than an artifact of individual edges, the network backbone, or any single site. These results indicate that coordinated, multi-edge pathology exists and can be localized even when pairwise analyses fail to detect it, positioning higher-order topology as a generalizable axis for mapping circuit pathology across psychiatric and neurological disorders.
- Open Access
- 22 October 2025
- Published by Springer Nature in Molecular Psychiatry
- Vol. 31 (3) , 1707-1722
Structural covariance network topology in individuals at clinical high risk for psychosis: the ENIGMA-CHR Study
Brain network architecture is anticipated to influence future grey matter loss in individuals at Clinical High Risk (CHR) for psychosis. However, existing studies on grey matter structural network properties in CHR are scarce and constrained by small sample sizes. Here, we examined network topology differences comparing a) CHR versus healthy controls (HC); b) CHR who transitioned to psychosis (CHR-T) versus those who did not (CHR-NT); and c) different subsyndromes. We included structural scans from 1842 CHR individuals and 1417 HC individuals from 31 sites within the Enhancing NeuroImaging Genetics through Meta-Analysis (ENIGMA) consortium. At the global level, CHR individuals exhibited lower structural covariance (q < 0.001; Cohen’s d = 0.164) and less optimal structural network configuration than HC (lower global efficiency and clustering coefficient, d = 0.100,0.087, qs <= 0.027). Though no global difference between CHR-T and CHR-NT, network distinctiveness of the frontal and temporal surface area networks was higher in CHR-T than CHR-NT (d = 0.223,0.237) and HC (d = 0.208,0.219) (qs < 0.001). Network distinctiveness of the frontal cortical thickness network was lower in CHR-T (d = 0.218, q < 0.001) than CHR-NT and HC (d = 0.165, q < 0.001). Importantly, higher network distinctiveness was associated with worse positive symptoms in CHR-NT (frontal surface area, q = 0.008, R2 = 0.013) and at trend with worse negative symptoms in CHR-T (frontal thickness, q = 0.063, R2 = 0.049). Further, the brief intermittent psychotic syndrome subgroup showed more severe network alterations. Together, brain structural networks inform symptoms and the risk of transition to psychosis in CHR individuals.
- 22 October 2025
- Published by Springer Nature in Molecular Psychiatry
- Vol. 31 (3) , 1707-1722
Open AccessStructural covariance network topology in individuals at clinical high risk for psychosis: the ENIGMA-CHR Study
Brain network architecture is anticipated to influence future grey matter loss in individuals at Clinical High Risk (CHR) for psychosis. However, existing studies on grey matter structural network properties in CHR are scarce and constrained by small sample sizes. Here, we examined network topology differences comparing a) CHR versus healthy controls (HC); b) CHR who transitioned to psychosis (CHR-T) versus those who did not (CHR-NT); and c) different subsyndromes. We included structural scans from 1842 CHR individuals and 1417 HC individuals from 31 sites within the Enhancing NeuroImaging Genetics through Meta-Analysis (ENIGMA) consortium. At the global level, CHR individuals exhibited lower structural covariance (q < 0.001; Cohen’s d = 0.164) and less optimal structural network configuration than HC (lower global efficiency and clustering coefficient, d = 0.100,0.087, qs <= 0.027). Though no global difference between CHR-T and CHR-NT, network distinctiveness of the frontal and temporal surface area networks was higher in CHR-T than CHR-NT (d = 0.223,0.237) and HC (d = 0.208,0.219) (qs < 0.001). Network distinctiveness of the frontal cortical thickness network was lower in CHR-T (d = 0.218, q < 0.001) than CHR-NT and HC (d = 0.165, q < 0.001). Importantly, higher network distinctiveness was associated with worse positive symptoms in CHR-NT (frontal surface area, q = 0.008, R2 = 0.013) and at trend with worse negative symptoms in CHR-T (frontal thickness, q = 0.063, R2 = 0.049). Further, the brief intermittent psychotic syndrome subgroup showed more severe network alterations. Together, brain structural networks inform symptoms and the risk of transition to psychosis in CHR individuals.
- Open Access
- 5 March 2025
- Published by Springer Nature in Neuroradiology
- Vol. 67 (4) , 865-874
Potential role of MELD and MAP18 in patients with structural temporal lobe epilepsy
Purpose: This study compared two image post-processing toolboxes primarily designed for focal cortical dysplasia (FCD): Multi-Centre Epilepsy Lesion Detection (MELD) and Morphometric Analysis Program (MAP18), in identifying temporal lobe epilepsy (TLE) structural lesions on MRI. Methods: This retrospective study examined 79 adults, 58 patients with confirmed TLE, and 21 healthy controls. All participants underwent an elective brain MRI between June 2007 – May 2023 at Hospital Clinic, Barcelona, Spain. All the 3D T1-weighted images were processed using MELD and MAP18 to detect potential epileptogenic lesions. The location (lateral or mesial) and laterality of the reference TLE structural lesion (refTLE) were determined through histopathology or multidisciplinary consensus based on clinical data. Toolboxes’ performance was evaluated using descriptive statistics, specificity, and diagnostic accuracy. Additionally, a second-look MRI was conducted for cases where abnormalities detected by MELD and MAP18 did not match the refTLE. Results: MELD and MAP18 demonstrated variability in specificity and diagnostic accuracy. Specificity ranged from 48% to 86%, with ProbMAP (MAP18) achieving the highest values. Global diagnostic accuracy ranged from 7% to 42%, with MELD showing the highest performance. In four patients with visible epileptogenic lesions on MRI, MELD and MAP18 identified additional abnormalities that were previously overlooked. Moreover, MELD detected one TLE lesion in one patient initially classified as MRI–negative (nonlesional). Conclusion: Incorporating tools like MELD and MAP18 into the diagnostic workflow can enhance the detection of TLE-related abnormalities on MRI, potentially improving patient outcomes and aiding in clinical decision-making.
- 5 March 2025
- Published by Springer Nature in Neuroradiology
- Vol. 67 (4) , 865-874
Open AccessPotential role of MELD and MAP18 in patients with structural temporal lobe epilepsy
Purpose: This study compared two image post-processing toolboxes primarily designed for focal cortical dysplasia (FCD): Multi-Centre Epilepsy Lesion Detection (MELD) and Morphometric Analysis Program (MAP18), in identifying temporal lobe epilepsy (TLE) structural lesions on MRI. Methods: This retrospective study examined 79 adults, 58 patients with confirmed TLE, and 21 healthy controls. All participants underwent an elective brain MRI between June 2007 – May 2023 at Hospital Clinic, Barcelona, Spain. All the 3D T1-weighted images were processed using MELD and MAP18 to detect potential epileptogenic lesions. The location (lateral or mesial) and laterality of the reference TLE structural lesion (refTLE) were determined through histopathology or multidisciplinary consensus based on clinical data. Toolboxes’ performance was evaluated using descriptive statistics, specificity, and diagnostic accuracy. Additionally, a second-look MRI was conducted for cases where abnormalities detected by MELD and MAP18 did not match the refTLE. Results: MELD and MAP18 demonstrated variability in specificity and diagnostic accuracy. Specificity ranged from 48% to 86%, with ProbMAP (MAP18) achieving the highest values. Global diagnostic accuracy ranged from 7% to 42%, with MELD showing the highest performance. In four patients with visible epileptogenic lesions on MRI, MELD and MAP18 identified additional abnormalities that were previously overlooked. Moreover, MELD detected one TLE lesion in one patient initially classified as MRI–negative (nonlesional). Conclusion: Incorporating tools like MELD and MAP18 into the diagnostic workflow can enhance the detection of TLE-related abnormalities on MRI, potentially improving patient outcomes and aiding in clinical decision-making.
- Open Access
- 7 March 2024
- Published by Springer Nature in Molecular Psychiatry
- Vol. 29 (4) , 1216
Correction: White matter diffusion estimates in obsessive-compulsive disorder across 1653 individuals: machine learning findings from the ENIGMA OCD Working Group
- 7 March 2024
- Published by Springer Nature in Molecular Psychiatry
- Vol. 29 (4) , 1216
Open AccessCorrection: White matter diffusion estimates in obsessive-compulsive disorder across 1653 individuals: machine learning findings from the ENIGMA OCD Working Group
- Open Access
- 7 February 2024
- Published by Springer Nature in Molecular Psychiatry
- Vol. 29 (4) , 1063-1074
White matter diffusion estimates in obsessive-compulsive disorder across 1653 individuals: machine learning findings from the ENIGMA OCD Working Group
White matter pathways, typically studied with diffusion tensor imaging (DTI), have been implicated in the neurobiology of obsessive-compulsive disorder (OCD). However, due to limited sample sizes and the predominance of single-site studies, the generalizability of OCD classification based on diffusion white matter estimates remains unclear. Here, we tested classification accuracy using the largest OCD DTI dataset to date, involving 1336 adult participants (690 OCD patients and 646 healthy controls) and 317 pediatric participants (175 OCD patients and 142 healthy controls) from 18 international sites within the ENIGMA OCD Working Group. We used an automatic machine learning pipeline (with feature engineering and selection, and model optimization) and examined the cross-site generalizability of the OCD classification models using leave-one-site-out cross-validation. Our models showed low-to-moderate accuracy in classifying (1) “OCD vs. healthy controls” (Adults, receiver operator characteristic-area under the curve = 57.19 ± 3.47 in the replication set; Children, 59.8 ± 7.39), (2) “unmedicated OCD vs. healthy controls” (Adults, 62.67 ± 3.84; Children, 48.51 ± 10.14), and (3) “medicated OCD vs. unmedicated OCD” (Adults, 76.72 ± 3.97; Children, 72.45 ± 8.87). There was significant site variability in model performance (cross-validated ROC AUC ranges 51.6–79.1 in adults; 35.9–63.2 in children). Machine learning interpretation showed that diffusivity measures of the corpus callosum, internal capsule, and posterior thalamic radiation contributed to the classification of OCD from HC. The classification performance appeared greater than the model trained on grey matter morphometry in the prior ENIGMA OCD study (our study includes subsamples from the morphometry study). Taken together, this study points to the meaningful multivariate patterns of white matter features relevant to the neurobiology of OCD, but with low-to-moderate classification accuracy. The OCD classification performance may be constrained by site variability and medication effects on the white matter integrity, indicating room for improvement for future research.
- 7 February 2024
- Published by Springer Nature in Molecular Psychiatry
- Vol. 29 (4) , 1063-1074
Open AccessWhite matter diffusion estimates in obsessive-compulsive disorder across 1653 individuals: machine learning findings from the ENIGMA OCD Working Group
White matter pathways, typically studied with diffusion tensor imaging (DTI), have been implicated in the neurobiology of obsessive-compulsive disorder (OCD). However, due to limited sample sizes and the predominance of single-site studies, the generalizability of OCD classification based on diffusion white matter estimates remains unclear. Here, we tested classification accuracy using the largest OCD DTI dataset to date, involving 1336 adult participants (690 OCD patients and 646 healthy controls) and 317 pediatric participants (175 OCD patients and 142 healthy controls) from 18 international sites within the ENIGMA OCD Working Group. We used an automatic machine learning pipeline (with feature engineering and selection, and model optimization) and examined the cross-site generalizability of the OCD classification models using leave-one-site-out cross-validation. Our models showed low-to-moderate accuracy in classifying (1) “OCD vs. healthy controls” (Adults, receiver operator characteristic-area under the curve = 57.19 ± 3.47 in the replication set; Children, 59.8 ± 7.39), (2) “unmedicated OCD vs. healthy controls” (Adults, 62.67 ± 3.84; Children, 48.51 ± 10.14), and (3) “medicated OCD vs. unmedicated OCD” (Adults, 76.72 ± 3.97; Children, 72.45 ± 8.87). There was significant site variability in model performance (cross-validated ROC AUC ranges 51.6–79.1 in adults; 35.9–63.2 in children). Machine learning interpretation showed that diffusivity measures of the corpus callosum, internal capsule, and posterior thalamic radiation contributed to the classification of OCD from HC. The classification performance appeared greater than the model trained on grey matter morphometry in the prior ENIGMA OCD study (our study includes subsamples from the morphometry study). Taken together, this study points to the meaningful multivariate patterns of white matter features relevant to the neurobiology of OCD, but with low-to-moderate classification accuracy. The OCD classification performance may be constrained by site variability and medication effects on the white matter integrity, indicating room for improvement for future research.