Michel Bilello
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Michel Bilello has a total of 146 co-authors. Here are the top co-authors by number of shared publications:
- Christos Davatzikos
Artificial Intelligence in Biomedical Imaging Laboratory (AIBIL), Center for AI and Data Science for Integrated Diagnostics (AI2D), Perelman School of Medicine, University of Pennsylvania, Philadelphia, USA
6 shared publications - Hamed Akbari
Department of Bioengineering, School of Engineering, Santa Clara University, Santa Clara, CA, USA
5 shared publications - Spyridon Bakas
Division of Computational Pathology, Department of Pathology and Laboratory Medicine, Indiana University School of Medicine, Indianapolis, IN, USA
5 shared publications - Sarthak Pati
Division of Computational Pathology, Department of Pathology and Laboratory Medicine, Indiana University School of Medicine, Indianapolis, IN, USA; Department of Informatics, Technical University of Munich, Munich, Germany
4 shared publications - Ashish Singh
Center for Biomedical Image Computing and Analytics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, USA
3 shared publications - Chiharu Sako
Center for Data Science and AI for Integrated Diagnostics (AI2D), University of Pennsylvania, Philadelphia, PA, USA
3 shared publications
Latest Publications
- 14 July 2026
- Published by Wolters Kluwer Health in Neurology
- Vol. 107 (1) , e218183
Application of the 2024 McDonald Criteria in Individuals With Nonspecific Symptoms or Incidental Imaging Findings in a Multicenter Study
The 2024 McDonald criteria allow diagnosis of multiple sclerosis (MS) in individuals presenting with symptoms not specific for MS or incidental imaging findings suggestive of demyelination when supported by biomarker evidence, reflecting a shift toward diagnostic definitions increasingly grounded in biological mechanisms of disease. The diagnostic yield of these criteria in such populations has not been evaluated in multicenter cohorts. We aimed to determine the proportion of individuals with nonspecific or incidental imaging presentations who meet the 2024 McDonald criteria and describe the contribution of central vein sign and cerebrospinal fluid (CSF) oligoclonal bands (OCBs) to diagnostic classification. This cross-sectional post-hoc analysis used data from the Central Vein Sign in Multiple Sclerosis (CAVS-MS) study, a multicenter observational cohort. Adults aged 18–65 years referred for diagnostic evaluation of possible MS were adjudicated by an expert panel (eFigure1). This analysis focused on participants with symptoms not specific for MS or incidental imaging findings suggestive of demyelination. Dissemination in space (DIS) and time (DIT) were assessed using 2017 MRI criteria. Fulfillment of the 2024 McDonald criteria at baseline—the primary outcome—was determined using the Select-6 central vein sign and CSF OCBs. Select-6 assessment was available for all participants, whereas OCB data were available for a subset based on prior clinical evaluation. Of 420 participants enrolled, 191 (45%) presented with either nonspecific symptoms (n=166) or incidental imaging findings (n=25). Mean age was 42 years, and 78% were female. Thirty-six (19%) met the 2024 McDonald criteria at baseline, including 28 (17%) in the nonspecific symptom cohort and 8 (32%) in the incidental imaging cohort. Among 51 participants meeting 2017 DIS, 22 (43%) were Select-6 positive, 17 (33%) had positive OCBs, and 4 (8%) met 2017 DIT. Nonspecific sensory symptoms, visual disturbances, and subacute cognitive decline were most associated with a diagnosis of MS. Application of the 2024 McDonald criteria identified nearly one-fifth of individuals without typical presentations as meeting diagnostic criteria for MS at baseline. Biomarker incorporation—particularly the central vein sign—accounted for a substantial proportion of diagnostic yield. Interpretation is limited by availability of CSF data and absence of longitudinal follow-up.
- 14 July 2026
- Published by Wolters Kluwer Health in Neurology
- Vol. 107 (1) , e218183
Application of the 2024 McDonald Criteria in Individuals With Nonspecific Symptoms or Incidental Imaging Findings in a Multicenter Study
The 2024 McDonald criteria allow diagnosis of multiple sclerosis (MS) in individuals presenting with symptoms not specific for MS or incidental imaging findings suggestive of demyelination when supported by biomarker evidence, reflecting a shift toward diagnostic definitions increasingly grounded in biological mechanisms of disease. The diagnostic yield of these criteria in such populations has not been evaluated in multicenter cohorts. We aimed to determine the proportion of individuals with nonspecific or incidental imaging presentations who meet the 2024 McDonald criteria and describe the contribution of central vein sign and cerebrospinal fluid (CSF) oligoclonal bands (OCBs) to diagnostic classification. This cross-sectional post-hoc analysis used data from the Central Vein Sign in Multiple Sclerosis (CAVS-MS) study, a multicenter observational cohort. Adults aged 18–65 years referred for diagnostic evaluation of possible MS were adjudicated by an expert panel (eFigure1). This analysis focused on participants with symptoms not specific for MS or incidental imaging findings suggestive of demyelination. Dissemination in space (DIS) and time (DIT) were assessed using 2017 MRI criteria. Fulfillment of the 2024 McDonald criteria at baseline—the primary outcome—was determined using the Select-6 central vein sign and CSF OCBs. Select-6 assessment was available for all participants, whereas OCB data were available for a subset based on prior clinical evaluation. Of 420 participants enrolled, 191 (45%) presented with either nonspecific symptoms (n=166) or incidental imaging findings (n=25). Mean age was 42 years, and 78% were female. Thirty-six (19%) met the 2024 McDonald criteria at baseline, including 28 (17%) in the nonspecific symptom cohort and 8 (32%) in the incidental imaging cohort. Among 51 participants meeting 2017 DIS, 22 (43%) were Select-6 positive, 17 (33%) had positive OCBs, and 4 (8%) met 2017 DIT. Nonspecific sensory symptoms, visual disturbances, and subacute cognitive decline were most associated with a diagnosis of MS. Application of the 2024 McDonald criteria identified nearly one-fifth of individuals without typical presentations as meeting diagnostic criteria for MS at baseline. Biomarker incorporation—particularly the central vein sign—accounted for a substantial proportion of diagnostic yield. Interpretation is limited by availability of CSF data and absence of longitudinal follow-up.
- 12 October 2022
- Published by IOP Publishing in Physics in Medicine & Biology
- Vol. 67 (20) , 204002
The federated tumor segmentation (FeTS) tool: an open-source solution to further solid tumor research
Objective. De-centralized data analysis becomes an increasingly preferred option in the healthcare domain, as it alleviates the need for sharing primary patient data across collaborating institutions. This highlights the need for consistent harmonized data curation, pre-processing, and identification of regions of interest based on uniform criteria. Approach. Towards this end, this manuscript describes the Fe derated T umor S egmentation (FeTS) tool, in terms of software architecture and functionality. Main results. The primary aim of the FeTS tool is to facilitate this harmonized processing and the generation of gold standard reference labels for tumor sub-compartments on brain magnetic resonance imaging, and further enable federated training of a tumor sub-compartment delineation model across numerous sites distributed across the globe, without the need to share patient data. Significance. Building upon existing open-source tools such as the Insight Toolkit and Qt, the FeTS tool is designed to enable training deep learning models targeting tumor delineation in either centralized or federated settings. The target audience of the FeTS tool is primarily the computational researcher interested in developing federated learning models, and interested in joining a global federation towards this effort. The tool is open sourced at https://github.com/FETS-AI/Front-End .
- 12 October 2022
- Published by IOP Publishing in Physics in Medicine & Biology
- Vol. 67 (20) , 204002
The federated tumor segmentation (FeTS) tool: an open-source solution to further solid tumor research
Objective. De-centralized data analysis becomes an increasingly preferred option in the healthcare domain, as it alleviates the need for sharing primary patient data across collaborating institutions. This highlights the need for consistent harmonized data curation, pre-processing, and identification of regions of interest based on uniform criteria. Approach. Towards this end, this manuscript describes the Fe derated T umor S egmentation (FeTS) tool, in terms of software architecture and functionality. Main results. The primary aim of the FeTS tool is to facilitate this harmonized processing and the generation of gold standard reference labels for tumor sub-compartments on brain magnetic resonance imaging, and further enable federated training of a tumor sub-compartment delineation model across numerous sites distributed across the globe, without the need to share patient data. Significance. Building upon existing open-source tools such as the Insight Toolkit and Qt, the FeTS tool is designed to enable training deep learning models targeting tumor delineation in either centralized or federated settings. The target audience of the FeTS tool is primarily the computational researcher interested in developing federated learning models, and interested in joining a global federation towards this effort. The tool is open sourced at https://github.com/FETS-AI/Front-End .
- Top Cited PapersOpen Access
- 29 July 2022
- Published by Springer Nature in Scientific Data
- Vol. 9 (1) , 1-12
The University of Pennsylvania glioblastoma (UPenn-GBM) cohort: advanced MRI, clinical, genomics, & radiomics
Glioblastoma is the most common aggressive adult brain tumor. Numerous studies have reported results from either private institutional data or publicly available datasets. However, current public datasets are limited in terms of: a) number of subjects, b) lack of consistent acquisition protocol, c) data quality, or d) accompanying clinical, demographic, and molecular information. Toward alleviating these limitations, we contribute the “University of Pennsylvania Glioblastoma Imaging, Genomics, and Radiomics” (UPenn-GBM) dataset, which describes the currently largest publicly available comprehensive collection of 630 patients diagnosed withde novoglioblastoma. The UPenn-GBM dataset includes (a) advanced multi-parametric magnetic resonance imaging scans acquired during routine clinical practice, at the University of Pennsylvania Health System, (b) accompanying clinical, demographic, and molecular information, (d) perfusion and diffusion derivative volumes, (e) computationally-derived and manually-revised expert annotations of tumor sub-regions, as well as (f) quantitative imaging (also known as radiomic) features corresponding to each of these regions. This collection describes our contribution towards repeatable, reproducible, and comparative quantitative studies leading to new predictive, prognostic, and diagnostic assessments.
- 29 July 2022
- Published by Springer Nature in Scientific Data
- Vol. 9 (1) , 1-12
Top Cited PapersOpen AccessThe University of Pennsylvania glioblastoma (UPenn-GBM) cohort: advanced MRI, clinical, genomics, & radiomics
Glioblastoma is the most common aggressive adult brain tumor. Numerous studies have reported results from either private institutional data or publicly available datasets. However, current public datasets are limited in terms of: a) number of subjects, b) lack of consistent acquisition protocol, c) data quality, or d) accompanying clinical, demographic, and molecular information. Toward alleviating these limitations, we contribute the “University of Pennsylvania Glioblastoma Imaging, Genomics, and Radiomics” (UPenn-GBM) dataset, which describes the currently largest publicly available comprehensive collection of 630 patients diagnosed withde novoglioblastoma. The UPenn-GBM dataset includes (a) advanced multi-parametric magnetic resonance imaging scans acquired during routine clinical practice, at the University of Pennsylvania Health System, (b) accompanying clinical, demographic, and molecular information, (d) perfusion and diffusion derivative volumes, (e) computationally-derived and manually-revised expert annotations of tumor sub-regions, as well as (f) quantitative imaging (also known as radiomic) features corresponding to each of these regions. This collection describes our contribution towards repeatable, reproducible, and comparative quantitative studies leading to new predictive, prognostic, and diagnostic assessments.
- 16 March 2022
- Published by Wiley in Journal of Neuroimaging
- Vol. 32 (4) , 656-666
Economic impact of selective use of contrast for routine follow‐up MRI of patients with multiple sclerosis
Background and Purpose: Imaging and autopsy studies show intracranial gadolinium deposition in patients who have undergone serial contrast‐enhanced MRIs. This observation has raised concerns when using contrast administration in patients who receive frequent MRIs. To address this, we implemented a contrast‐conditional protocol wherein gadolinium is administered only for multiple sclerosis (MS) patients with imaging evidence of new disease activity on precontrast imaging. In this study, we explore the economic impact of our new MRI protocol.Methods: We compared scanner time and Medicare reimbursement using our contrast‐conditional methodology versus that of prior protocols where all patients received gadolinium.Results: For 422 patients over 4 months, the contrast‐conditional protocol amounted to 60% decrease in contrast injection and savings of approximately 20% of MRI scanner time. If the extra scanner time was used for performing MS follow‐up MRIs in additional patients, the contrast‐conditional protocol would amount to net revenue loss of $21,707 (∼3.7%).Conclusions: Implementation of a new protocol to limit contrast in MS follow‐up MRIs led to a minimal decrease in revenue when controlled for scanner time utilized and is outweighed by other benefits, including substantial decreased gadolinium administration, increased patient comfort, and increased availability of scanner time, which depending on type of studies performed could result in additional financial benefit.
- 16 March 2022
- Published by Wiley in Journal of Neuroimaging
- Vol. 32 (4) , 656-666
Economic impact of selective use of contrast for routine follow‐up MRI of patients with multiple sclerosis
Background and Purpose: Imaging and autopsy studies show intracranial gadolinium deposition in patients who have undergone serial contrast‐enhanced MRIs. This observation has raised concerns when using contrast administration in patients who receive frequent MRIs. To address this, we implemented a contrast‐conditional protocol wherein gadolinium is administered only for multiple sclerosis (MS) patients with imaging evidence of new disease activity on precontrast imaging. In this study, we explore the economic impact of our new MRI protocol.Methods: We compared scanner time and Medicare reimbursement using our contrast‐conditional methodology versus that of prior protocols where all patients received gadolinium.Results: For 422 patients over 4 months, the contrast‐conditional protocol amounted to 60% decrease in contrast injection and savings of approximately 20% of MRI scanner time. If the extra scanner time was used for performing MS follow‐up MRIs in additional patients, the contrast‐conditional protocol would amount to net revenue loss of $21,707 (∼3.7%).Conclusions: Implementation of a new protocol to limit contrast in MS follow‐up MRIs led to a minimal decrease in revenue when controlled for scanner time utilized and is outweighed by other benefits, including substantial decreased gadolinium administration, increased patient comfort, and increased availability of scanner time, which depending on type of studies performed could result in additional financial benefit.
- Open Access
- 30 August 2021
- Published by Wolters Kluwer Health in Stroke
- Vol. 52 (11) , 3419-3426
Patent Foramen Ovale Closure Decreases the Incidence but Not the Size of New Brain Infarction on Magnetic Resonance Imaging: An Analysis of the REDUCE Trial
Background and Purpose: Randomized patent foramen ovale closure trials have used open-label end point ascertainment which increases the risk of bias and undermines confidence in the conclusions. The Gore REDUCE trial prospectively performed baseline and follow-up magnetic resonance imaging (MRIs) for all subjects providing an objective measure of the effectiveness of closure. Methods: We performed blinded evaluations of the presence, location, and volume of new infarct on diffusion-weighted imaging of recurrent clinical stroke or new infarct (>3 mm) on T2/fluid attenuated inversion recovery from baseline to follow-up MRI at 2 years, comparing closure to medical therapy alone. We also examined the effect of shunt size and the development of atrial fibrillation on infarct burden at follow-up. Results: At follow-up, new clinical stroke or silent MRI infarct occurred in 18/383 (4.7%) patients who underwent closure and 19/177 (10.7%) medication-only patients (relative risk, 0.44 [95% CI, 0.24–0.81], P =0.02). Clinical strokes were less common in closure patients compared with medically treated patients, 5 (1.3%) versus 12 (6.8%), P =0.001, while silent MRI infarcts were similar, 13 (3.4%) versus 7 (4.0%), P =0.81. There were no differences in number, volumes, and distribution of new infarct comparing closure patients to those treated with medication alone. There were also no differences of number, volumes, and distribution comparing silent infarcts to clinical strokes. Infarct burden was also similar for patients who developed atrial fibrillation and for those with large shunts. Conclusions: The REDUCE trial demonstrates that patent foramen ovale closure prevents recurrent brain infarction based on the objective outcome of new infarcts on MRI. Only clinical strokes were reduced by closure while silent infarctions were similar between study arms, and there were no differences in infarct volume or location comparing silent infarcts to clinical strokes. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT00738894.
- 30 August 2021
- Published by Wolters Kluwer Health in Stroke
- Vol. 52 (11) , 3419-3426
Open AccessPatent Foramen Ovale Closure Decreases the Incidence but Not the Size of New Brain Infarction on Magnetic Resonance Imaging: An Analysis of the REDUCE Trial
Background and Purpose: Randomized patent foramen ovale closure trials have used open-label end point ascertainment which increases the risk of bias and undermines confidence in the conclusions. The Gore REDUCE trial prospectively performed baseline and follow-up magnetic resonance imaging (MRIs) for all subjects providing an objective measure of the effectiveness of closure. Methods: We performed blinded evaluations of the presence, location, and volume of new infarct on diffusion-weighted imaging of recurrent clinical stroke or new infarct (>3 mm) on T2/fluid attenuated inversion recovery from baseline to follow-up MRI at 2 years, comparing closure to medical therapy alone. We also examined the effect of shunt size and the development of atrial fibrillation on infarct burden at follow-up. Results: At follow-up, new clinical stroke or silent MRI infarct occurred in 18/383 (4.7%) patients who underwent closure and 19/177 (10.7%) medication-only patients (relative risk, 0.44 [95% CI, 0.24–0.81], P =0.02). Clinical strokes were less common in closure patients compared with medically treated patients, 5 (1.3%) versus 12 (6.8%), P =0.001, while silent MRI infarcts were similar, 13 (3.4%) versus 7 (4.0%), P =0.81. There were no differences in number, volumes, and distribution of new infarct comparing closure patients to those treated with medication alone. There were also no differences of number, volumes, and distribution comparing silent infarcts to clinical strokes. Infarct burden was also similar for patients who developed atrial fibrillation and for those with large shunts. Conclusions: The REDUCE trial demonstrates that patent foramen ovale closure prevents recurrent brain infarction based on the objective outcome of new infarcts on MRI. Only clinical strokes were reduced by closure while silent infarctions were similar between study arms, and there were no differences in infarct volume or location comparing silent infarcts to clinical strokes. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT00738894.
- Open Access
- 15 August 2021
- Published by MDPI AG in Applied Sciences
- Vol. 11 (16) , 7488
Interactive Machine Learning-Based Multi-Label Segmentation of Solid Tumors and Organs
We seek the development and evaluation of a fast, accurate, and consistent method for general-purpose segmentation, based on interactive machine learning (IML). To validate our method, we identified retrospective cohorts of 20 brain, 50 breast, and 50 lung cancer patients, as well as 20 spleen scans, with corresponding ground truth annotations. Utilizing very brief user training annotations and the adaptive geodesic distance transform, an ensemble of SVMs is trained, providing a patient-specific model applied to the whole image. Two experts segmented each cohort twice with our method and twice manually. The IML method was faster than manual annotation by 53.1% on average. We found significant (p < 0.001) overlap difference for spleen (DiceIML/DiceManual = 0.91/0.87), breast tumors (DiceIML/DiceManual = 0.84/0.82), and lung nodules (DiceIML/DiceManual = 0.78/0.83). For intra-rater consistency, a significant (p = 0.003) difference was found for spleen (DiceIML/DiceManual = 0.91/0.89). For inter-rater consistency, significant (p < 0.045) differences were found for spleen (DiceIML/DiceManual = 0.91/0.87), breast (DiceIML/DiceManual = 0.86/0.81), lung (DiceIML/DiceManual = 0.85/0.89), the non-enhancing (DiceIML/DiceManual = 0.79/0.67) and the enhancing (DiceIML/DiceManual = 0.79/0.84) brain tumor sub-regions, which, in aggregation, favored our method. Quantitative evaluation for speed, spatial overlap, and consistency, reveals the benefits of our proposed method when compared with manual annotation, for several clinically relevant problems. We publicly release our implementation through CaPTk (Cancer Imaging Phenomics Toolkit) and as an MITK plugin.
- 15 August 2021
- Published by MDPI AG in Applied Sciences
- Vol. 11 (16) , 7488
Open AccessInteractive Machine Learning-Based Multi-Label Segmentation of Solid Tumors and Organs
We seek the development and evaluation of a fast, accurate, and consistent method for general-purpose segmentation, based on interactive machine learning (IML). To validate our method, we identified retrospective cohorts of 20 brain, 50 breast, and 50 lung cancer patients, as well as 20 spleen scans, with corresponding ground truth annotations. Utilizing very brief user training annotations and the adaptive geodesic distance transform, an ensemble of SVMs is trained, providing a patient-specific model applied to the whole image. Two experts segmented each cohort twice with our method and twice manually. The IML method was faster than manual annotation by 53.1% on average. We found significant (p < 0.001) overlap difference for spleen (DiceIML/DiceManual = 0.91/0.87), breast tumors (DiceIML/DiceManual = 0.84/0.82), and lung nodules (DiceIML/DiceManual = 0.78/0.83). For intra-rater consistency, a significant (p = 0.003) difference was found for spleen (DiceIML/DiceManual = 0.91/0.89). For inter-rater consistency, significant (p < 0.045) differences were found for spleen (DiceIML/DiceManual = 0.91/0.87), breast (DiceIML/DiceManual = 0.86/0.81), lung (DiceIML/DiceManual = 0.85/0.89), the non-enhancing (DiceIML/DiceManual = 0.79/0.67) and the enhancing (DiceIML/DiceManual = 0.79/0.84) brain tumor sub-regions, which, in aggregation, favored our method. Quantitative evaluation for speed, spatial overlap, and consistency, reveals the benefits of our proposed method when compared with manual annotation, for several clinically relevant problems. We publicly release our implementation through CaPTk (Cancer Imaging Phenomics Toolkit) and as an MITK plugin.
- 1 July 2021
- Published by Radiological Society of North America (RSNA) in Radiology: Imaging Cancer
- Vol. 3 (4) , e200108
Combining MRI and Histologic Imaging Features for Predicting Overall Survival in Patients with Glioma
Combining features extracted from MRI with digital histopathologic features resulted in a higher prediction accuracy of overall survival in patients with glioma compared with using MRI or histopath... Purpose To test the hypothesis that combined features from MR and digital histopathologic images more accurately predict overall survival (OS) in patients with glioma compared with MRI or histopath...
- 1 July 2021
- Published by Radiological Society of North America (RSNA) in Radiology: Imaging Cancer
- Vol. 3 (4) , e200108
Combining MRI and Histologic Imaging Features for Predicting Overall Survival in Patients with Glioma
Combining features extracted from MRI with digital histopathologic features resulted in a higher prediction accuracy of overall survival in patients with glioma compared with using MRI or histopath... Purpose To test the hypothesis that combined features from MR and digital histopathologic images more accurately predict overall survival (OS) in patients with glioma compared with MRI or histopath...
- 4 December 2020
- Published by Wiley in Medical Physics
- Vol. 47 (12) , 6039-6052
Reproducibility analysis of multi‐institutional paired expert annotations and radiomic features of the Ivy Glioblastoma Atlas Project (Ivy GAP) dataset
Purpose: The availability of radiographic magnetic resonance imaging (MRI) scans for the Ivy Glioblastoma Atlas Project (Ivy GAP) has opened up opportunities for development of radiomic markers for prognostic/predictive applications in glioblastoma (GBM). In this work, we address two critical challenges with regard to developing robust radiomic approaches: (a) the lack of availability of reliable segmentation labels for glioblastoma tumor sub‐compartments (i.e., enhancing tumor, non‐enhancing tumor core, peritumoral edematous/infiltrated tissue) and (b) identifying “reproducible” radiomic features that are robust to segmentation variability across readers/sites.Acquisition and validation methods: From TCIA’s Ivy GAP cohort, we obtained a paired set (n = 31) of expert annotations approved by two board‐certified neuroradiologists at the Hospital of the University of Pennsylvania (UPenn) and at Case Western Reserve University (CWRU). For these studies, we performed a reproducibility study that assessed the variability in (a) segmentation labels and (b) radiomic features, between these paired annotations. The radiomic variability was assessed on a comprehensive panel of 11 700 radiomic features including intensity, volumetric, morphologic, histogram‐based, and textural parameters, extracted for each of the paired sets of annotations. Our results demonstrated (a) a high level of inter‐rater agreement (median value of DICE ≥0.8 for all sub‐compartments), and (b) ≈24% of the extracted radiomic features being highly correlated (based on Spearman’s rank correlation coefficient) to annotation variations. These robust features largely belonged to morphology (describing shape characteristics), intensity (capturing intensity profile statistics), and COLLAGE (capturing heterogeneity in gradient orientations) feature families.Data format and usage notes: We make publicly available on TCIA’s Analysis Results Directory (https://doi.org/10.7937/9j41‐7d44), the complete set of (a) multi‐institutional expert annotations for the tumor sub‐compartments, (b) 11 700 radiomic features, and (c) the associated reproducibility meta‐analysis.Potential applications: The annotations and the associated meta‐data for Ivy GAP are released with the purpose of enabling researchers toward developing image‐based biomarkers for prognostic/predictive applications in GBM.
- 4 December 2020
- Published by Wiley in Medical Physics
- Vol. 47 (12) , 6039-6052
Reproducibility analysis of multi‐institutional paired expert annotations and radiomic features of the Ivy Glioblastoma Atlas Project (Ivy GAP) dataset
Purpose: The availability of radiographic magnetic resonance imaging (MRI) scans for the Ivy Glioblastoma Atlas Project (Ivy GAP) has opened up opportunities for development of radiomic markers for prognostic/predictive applications in glioblastoma (GBM). In this work, we address two critical challenges with regard to developing robust radiomic approaches: (a) the lack of availability of reliable segmentation labels for glioblastoma tumor sub‐compartments (i.e., enhancing tumor, non‐enhancing tumor core, peritumoral edematous/infiltrated tissue) and (b) identifying “reproducible” radiomic features that are robust to segmentation variability across readers/sites.Acquisition and validation methods: From TCIA’s Ivy GAP cohort, we obtained a paired set (n = 31) of expert annotations approved by two board‐certified neuroradiologists at the Hospital of the University of Pennsylvania (UPenn) and at Case Western Reserve University (CWRU). For these studies, we performed a reproducibility study that assessed the variability in (a) segmentation labels and (b) radiomic features, between these paired annotations. The radiomic variability was assessed on a comprehensive panel of 11 700 radiomic features including intensity, volumetric, morphologic, histogram‐based, and textural parameters, extracted for each of the paired sets of annotations. Our results demonstrated (a) a high level of inter‐rater agreement (median value of DICE ≥0.8 for all sub‐compartments), and (b) ≈24% of the extracted radiomic features being highly correlated (based on Spearman’s rank correlation coefficient) to annotation variations. These robust features largely belonged to morphology (describing shape characteristics), intensity (capturing intensity profile statistics), and COLLAGE (capturing heterogeneity in gradient orientations) feature families.Data format and usage notes: We make publicly available on TCIA’s Analysis Results Directory (https://doi.org/10.7937/9j41‐7d44), the complete set of (a) multi‐institutional expert annotations for the tumor sub‐compartments, (b) 11 700 radiomic features, and (c) the associated reproducibility meta‐analysis.Potential applications: The annotations and the associated meta‐data for Ivy GAP are released with the purpose of enabling researchers toward developing image‐based biomarkers for prognostic/predictive applications in GBM.
- Open Access
- 10 July 2020
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 41 (9) , E78-E79
Neuro-Thoracic Radiologists “Corner”: Incidental Pulmonary Findings on a Neck MRI Leading to the Diagnosis of COVID-19
- 10 July 2020
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 41 (9) , E78-E79
Open AccessNeuro-Thoracic Radiologists “Corner”: Incidental Pulmonary Findings on a Neck MRI Leading to the Diagnosis of COVID-19
- Open Access
- 11 March 2020
- Published by MDPI AG in Applied Sciences
- Vol. 10 (6) , 1903
Multimodal Ensemble-Based Segmentation of White Matter Lesions and Analysis of Their Differential Characteristics across Major Brain Regions
White matter lesions (WML) are common in a variety of brain pathologies, including ischemia affecting blood vessels deeper inside the brain’s white matter, and show an abnormal signal in T1-weighted and FLAIR images. The emergence of personalized medicine requires quantification and analysis of differential characteristics of WML across different brain regions. Manual segmentation and analysis of WMLs is laborious and time-consuming; therefore, automated methods providing robust, reproducible, and fast WML segmentation and analysis are highly desirable. In this study, we tackled the segmentation problem as a voxel-based classification problem. We developed an ensemble of different classification models, including six models of support vector machine, trained on handcrafted and transfer learning features, and five models of Residual neural network, trained on varying window sizes. The output of these models was combined through majority-voting. A series of image processing operations was applied to remove false positives in a post-processing step. Moreover, images were mapped to a standard atlas template to quantify the spatial distribution of WMLs, and a radiomic analysis of all the lesions across different brain regions was carried out. The performance of the method on multi-institutional WML Segmentation Challenge dataset (n = 150) comprising T1-weighted and FLAIR images was >90% within data of each institution, multi-institutional data pooled together, and across-institution training–testing. Forty-five percent of lesions were found in the temporal lobe of the brain, and these lesions were easier to segment (95.67%) compared to lesions in other brain regions. Lesions in different brain regions were characterized by their differential characteristics of signal strength, size/shape, heterogeneity, and texture (p < 0.001). The proposed multimodal ensemble-based segmentation of WML showed effective performance across all scanners. Further, the radiomic characteristics of WMLs of different brain regions provide an in vivo portrait of phenotypic heterogeneity in WMLs, which points to the need for precision diagnostics and personalized treatment.
- 11 March 2020
- Published by MDPI AG in Applied Sciences
- Vol. 10 (6) , 1903
Open AccessMultimodal Ensemble-Based Segmentation of White Matter Lesions and Analysis of Their Differential Characteristics across Major Brain Regions
White matter lesions (WML) are common in a variety of brain pathologies, including ischemia affecting blood vessels deeper inside the brain’s white matter, and show an abnormal signal in T1-weighted and FLAIR images. The emergence of personalized medicine requires quantification and analysis of differential characteristics of WML across different brain regions. Manual segmentation and analysis of WMLs is laborious and time-consuming; therefore, automated methods providing robust, reproducible, and fast WML segmentation and analysis are highly desirable. In this study, we tackled the segmentation problem as a voxel-based classification problem. We developed an ensemble of different classification models, including six models of support vector machine, trained on handcrafted and transfer learning features, and five models of Residual neural network, trained on varying window sizes. The output of these models was combined through majority-voting. A series of image processing operations was applied to remove false positives in a post-processing step. Moreover, images were mapped to a standard atlas template to quantify the spatial distribution of WMLs, and a radiomic analysis of all the lesions across different brain regions was carried out. The performance of the method on multi-institutional WML Segmentation Challenge dataset (n = 150) comprising T1-weighted and FLAIR images was >90% within data of each institution, multi-institutional data pooled together, and across-institution training–testing. Forty-five percent of lesions were found in the temporal lobe of the brain, and these lesions were easier to segment (95.67%) compared to lesions in other brain regions. Lesions in different brain regions were characterized by their differential characteristics of signal strength, size/shape, heterogeneity, and texture (p < 0.001). The proposed multimodal ensemble-based segmentation of WML showed effective performance across all scanners. Further, the radiomic characteristics of WMLs of different brain regions provide an in vivo portrait of phenotypic heterogeneity in WMLs, which points to the need for precision diagnostics and personalized treatment.