TY - JOUR
T1 - MRI-based identification and classification of major intracranial tumor types by using a 3D convolutional neural network
T2 - A retrospective multi-institutional analysis
AU - Chakrabarty, Satrajit
AU - Sotiras, Aristeidis
AU - Milchenko, Mikhail
AU - Lamontagne, Pamela
AU - Hileman, Michael
AU - Marcus, Daniel
N1 - Funding Information:
Disclosures of Conflicts of Interest: S.C. institution received grant from National Institutes of Health. A.S. author has stocks in TheraPanacea, which is a startup in France founded by PhD supervisor. TheraPanacea develops a medical analysis platform intended to improve cancer treatment. This platform leverages artificial intelligence technology to enable medical practitioners to treat patients with cancer with radiation therapy with improved success and lesser risks. M.M. institution received grant from NIH, P30 NS048056, NINDS Center Core for Brain Imaging (NCCBI). The NCCBI provides informatics, analysis, and imaging methodology service and consultation to the Washington University neuroimaging community; author employed by Washington University, St Louis as a research instructor in the department of radiology. M.L. employed by Washington University Medical School. M.H. institution received grant from National Institutes of Health. D.M. institution received grant from National Institutes of Health.
Funding Information:
D.M. is supported by the National Institutes of Health (grants P30 NS098577, U24 CA204854, and R01 EB009352).
Publisher Copyright:
© RSNA, 2021.
PY - 2021/9
Y1 - 2021/9
N2 - Purpose: To develop an algorithm to classify postcontrast T1-weighted MRI scans by tumor classes (high-grade glioma, low-grade glioma [LGG], brain metastasis, meningioma, pituitary adenoma, and acoustic neuroma) and a healthy tissue (HLTH) class. Materials and Methods: In this retrospective study, preoperative postcontrast T1-weighted MR scans from four publicly available da-tasets—the Brain Tumor Image Segmentation dataset (n = 378), the LGG-1p19q dataset (n = 145), The Cancer Genome Atlas Glioblastoma Multiforme dataset (n = 141), and The Cancer Genome Atlas Low Grade Glioma dataset (n = 68)—and an internal clinical dataset (n = 1373) were used. In all, a total of 2105 images were split into a training dataset (n = 1396), an internal test set (n = 361), and an external test dataset (n = 348). A convolutional neural network was trained to classify the tumor type and to discriminate between images depicting HLTH and images depicting tumors. The performance of the model was evaluated by using cross-validation, internal testing, and external testing. Feature maps were plotted to visualize network attention. The accuracy, positive predictive value (PPV), negative predictive value, sensitivity, specificity, F1 score, area under the receiver operating characteristic curve (AUC), and area under the precision-recall curve (AUPRC) were calculated. Results: On the internal test dataset, across the seven different classes, the sensitivities, PPVs, AUCs, and AUPRCs ranged from 87% to 100%, 85% to 100%, 0.98 to 1.00, and 0.91 to 1.00, respectively. On the external data, they ranged from 91% to 97%, 73% to 99%, 0.97 to 0.98, and 0.9 to 1.0, respectively. Conclusion: The developed model was capable of classifying postcontrast T1-weighted MRI scans of different intracranial tumor types and discriminating images depicting pathologic conditions from images depicting HLTH.
AB - Purpose: To develop an algorithm to classify postcontrast T1-weighted MRI scans by tumor classes (high-grade glioma, low-grade glioma [LGG], brain metastasis, meningioma, pituitary adenoma, and acoustic neuroma) and a healthy tissue (HLTH) class. Materials and Methods: In this retrospective study, preoperative postcontrast T1-weighted MR scans from four publicly available da-tasets—the Brain Tumor Image Segmentation dataset (n = 378), the LGG-1p19q dataset (n = 145), The Cancer Genome Atlas Glioblastoma Multiforme dataset (n = 141), and The Cancer Genome Atlas Low Grade Glioma dataset (n = 68)—and an internal clinical dataset (n = 1373) were used. In all, a total of 2105 images were split into a training dataset (n = 1396), an internal test set (n = 361), and an external test dataset (n = 348). A convolutional neural network was trained to classify the tumor type and to discriminate between images depicting HLTH and images depicting tumors. The performance of the model was evaluated by using cross-validation, internal testing, and external testing. Feature maps were plotted to visualize network attention. The accuracy, positive predictive value (PPV), negative predictive value, sensitivity, specificity, F1 score, area under the receiver operating characteristic curve (AUC), and area under the precision-recall curve (AUPRC) were calculated. Results: On the internal test dataset, across the seven different classes, the sensitivities, PPVs, AUCs, and AUPRCs ranged from 87% to 100%, 85% to 100%, 0.98 to 1.00, and 0.91 to 1.00, respectively. On the external data, they ranged from 91% to 97%, 73% to 99%, 0.97 to 0.98, and 0.9 to 1.0, respectively. Conclusion: The developed model was capable of classifying postcontrast T1-weighted MRI scans of different intracranial tumor types and discriminating images depicting pathologic conditions from images depicting HLTH.
UR - http://www.scopus.com/inward/record.url?scp=85117844619&partnerID=8YFLogxK
U2 - 10.1148/ryai.2021200301
DO - 10.1148/ryai.2021200301
M3 - Article
C2 - 34617029
AN - SCOPUS:85117844619
SN - 2638-6100
VL - 3
JO - Radiology: Artificial Intelligence
JF - Radiology: Artificial Intelligence
IS - 5
M1 - e200301
ER -