التفاصيل البيبلوغرافية
| العنوان: |
MRI-Based Deep Learning Method for Classification of IDH Mutation Status. |
| المؤلفون: |
Bangalore Yogananda, Chandan Ganesh1 (AUTHOR) ben.wagner@utsouthwestern.edu, Wagner, Benjamin C.1 (AUTHOR) nghi.truong@utsouthwestern.edu, Truong, Nghi C. D.1 (AUTHOR) james.holcomb@utsouthwestern.edu, Holcomb, James M.1 (AUTHOR) divya.reddy@utsouthwestern.edu, Reddy, Divya D.1 (AUTHOR) niloufar.saadat@utsouthwestern.edu, Saadat, Niloufar1 (AUTHOR) bfei@utdallas.edu, Hatanpaa, Kimmo J.2 (AUTHOR) kimmo.hatanpaa@utsouthwestern.edu, Patel, Toral R.3 (AUTHOR) toral.patel@utsouthwestern.edu, Fei, Baowei1,4 (AUTHOR) marco.pinho@utsouthwestern.edu, Lee, Matthew D.5 (AUTHOR) matthew.lee4@nyulangone.org, Jain, Rajan5,6 (AUTHOR) rajan.jain@nyulangone.org, Bruce, Richard J.7 (AUTHOR) rbruce@uwhealth.org, Pinho, Marco C.1 (AUTHOR) ananth.madhuranthakam@utsouthwestern.edu, Madhuranthakam, Ananth J.1 (AUTHOR) joseph.maldjian@utsouthwestern.edu, Maldjian, Joseph A.1 (AUTHOR) |
| المصدر: |
Bioengineering (Basel). Sep2023, Vol. 10 Issue 9, p1045. 14p. |
| مصطلحات موضوعية: |
*DEEP learning, *MACHINE learning, *RECEIVER operating characteristic curves, *DATABASES, *ISOCITRATE dehydrogenase, *BRAIN tumors |
| الشركة/الكيان: |
UNIVERSITY of Wisconsin (Madison, Wis.) , NEW York University , UNIVERSITY of California, San Francisco |
| مستخلص: |
Isocitrate dehydrogenase (IDH) mutation status has emerged as an important prognostic marker in gliomas. This study sought to develop deep learning networks for non-invasive IDH classification using T2w MR images while comparing their performance to a multi-contrast network. Methods: Multi-contrast brain tumor MRI and genomic data were obtained from The Cancer Imaging Archive (TCIA) and The Erasmus Glioma Database (EGD). Two separate 2D networks were developed using nnU-Net, a T2w-image-only network (T2-net) and a multi-contrast network (MC-net). Each network was separately trained using TCIA (227 subjects) or TCIA + EGD data (683 subjects combined). The networks were trained to classify IDH mutation status and implement single-label tumor segmentation simultaneously. The trained networks were tested on over 1100 held-out datasets including 360 cases from UT Southwestern Medical Center, 136 cases from New York University, 175 cases from the University of Wisconsin–Madison, 456 cases from EGD (for the TCIA-trained network), and 495 cases from the University of California, San Francisco public database. A receiver operating characteristic curve (ROC) was drawn to calculate the AUC value to determine classifier performance. Results: T2-net trained on TCIA and TCIA + EGD datasets achieved an overall accuracy of 85.4% and 87.6% with AUCs of 0.86 and 0.89, respectively. MC-net trained on TCIA and TCIA + EGD datasets achieved an overall accuracy of 91.0% and 92.8% with AUCs of 0.94 and 0.96, respectively. We developed reliable, high-performing deep learning algorithms for IDH classification using both a T2-image-only and a multi-contrast approach. The networks were tested on more than 1100 subjects from diverse databases, making this the largest study on image-based IDH classification to date. [ABSTRACT FROM AUTHOR] |
| قاعدة البيانات: |
Academic Search Index |
Full Text Finder