Title |
Residual Deep Convolutional Neural Network Predicts MGMT Methylation Status
|
---|---|
Published in |
Journal of Imaging Informatics in Medicine, August 2017
|
DOI | 10.1007/s10278-017-0009-z |
Pubmed ID | |
Authors |
Panagiotis Korfiatis, Timothy L. Kline, Daniel H. Lachance, Ian F. Parney, Jan C. Buckner, Bradley J. Erickson |
Abstract |
Predicting methylation of the O6-methylguanine methyltransferase (MGMT) gene status utilizing MRI imaging is of high importance since it is a predictor of response and prognosis in brain tumors. In this study, we compare three different residual deep neural network (ResNet) architectures to evaluate their ability in predicting MGMT methylation status without the need for a distinct tumor segmentation step. We found that the ResNet50 (50 layers) architecture was the best performing model, achieving an accuracy of 94.90% (+/- 3.92%) for the test set (classification of a slice as no tumor, methylated MGMT, or non-methylated). ResNet34 (34 layers) achieved 80.72% (+/- 13.61%) while ResNet18 (18 layers) accuracy was 76.75% (+/- 20.67%). ResNet50 performance was statistically significantly better than both ResNet18 and ResNet34 architectures (p < 0.001). We report a method that alleviates the need of extensive preprocessing and acts as a proof of concept that deep neural architectures can be used to predict molecular biomarkers from routine medical images. |
X Demographics
Geographical breakdown
Country | Count | As % |
---|---|---|
Netherlands | 1 | 25% |
United States | 1 | 25% |
Unknown | 2 | 50% |
Demographic breakdown
Type | Count | As % |
---|---|---|
Members of the public | 3 | 75% |
Practitioners (doctors, other healthcare professionals) | 1 | 25% |
Mendeley readers
Geographical breakdown
Country | Count | As % |
---|---|---|
Unknown | 152 | 100% |
Demographic breakdown
Readers by professional status | Count | As % |
---|---|---|
Student > Master | 18 | 12% |
Researcher | 14 | 9% |
Student > Bachelor | 14 | 9% |
Student > Doctoral Student | 13 | 9% |
Student > Ph. D. Student | 11 | 7% |
Other | 27 | 18% |
Unknown | 55 | 36% |
Readers by discipline | Count | As % |
---|---|---|
Medicine and Dentistry | 21 | 14% |
Computer Science | 20 | 13% |
Engineering | 18 | 12% |
Neuroscience | 6 | 4% |
Physics and Astronomy | 3 | 2% |
Other | 11 | 7% |
Unknown | 73 | 48% |