Current Result Document : 7 / 184
| 한글제목(Korean Title) |
딥러닝 기반의 MRI 영상에서의 직장암 병기 분류 |
| 영문제목(English Title) |
Rectal Cancer Classification using Deep Convolutional Neural Network |
| 저자(Author) |
오지은
안선희
최영상
김민희
김유성
손대경
김태성
Ji Eun Oh
Seon Hui Ahn
Young-sang Choi
Min Hee Kim
You-sung Kim
Dae Kyung Sohn
Tae-sung Kim
|
| 원문수록처(Citation) |
VOL 45 NO. 01 PP. 2525 ~ 2526 (2022. 06) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
In this study, we introduced a deep convolutional neural network (DCNN) model to predict T-staging of rectal cancer. The experiments were performed using T2-weighted MR images of 567 patients with histologically confirmed rectal cancer. The patients consisted of 168 patients with T2 tumor stage and 399 patients with T3 tumor stage. The MR images were obtained from the National Cancer Center (NCC) in the Republic of Korea. We used the ResNet18 architecture which is a Residual Network (ResNet) variant composed of 18 layers as backbone. A DCNN-based model achieved an AUC of 68.60±7.27 (with an accuracy of 64.17 ± 12.42). By applying the rectum cropping and intensity correction, the AUC and accuracy were more increased to 0.7960 ± 0.0508 and 76.62 ± 3.11. The trained model with preprocessed images were statistically significantly outperformed.
|
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
