講演抄録/キーワード |
講演名 |
2020-03-16 15:25
Assessing robustness of deep learning methods in dermoscopic workflow ○Sourav Mishra(Univ. of Tokyo)・Hideaki Imaizumi(exMedio)・Toshihiko Yamasaki(Univ. of Tokyo) PRMU2019-79 |
抄録 |
(和) |
(まだ登録されていません) |
(英) |
Our paper aims to evaluate current deep learning methods for clinical workflow in the domain of dermatology. Although deep learning methods have been successful in many cases, it has not been rigorously tested for common clinical complaints. Most projects involve data acquired in well-controlled laboratory conditions which may not reflect regular clinical evaluation. We test the robustness of deep learning methods by simulating non-ideal characteristics on user submitted images of ten classes of diseases. Assessing via imitated conditions, we have found the overall accuracy to drop and individual predictions change significantly in many cases despite of robust training. |
キーワード |
(和) |
/ / / / / / / |
(英) |
Deep Learning / Dermatology / Robustness / / / / / |
文献情報 |
信学技報, vol. 119, no. 481, PRMU2019-79, pp. 77-78, 2020年3月. |
資料番号 |
PRMU2019-79 |
発行日 |
2020-03-09 (PRMU) |
ISSN |
Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
PRMU2019-79 |