| Presentation | 2019-01-23 Feature Selection from Imbalanced Data Hayato Itoh, Yuichi Mori, Masashi Misawa, Masahiro Oda, Shin-Ei Kudo, Kensaku Mori, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Endocytoscope gives ultramagnified observation that enables physicians to achieve minimally invasive and real-time diagnosis in colonoscopy. Since this modality is a quite new, a pathological image classifier is required for the support of non-expert physicians. In addition to the ununiformity of occurrence frequency of pathological patterns, data sampling by physicians includes bias. We have to handle imbalances of data to design an accurate pathological classifier. We propose feature-selection method that selects discriminative feature from imbalanced data for training of pathological classifier. We experimentally evaluated the proposed method by comparing the classification accuracy between before and after feature selection with about 50,000 endocytoscopic images. Our method achieves 2.7% improvement of classification accuracy with more accurate likelihood estimation than original texture features. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Endocytoscopy / automated pathological diagnosis / feature selection / texture feature / manifold learning / definite canonicalisation |
| Paper # | MI2018-87 |
| Date of Issue | 2019-01-15 (MI) |
| Conference Information | |
| Committee | MI |
|---|---|
| Conference Date | 2019/1/22(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Medical Image Engineering, Analysis, Recognition, etc. |
| Chair | Kensaku Mori(Nagoya Univ.) |
| Vice Chair | Yoshiki Kawata(Tokushima Univ.) / Yuichi Kimura(Kinki Univ.) |
| Secretary | Yoshiki Kawata(Aichi Inst. of Tech.) / Yuichi Kimura(Nagoya Inst. of Tech.) |
| Assistant | Ryo Haraguchi(Univ. of Hyogo) / Yasushi Hirano(Yamaguchi Univ.) |
| Paper Information | |
| Registration To | Medical Imaging |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Feature Selection from Imbalanced Data |
| Sub Title (in English) | Pathological Pattern Classification in Endocytoscopic Images |
| Keyword(1) | Endocytoscopy |
| Keyword(2) | automated pathological diagnosis |
| Keyword(3) | feature selection |
| Keyword(4) | texture feature |
| Keyword(5) | manifold learning |
| Keyword(6) | definite canonicalisation |
| 1st Author's Name | Hayato Itoh |
| 1st Author's Affiliation | Nagoya University(Nagoya Univ.) |
| 2nd Author's Name | Yuichi Mori |
| 2nd Author's Affiliation | Showa University(Showa Univ.) |
| 3rd Author's Name | Masashi Misawa |
| 3rd Author's Affiliation | Showa University(Showa Univ.) |
| 4th Author's Name | Masahiro Oda |
| 4th Author's Affiliation | Nagoya University(Nagoya Univ.) |
| 5th Author's Name | Shin-Ei Kudo |
| 5th Author's Affiliation | Showa University(Showa Univ.) |
| 6th Author's Name | Kensaku Mori |
| 6th Author's Affiliation | Nagoya University(Nagoya Univ.) |
| Date | 2019-01-23 |
| Paper # | MI2018-87 |
| Volume (vol) | vol.118 |
| Number (no) | MI-412 |
| Page | pp.pp.109-114(MI), |
| #Pages | 6 |
| Date of Issue | 2019-01-15 (MI) |