| Presentation | 2017-03-06 Recurrent Neural Networks for task-evoked fMRI data classification Koya Ohashi, Taiji Suzuki, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | We consider a classification problem in which the task that a subject is performing is identified from the brain activity data observed by fMRI. It has been shown that a classification method using FFNN (Feedforward Neural Network) achieved better classification accuracy than existing methods such as logistic regression. However, their method did not use temporal information and there was room for improvement in classification accuracy. In this study, we propose a classification method using RNN (Recurrent Neural Network) to incorporate temporal information. The proposed method adopts the idea of text classification method proposed in the field of natural language processing. Numerical experiments using real data confirmed that the proposed method achieves classification accuracy higher than those of existing methods. Furthermore, we show that through the sensitivity analysis of the learned task classifier, it is possible to estimate the part of the brain that has an important role for a specific task. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Deep learning / RNN / fMRI / Classification / Brain decoding / Brain machine interface |
| Paper # | IBISML2016-104 |
| Date of Issue | 2017-02-27 (IBISML) |
| Conference Information | |
| Committee | IBISML |
|---|---|
| Conference Date | 2017/3/6(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Tokyo Institute of Technology |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Statistical Mathematics, Machine Learning, Data Mining, etc. |
| Chair | Kenji Fukumizu(ISM) |
| Vice Chair | Masashi Sugiyama(Univ. of Tokyo) / Hisashi Kashima(Kyoto Univ.) |
| Secretary | Masashi Sugiyama(Univ. of Tokyo) / Hisashi Kashima(Nagoya Inst. of Tech.) |
| Assistant | Toshihiro Kamishima(AIST) / Tomoharu Iwata(NTT) |
| Paper Information | |
| Registration To | Technical Committee on Infomation-Based Induction Sciences and Machine Learning |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Recurrent Neural Networks for task-evoked fMRI data classification |
| Sub Title (in English) | |
| Keyword(1) | Deep learning |
| Keyword(2) | RNN |
| Keyword(3) | fMRI |
| Keyword(4) | Classification |
| Keyword(5) | Brain decoding |
| Keyword(6) | Brain machine interface |
| 1st Author's Name | Koya Ohashi |
| 1st Author's Affiliation | Tokyo Institute of Technology(Tokyo Tech) |
| 2nd Author's Name | Taiji Suzuki |
| 2nd Author's Affiliation | Tokyo Institute of Technology・PRESTO,Japan Science and Technorogy Agency/RIKEN(Tokyo Tech/JST/RIKEN) |
| Date | 2017-03-06 |
| Paper # | IBISML2016-104 |
| Volume (vol) | vol.116 |
| Number (no) | IBISML-500 |
| Page | pp.pp.33-40(IBISML), |
| #Pages | 8 |
| Date of Issue | 2017-02-27 (IBISML) |