| 諏訪部 開 | D, 中間発表 | ユビキタスコンピューティングシステム | 安本 慶一, 池田 和司, 荒川 豊, 諏訪 博彦 |
|
title: Investigating inter-subjects generalization performance in cognitive load estimation using biological information
abstract: Studies on behavior change by imposing some tasks on subjects are effective to solve problems in education and health areas. However, when the skill of the subject and the challenge of the subject's task do not have a good match, the subject may not be able to concentrate on the task and its utility is diminished. Although previous studies tried to solve the problem by estimating cognitive load from subject's biological information using classifiers constructed for each subject, they require a pre-trained personalized recognition model. In this work, we aim to overcome this restriction through inter-subjects generalization. For this purpose, we constructed a inter-subject model for recognizing cognitive workload with N-back problem and evaluated its performance. In this presentation, we report the current status and show how we can improve the performance from the viewpoint of features selection. language of the presentation: Japanese | |||
| 栗田 侑樹 | M, 2回目発表 | 生体医用画像 | 佐藤 嘉伸, 池田 和司, 大竹 義人, Soufi Mazen |
|
title: Automated Kinematic Analysis of Sacroiliac Joint Using Upright MRI
abstract: Three-dimensional evaluation of musculoskeletal anatomy in standing position is important because many musculoskeletal disorders provoke symptoms under the load of gravity. Therefore, we aim to construct an automated method to evaluate musculoskeletal dynamics of physiological and pathological motion using upright MRI, G-scan brio (Esaote Spa, Italy), that can obtain 3D-MR images while the subject is in standing position. In this study, we focused on the movement of sacroiliac joint. language of the presentation: Japanese | |||