This item is licensed Korea Open Government License
Title
정신질환 판별 지원을 위한 기계학습 기술 개발
Alternative Title
Machine learning technology development for diagnosis of psychiatric disorders
Publisher
한국과학기술정보연구원 Korea Institute of Science and Technology Information
Publication Year
2017-01
Description
funder : 미래창조과학부 funder : KA
Abstract
본 과제는 정신질환 환자의 뇌파 데이터를 처리하는 뇌파 분석 기술을 확보하고, 기계학습 및 다양한 통계기술을 사용하여 뇌파 데이터 기반 정신질환 판별 지원 기술 및 시스템을 개발하는 데에 목적을 둔다.
○ 정신질환 뇌파 데이터 수집
○ 정신질환 뇌파 데이터 분석을 위한 파이프라인 구축
○ 기계학습 기술 기반 정신질환 판별기술 연구 및 개발
Ⅳ. Results of the research
○ Collect psychiatric disorders’ EEG data
Collect two independent EEG datasets related to alcohols (alcohol drinking data and alcoholism data)
Collect clinical and meta-data for the patients for the EEG
○ Discovered EEG-features related to alcohols Identified
Theta-Gamma coupling and gamma reduction after alcohol drinking condition
Discovered common EEG-features between alcohol drinking and alcoholism features (gamma waves in parietal lobe)
Identfied the increasing features in increasing of alcohol-use dependence
○ Developed discriminative models for alcoholism based on machine learning
Developed the scoring method for the severity of alcohol-use-dependence
Developed deep-learning models (DBN,CNN) for alcoholism classifications