现代信息科技2024，Vol.8Issue(7)：107-112,117,7.DOI:10.19850/j.cnki.2096-4706.2024.07.023
基于PH-GAT的精分患者分类预测模型研究
Research on a Classification Prediction Model for Schizophrenic Patients Based on PH-GAT
盛志林 1阴桂梅 1符永灿1
作者信息
- 1. 太原师范学院计算机科学与技术学院,山西 晋中 030619
- 折叠
摘要
Abstract
This paper studies the current analysis based on cerebral network,the study shows that the analysis methods can be broadly categorized into two main approaches:analysis based on continuous homotopy methods and analysis based on Deep Learning models.In order to enhance the predictive capabilities of brain disease diagnosis,this model incorporates continuous homotopy into the GAT model,endowing it with a"topological awareness".Towards the end of the model,the Long Short-Term Memory(LSTM)model is employed to capture temporal information embedded within the extracted features,thereby enhancing the effectiveness of classification prediction.Under the PH-GAT model,a fusion of local and global features is applied for classifying data in the Theta frequency range,achieving a high classification accuracy of 0.930 9.This approach not only enables the discovery of objective and effective imaging biomarkers for early schizophrenia diagnosis,but also enhances the predictive capabilities of brain disease diagnosis.关键词
脑网络/持续同调/图注意力网络/精神分裂症Key words
cerebral network/continuous homotopy/Graph Attention Network/schizophrenia分类
信息技术与安全科学引用本文复制引用
盛志林,阴桂梅,符永灿..基于PH-GAT的精分患者分类预测模型研究[J].现代信息科技,2024,8(7):107-112,117,7.
