Automated Risk Prediction of Post-Stroke Adverse Mental Outcomes Using Deep Learning Methods and Sequential Data

• DOI: 10.3390/bioengineering12050517
• Publication Date: 2025-05-14T14:27:41Z
• Authors: Chien Wei Oei; Eddie Yin Kwee Ng; Matthew Hok Shan Ng; Yam Meng Chan; Vinithasree Subbh...; Lai Gwen Chan; U. Rajendra Acharya
• Keywords: Stroke

Depression and anxiety are common comorbidities of stroke. Research has shown that about 30% stroke survivors develop depression 20% anxiety. Stroke with such adverse mental outcomes often attributed to poorer health outcomes, as higher mortality rates. The objective this study is use deep learning (DL) methods predict the risk a survivor experiencing post-stroke and/or anxiety, which collectively known (PSAMO). This studied 179 patients stroke, who were further classified into PSAMO versus no group based on results validated questionnaires, industry's gold standard. collected demographic sociological data, quality life scores, stroke-related information, medical medication history, comorbidities. In addition, sequential data daily lab taken seven consecutive days after admission also collected. combination using DL algorithms, multi-layer perceptron (MLP) long short-term memory (LSTM), can process complex patterns in inclusion new types, helped improve model performance. Accurate prediction helps clinicians make early intervention care plans potentially reduce incidence PSAMO.