Predicting Insomnia Response to Acupuncture With the Development of Innovative Machine Learning

Insomnia is one of the most prevalent mental health disorders, and acupuncture has shown promise as an effective non-pharmacological treatment with minimal side effects. However, the response to acupuncture varies among individuals, emphasizing the need for personalized approaches to treatment. To a...

Full description

Saved in:
Bibliographic Details
Main Authors: Qingyun Wan, Kai Liu, Yuyang Bo, Xiya Yuan, Mufeng Li, Xiaoqiu Wang, Chuang Chen, Lanying Liu, Wenzhong Wu
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10918726/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Insomnia is one of the most prevalent mental health disorders, and acupuncture has shown promise as an effective non-pharmacological treatment with minimal side effects. However, the response to acupuncture varies among individuals, emphasizing the need for personalized approaches to treatment. To address this, an innovative machine learning algorithm, Relief-NDPGWO-WSVM, is developed to predict insomnia response to acupuncture. Data from 51 patients, considering 19 key factors such as age, sleep quality, anxiety level, and insomnia severity, were analyzed to identify the most influential predictors of treatment outcomes. The proposed model combines the Relief algorithm for feature selection, a weighted support vector machine (WSVM) to map these factors to treatment efficacy, and the NDPGWO optimization method, which incorporates a nonlinear convergence factor, dynamic weight, and probability perturbation. Experimental results demonstrate that this model outperforms existing models in terms of precision, recall, F1-score, accuracy, and area under the curve (AUC). These results highlight the model’s potential in accurately predicting acupuncture treatment outcomes for insomnia and suggest its broader applicability for enhancing clinical decision-making and optimizing healthcare resources for other related conditions.
ISSN:2169-3536