Related Experiment Video
Updated: Aug 24, 2025
08:24
Author Spotlight: Exploring the Potential of Massage Therapy in Cerebral Palsy Using Animal Models
Published on: August 11, 2023
2.2K
Knowledge-Based Recurrent Neural Network for TCM Cerebral Palsy Diagnosis
Dongmei Li1,2, Jintao Qu1,2, Ziwei Tian1,2
1School of Information Science and Technology, Beijing Forestry University, Beijing 100083, China.
Evidence-Based Complementary and Alternative Medicine : Ecam
|October 24, 2022
Summary
Artificial intelligence (AI) enhances Traditional Chinese Medicine (TCM) for cerebral palsy diagnosis. A knowledge-based recurrent neural network (KBRNN) improved diagnostic accuracy by integrating TCM knowledge graphs with electronic medical records.
Area of Science:
- Neurology
- Artificial Intelligence
- Traditional Chinese Medicine
Background:
- Cerebral palsy is a leading cause of neurological disability.
- Traditional Chinese Medicine (TCM) diagnoses cerebral palsy based on symptom identification.
- Artificial intelligence (AI) offers potential for improving TCM diagnostic accuracy.
Purpose of the Study:
- To develop an AI decision-making model for cerebral palsy syndrome diagnosis.
- To integrate TCM domain knowledge with electronic medical records (EMRs) for enhanced diagnostic precision.
- To improve the accuracy and reliability of cerebral palsy diagnosis using AI.
Main Methods:
- Constructed a knowledge-based recurrent neural network (KBRNN) using a cerebral palsy knowledge graph.
- Employed an evolution algorithm to extract knowledge for the knowledge graph.
- Injected extracted TCM knowledge into the RNN and fine-tuned the model with labeled EMRs.
Main Results:
- Knowledge injection alone improved KBRNN diagnostic accuracy to 79.31%.
- Further training with EMRs increased the fully trained KBRNN accuracy to 83.12%.
- The KBRNN model demonstrated significant improvement in diagnostic performance.
Conclusions:
- Integrating domain knowledge (TCM) into AI models (RNN) enhances diagnostic accuracy.
- The KBRNN model provides a promising approach for AI-assisted TCM diagnosis of cerebral palsy.
- AI and knowledge graphs can bridge the gap between TCM principles and modern data analysis.
