仲思智慧中医平台

Search documents
如同在足球场找到一粒特定纹路沙子
Nan Fang Du Shi Bao· 2025-07-23 23:08
Core Insights - The Guangdong Provincial Health Commission has released the first batch of "AI + Healthcare" application scenarios, featuring 289 cases, with Shenzhen's Bao'an District leading with 30 cases, including five from Bao'an District People's Hospital [2][7] - The AI-assisted diagnostic systems in Bao'an District have significantly improved diagnostic accuracy, reducing misdiagnosis rates by 32% and enhancing efficiency in medical processes [3][4] Group 1: AI Applications in Healthcare - Bao'an District People's Hospital's AI-assisted diagnostic system utilizes a "pathological feature knowledge base" to identify subtle morphological changes, achieving a misdiagnosis rate reduction of 32% [3] - The imaging AI system employs a "multi-modal imaging fusion algorithm" to quickly locate lesions within 0.3 seconds, enhancing diagnostic transparency and efficiency by four times [3] - The AI medical assistant provides personalized recommendations based on patient-specific contexts, transforming the process of medical record generation and significantly reducing the time required for documentation from 1-2 hours to 30 minutes [4] Group 2: Predictive Health Management - The multi-modal health warning system analyzes various health data to provide predictive interventions, reducing the acute attack rate of hypertension patients by 57% [5] - The intelligent health assistant interprets health check reports efficiently, acting as a proactive health manager [5] Group 3: Innovations in Gastrointestinal Preparation - The "multi-modal AI-based intestinal preparation quality detection platform" at Fuyong People's Hospital enhances the preparation process for patients undergoing colonoscopy, addressing traditional pain points and improving communication between patients and healthcare providers [7] Group 4: Traditional Chinese Medicine Integration - The "Zhongsi" smart TCM platform developed by Guangzhong Medicine Shenzhen Hospital integrates over 380 million medical instructions and 60,000 TCM-specific instructions, significantly improving service efficiency and quality in traditional Chinese medicine [8][9]
深圳 “仲思”智慧中医平台入选广东“AI+医疗”典型案例
Nan Fang Du Shi Bao· 2025-07-21 13:17
Core Insights - The Guangdong Provincial Health Commission has announced the first batch of "Artificial Intelligence + Healthcare" application scenarios, with the "Zhongsi" Smart Traditional Chinese Medicine Platform from Guangzhong Medicine Shenzhen Hospital being selected for provincial promotion, indicating a significant achievement in the integration of traditional Chinese medicine and artificial intelligence [1] Group 1 - The "Zhongsi" platform is based on a large language model with 34 billion parameters, utilizing over 3.8 million high-quality medical instruction data and 60,000 detailed traditional Chinese medicine instructions, integrating technologies such as ASR voice recognition and knowledge graphs [1] - The platform enables real-time transcription of doctor-patient dialogue, automatic extraction of patient symptoms and examination indicators, and ensures medical record quality through an intelligent error correction mechanism [1] - The platform has been deployed in 31 community health institutions and one community hospital under Guangzhou University of Traditional Chinese Medicine Shenzhen Hospital, serving 200 clinical physicians with a total consultation volume of 5,000, improving community health service efficiency by 20% [2] Group 2 - The platform analyzes and integrates multimodal diagnostic data to provide comprehensive diagnostic support for doctors, enhancing follow-up efficiency and achieving a closed-loop management of "diagnosis - follow-up" [2] - Features such as intelligent inquiry recommendations, disease differentiation, prescription recommendations, and care plan suggestions are available to assist physicians during consultations [2] - After consultations, the system automatically synchronizes diagnostic data and health reports to the platform, supporting the intelligent follow-up system and gradually building a local data pool [2]