Deep learning-based multi-modal framework for integrating electronic health records, imaging, and genetic data for cardiovascular disease prediction, processes heterogeneous data sources to generate accurate risk assessments
	REDDY C P; MAHENDER G
	2025-03-24
专利权人	UNIV WARANGAL SR (UYWA-Non-standard)
申请日期	2025-03-24
专利号	IN202541026875-A
成果简介	NOVELTY - The framework has a framework for processing heterogeneous data sources to generate accurate risk assessments. A feature extraction mechanism utilizes convolutional neural networks (CNNs) and transformer models processes medical imaging and genetic sequences to enhance predictive performance. An explainable AI module enhances model interpretability using Shapley Additive Explanations (SHAP) and Gradient-weighted Class Activation Mapping (Grad-CAM) techniques. A secure cloud-based architecture employs federated learning for decentralized training across multiple healthcare institutions fo ensuring data privacy and security. A real-time risk assessment system is connected with an interactive dashboard for clinical decision support for visualizing key contributing factors and personalized risk scores. USE - Deep learning-based multi-modal framework for integrating EHR, imaging, and genetic data for cardiovascular disease prediction. ADVANTAGE - The framework realizes integration of multiple data modalities to enhance predictive accuracy and enable personalized treatment strategies and utilizes a deep learning model that harmonizes diverse data sources. The framework provides clinicians with actionable insights for enabling early intervention, thus reducing CVD-related mortality. The framework bridges gap between multi-modal data integration and AI-based prediction to revolutionize cardiovascular disease management. DESCRIPTION OF DRAWING(S) - The drawing shows a block diagram of a deep learning based multi-modal framework.
IPC 分类号	G06N-003/045 ; G06N-003/08 ; G06T-007/00 ; G16H-010/60 ; G16H-050/20
国家	印度
专业领域	信息技术
语种	英语
成果类型	专利
文献类型	科技成果
条目标识符	http://119.78.100.226:8889/handle/3KE4DYBR/13535
专题	中国科学院新疆生态与地理研究所
作者单位	UNIV WARANGAL SR (UYWA-Non-standard)
推荐引用方式 GB/T 7714	REDDY C P,MAHENDER G. Deep learning-based multi-modal framework for integrating electronic health records, imaging, and genetic data for cardiovascular disease prediction, processes heterogeneous data sources to generate accurate risk assessments. IN202541026875-A[P]. 2025.

条目包含的文件
条目无相关文件。

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。