Análise preditiva baseada em inteligência artificial: para a vigilância de doenças raras.

Abstract

This doctoral dissertation titled "Predictive Analysis Based on Artificial Intelligence: For the Surveillance of Rare Diseases" presents a comprehensive study developed by Paula Lopes Alvim Santilli for the Master's program in Strategic Direction of Information Technologies. The research addresses the critical challenges in diagnosing and monitoring rare diseases through the implementation of artificial intelligence and predictive analytics. The study explores how digital health technologies, particularly AI-driven predictive models, can transform the landscape of rare disease surveillance and patient care. Key Components: Research Focus: The dissertation investigates the application of predictive medicine and AI algorithms to improve early detection, diagnosis, and continuous monitoring of rare diseases. It examines the integration of synthetic data generation, natural language processing, and machine learning models to overcome the inherent data scarcity challenges in rare disease research. Technological Framework: The study presents a comprehensive technological architecture including: A mobile health application (DR Health) developed in FlutterFlow for patient data collection Cloud-based data processing infrastructure utilizing N8N workflows Integration of statistical APIs and generative AI models RAG (Retrieval-Augmented Generation) architecture for clinical decision support Interoperability standards including FHIR (Fast Healthcare Interoperability Resources) Methodology: The research employs a mixed-methods approach combining: Literature review of international and national experiences in predictive platforms Development of synthetic patient data for model training Implementation of machine learning algorithms for risk prediction Testing scenarios comparing statistical APIs with AI-enhanced approaches Privacy and security frameworks aligned with LGPD (Brazilian General Data Protection Law) Clinical Applications: The dissertation covers various aspects of rare disease management including: Daily symptom and vital signs monitoring Examination and hospitalization tracking Adverse event surveillance Personalized clinical alerts and recommendations Integration with healthcare provider workflows Ethical and Legal Framework: Comprehensive analysis of: Data privacy and security in digital health Patient consent management Civil and medical liability with AI systems Compliance with Brazilian and international healthcare regulations Data minimization and pseudonymization strategies Expected Outcomes: The research aims to demonstrate how AI-powered predictive analytics can reduce diagnostic delays, improve patient outcomes, enable early intervention, and support healthcare professionals in managing rare disease patients through evidence-based, data-driven approaches. Keywords: Predictive Medicine, Rare Diseases, Artificial Intelligence, Synthetic Data, Personalization, Digital Health This dissertation contributes to the growing field of digital health transformation by providing both theoretical foundations and practical implementation frameworks for AI-based rare disease surveillance systems, with particular relevance to the Brazilian healthcare context and global health challenges.

Ähnliche Arbeiten