Data Science in Pre-diagnosis to Aid Decision-Making in Teleconsultations at CISAM

Abstract

The overburdened healthcare systems and the difficulty of accessing care are global challenges. Telehealth, supported by Data Science, emerges as a viable solution, facilitating assistance. This work, developed in partnership with NUTES/CISAM, proposes an integrated Data Mining solution to optimize the flow of teleconsultations. The central objective is to develop a unified platform for the secure storage of patient data and the creation of a dashboard to support strategic decision-making. The methodology combined CRISP-DM for data flow and SCRUM/Kanban for project management. Data were extracted from heterogeneous sources (spreadsheets, medical records, and imaging exams) and centralized in a PostgreSQL database with MinIO. The main results are an interactive dashboard that reveals metrics and bottlenecks in care, and a web interface for exam management, aiming to improve the efficiency and assertiveness of medical decision-making.