Odontología · 2026

Desarrollo de una aplicación por medio de inteligencia artificial para la extracción de datos de historias clínicas en el área de periodoncia en la UAN sede Armenia 2020-2025

Resumen

The objective of this study was the AI-assisted design and implementation of an application whose primary purpose was the automation of data extraction and analysis from Periodontics clinical records at the Universidad Antonio Nariño, Armenia campus. Utilizing clinical records from patients treated at the university clinic between 2020 and 2025, a workflow was structured to integrate data processing from PDF files with machine learning techniques. This allowed for an efficient transition from isolated clinical records to a structured database ready for statistical analysis. In the initial phase, development was conducted within the Google Colab environment, where proof-of-concept testing and validation of the extraction algorithms were performed. This stage was fundamental for debugging the programming logic and ensuring that the Python scripts interacted accurately with the information. Subsequently, to comprehensively improve, review, and unify the cleaning, debugging, and data structuring functions, the project concluded its development phase within the Google Antigravity environment. This transition enhanced the data extraction application through a more advanced development ecosystem, optimizing the software’s responsiveness to the volume and complexity of the clinical data. The information management process began by consolidating periodontal disease data into a questionnaire format using spreadsheet tools. Once the information was unified, it was converted into PDF format to serve as the primary source for the data extraction engine. Through this approach, unstructured clinical documentation was successfully transformed into an analytical database under an observational and retrospective focus. The implementation of this tool not only facilitates the acquisition of precise statistical indicators—such as the association of variables with periodontal disease—but also sets a precedent at the UAN Clinic regarding the use of AI for knowledge management and the improved interpretation of oral health epidemiological profiles. It is concluded that the integration of artificial intelligence and machine learning techniques into clinical record management represents a significant advancement for the UAN Armenia Clinic. The efficacy of the developed software allows for the transformation of retrospective records into structured data of high analytical value, mitigating human error and maximizing information quality. Beyond technical processing, this system serves as a clinical-strategic support tool that empowers decision-making and drives the development of preventive protocols tailored to the epidemiological reality of the population treated at the clinic, based on the analysis for the 2020– 2025 period.

Palabras clave