ENHANCING HEALTHCARE DATA ACCESSIBILITY: A WEB-BASED LABORATORY TEST RESULT ANALYSIS SYSTEM

Abstract

Medical laboratory test results play a crucial role in healthcare by supporting disease diagnosis, treatment, monitoring, and prevention. However, the technical and complex nature of these results can be difficult for non-medical individuals to understand, often leading to misinterpretation, emotional distress, and difficulty in making informed health decisions. As a result, patients frequently require professional assistance from physicians, which may cause inconvenience and additional financial burden. This study aimed to address these challenges by developing a web-based laboratory test result analysis system to enhance healthcare data accessibility and interpretation. This study employed an experimental research design to assess the feasibility of the proposed system. The development process involved several stages, including the installation of necessary software and programs, integration of the EasyOCR model and GPT-3 API, and system implementation using HTML, CSS, Python, and JavaScript. The system’s efficacy and reliability were evaluated through multiple testing procedures, including comparative analysis between physician-provided interpretations and system-generated results, repeated trials to measure consistency, response time recording, and text box detection for assessing OCR performance. The findings showed that the system achieved a natural language processing (NLP) accuracy of 87.11% and a consistency rating of 92.85%. The average system response time was recorded at 52.06 seconds. Additionally, the optical character recognition (OCR) component demonstrated a high accuracy rate of 98.84%. The results indicated that the web-based laboratory test result analysis system performed favorably compared to findings from previous studies, demonstrating strong potential for improving healthcare data accessibility and interpretation. Although the system showed excellent overall performance, there remains room for enhancement. Future improvements may include the integration of alternative NLP and OCR models, as well as expanding the range of laboratory parameters covered by the system to further improve accuracy, efficiency, and usability in healthcare applications.