AI-Powered Liver Disease Prediction Using ANN and Mobilenetv2

Resumen

Liver disease is a serious health issue affecting people worldwide, and detecting it at an early stage is essential to avoid severe complications and improve survival rates. In recent years, Artificial Intelligence (AI) and Machine Learning (ML) have shown great potential in supporting medical diagnosis by providing faster and more accurate results. This paper presents an AI-based system for predicting liver disease using both blood test data and medical images. For analyzing structured clinical data, an Artificial Neural Network (ANN) is used to evaluate important biochemical parameters related to liver function. For image-based analysis, a deep learning model based on MobileNetV2 is applied to classify different stages of liver disease. The proposed system was trained and evaluated using approximately 30,000 clinical records and 6,323 liver ultrasound images. Before training the models, the data is processed through several steps such as handling missing values, converting categorical data, normalizing features, and dividing the dataset into training and testing sets. The complete system is developed as a web application using the Flask framework, which allows users to input data, obtain real-time prediction results, and generate reports. Experimental results show that the ANN model achieved 94% accuracy, while the MobileNetV2 model achieved 92% accuracy in liver disease prediction. This system can be useful in hospitals, diagnostic centers, and telemedicine platforms for early screening and continuous monitoring of liver disease patients. It helps doctors make quicker and more reliable decisions, ultimately improving patient care.