Multi-Disease Prediction Website and Android Application Using Explainable Machine Learning and Deep learning Techniques
Date
2022
Student ID
Research Supervisor
Editor
Journal Title
Volume
Issue
Journal Title
Journal ISSN
Volume Title
Abstract
As the world advances towards a more computerized future, disruptive technologies are arriving at a pace second to no other innovation in history. The rise of computer vision and artificial intelligence techniques empowered innovations in the healthcare sector, which are helping to save lives, detect diseases, and extend life expectancy. This paper uses machine learning and deep learning techniques that fall into supervised learning to detect five diseases, e.g., diabetes, coronavirus, cardiac disorder, liver illness, and chronic kidney disease. Consequently, the automatic disease prediction system has been deployed on a website and an Android smartphone application. Initially, we enter necessary values from the pathology report or insert X-ray images into the website or the Android application to predict whether the person suffers from these diseases or not. The prediction is made in real time using machine learning models trained on various open-source datasets of different illnesses. We have used Random Forest as the supervised machine learning model as it produced an accuracy of 85%, 97%, 75%, and 85% to predict diabetes, cardiac disorder, liver illness, and chronic kidney disease, respectively. We have employed an attention-based CNN model, which produces a validation accuracy of 92% to predict coronavirus. Explainable: All models were used to interpret them.
Description
Keywords
Citation
Department Name
Electrical and Computer Engineering
Publisher
North South University