Automated Stroke Prediction and Prevention Recommendations: Development of an Android Application

Stroke is currently one of the leading causes of mortality worldwide. Extensive research has identified several risk factors associated with stroke, which have been extensively studied to enhance prediction and categorization of the disease. Machine learning (ML) has proven to be a powerful tool for analyzing vast amounts of data, enabling accurate predictions and informed decision-making. Researchers are actively working to develop automated ML models for stroke prediction with the aim of facilitating early interventions and saving lives. As the global population ages and the number of individuals at risk for stroke continues to rise, the need for accurate prediction algorithms has become increasingly critical. The widespread adoption of Android applications provides an opportunity to make predictive tools accessible to a broader audience. This study investigates the application of ML algorithms in stroke risk prediction and demonstrates their integration into a functional Android application. The dataset used for model development was sourced from Kaggle, where a significant imbalance was observed with a ratio of 19:1 between instances of no stroke and stroke. To address this imbalance, Synthetic Minority Oversampling Technique (SMOTE) analysis was employed, ensuring a balanced dataset for training the models. Eight robust ML algorithms were utilized to develop predictive models. Among these, the Ensemble method leveraging a voting classifier achieved the best performance, attaining an accuracy of 95% and a recall of 95%. TensorFlow was used to integrate the machine learning model into an Android application, enabling real-time predictions. The Android app, developed in Java and built using the Android Studio platform, is designed to offer recommendations for stroke prevention and management. This user-friendly application aims to enhance accessibility to stroke risk predictions, empowering users to take proactive measures to safeguard their health..