A Bidirectional LSTM and Stochastic Fuzzy Systems for Improved Chronic Kidney Disease Prediction in IoMT-Based Robotic Automation

Improved Chronic Kidney Disease

Abstract

Chronic kidney disease (CKD) is a huge threat to health globally. For better patient prognoses, early diagnosis and prediction are necessary. In order to improve CKD prediction in Internet of Medical Things (IoMT)-based robotic systems, this study presents a novel hybrid model that incorporates stochastic fuzzy system and bidirectional long short term memory (Bi-LSTM). On the one hand, stochastic fuzzy systems manage uncertainty in medical data to improve decision-making; on the other, bi-LSTM is able to include information from data trends occurring either direction since they onset. Compared to traditional methods, for the detection of early CKD and real-time monitoring with 99% accuracy, 98% precision and 97% recall so it outperforms generic solutions.