Kidney Disease Prediction Using Multiple LSTM Neural Networks

Over the last few years, artificial intelligence in healthcare has transformed everything regarding disease prediction and diagnosis. Artificial intelligence has proven useful in analyzing medical data and patient symptoms. Likewise, as this paper shows, artificial intelligence has been extremely accurate and effective at predicting kidney diseases, which are the leading causes of chronic kidney disease. Long-short-term memory was applied and implemented in this paper on a created dataset collected and organized manually from three local laboratories in Iraq. Five models were evaluated, including (long short-term memory, bidirectional long short-term memory, gated recurrent units, a combination of long short-term memory and bidirectional long short-term memory, and a hybrid model incorporating gated recurrent units and long short-term memory). It can carry out that the highest accuracy of 98% obtained from bidirectional long short-term memory, and the binary cross entropy loss is about 0.089.