Advanced Kidney Disease Classification Using Moth Fly Optimized Deep Convolutional Neural Network

Abstract

Kidney disease significantly impairs kidney function, making timely diagnosis essential to prevent progression. Artificial Intelligence (AI) has transformed healthcare, enabling precise disease classification. However, a key challenge in kidney disease diagnosis lies in processing data effectively for accurate results. This research introduces a novel technique for classification of kidney disease utilizing optimized neural network classification. The process initiates by preprocessing of input kidney dataset using Adaptive Notch Filter (ANF), for elimination of noise and to preserve critical information needed for classification, then it is segmented by Fuzzy C- Means algorithm that enhances the identification of different regions within the kidney. Next, structural and texture features are extracted through the Histogram of Oriented Gradients (HOG). At last, classification is done by Moth Fly Optimized Deep Convolutional Neural Network (DCNN), which improves the efficiency and accuracy of classification process by optimizing the parameters of network for better convergence.