Abstract: chronic kidney disease (CKD) is a long-term condition characterized by the gradual loss of kidney function over time. The kidneys are essential organs that filter waste and extra fluid from the blood so that it can be expelled as urine. Early detection is crucial for managing progression and associated complications. However, manual detection of kidney disease is time-consuming and challenging task in the current setup. In this work, a novel deep learning-based CKD-DBN model has been proposed for identifying CKD in its early stages. Initially, the CT images are gathered from the publicly available dataset and these gathered images are pre-processed using median filter to eliminate noise while preserving edges in the image. The pre-processed images are fed into the GhostNet for feature extraction that extracting significant features from the input images. Finally Deep Belief Network (DBN) classifies the images into Normal or CKD based on these features. The proposed CKD-DBN model is evaluated based on its f1 score, recall, specificity, precision and accuracy. The classification accuracy of 98.82% for the proposed are highly reliable for public available dataset. The proposed CKD-DBN model achieves the overall accuracy by 1.84%, 0.82% and 2.85% comparing to the existing method such as ALO, DNN and RNN

Abstract: Premature newborns cry to let their parents and others know what they need. It is essential to remember that premature screamers may scream for a variety of causes. Based on weeping, parents can identify a baby's emotional and physical changes and needs. However, quite challenging to pinpoint the requirement for the incubated newborns that have jaundice. To overcome these challenges, a novel deep learning-based CLJFS model is proposed for Cry-based Jaundiced Infant Signal (CLJFS) classification model. The crying is a newborn's primary communication, understanding its acoustic features can provide critical insights into the infant's condition. The proposed CLJFS model employs a multi-step process beginning with signal pre-processing using Stationary Wavelet Transform (SWT) for noise reduction and feature enhancement. Linear Prediction Coefficients (LPC) are extracted, followed by feature selection using the Least Absolute Shrinkage and Selection Operator (LASSO). A Spiking Neural Network (SNN) then categorizes the cries into three classes: hunger, fear, and discomfort. The effectiveness of the proposed CLJFS was evaluate using F1 score, accuracy, precision, recall, and specificity. The proposed CLJFS model achieved a classification accuracy 98.9%. The proposed model enhanced the total accuracy by 2.24%, 4.03%, 10.1%, and 7.03%, respectively.

Abstract: Tuberculosis (TB) is among the most common communicable diseases caused by a bacterial infection namely Mycobacterium tuberculosis. The radiologists spend more time for detecting the TB when analysing with traditional methods. However, manual detection of TB is time-consuming and challenging task in the current scenario. To overcome these challenges, a novel deep learning-based TOUR-STAR model has been proposed for classifying the TB into three classes. Initially, the chest x-ray (CXR) images are gathered from the Pad Chest dataset and the collected images are pre-processed by Gaussian filter for reducing the noises and smoothen the edges. The Dual Attention U-Network is used to segment the liver region separately from the pre-processed CXR images. Finally, the segmented images are fed into spiking dilated convolutional neural network (SDCNN) to classify the TB into tri classes such as normal controls (NC), Pulmonary Tuberculosis (PT) and Miliary Tuberculosis (MT). The proposed TOUR-STAR model is evaluated based on its f1 score, precision, specificity, recall and accuracy. The classification accuracy of 98.90% for the proposed SDCNN are highly reliable for Pad Chest dataset. The proposed SDCNN outperforms ResNet, AlexNet, RegNet, and GoogleNet by 5.30%, 1.63%, 4.49% and 0.83% in terms of overall accuracy range. The suggested TOUR-STAR model achieves the overall accuracy by 0.20%, 0.76% and 4.14% comparing to the existing method such as CNN, DNN and EfficientNet.

Abstract: Huntington disease is a neurodegenerative genetic condition brought on by an HTT gene mutation that produces a toxic protein that damages neurons. This progressive disease primarily affects motor function, cognition, and psychiatric health. Huntington disease detection is the need for more accurate, objective, and early diagnostic tools to predict disease progression. To overcome these challenges, a novel AML-HD model is proposed for the classification of Huntington disease. Initially, the Huntington images are preprocessed using a Contrast limited adaptive histogram equalization to enhance the image quality and remove the noise artifacts. In segmentation using Graph based segmentation used in HD brain images. Then feature extraction using a Random Forest for segmented images used to extract features. Finally, Support Vector Machine is used to classify types of HD namely Normal, Intermediate, Reduced penetrance and penetrance respectively. The effectiveness of the proposed AML-HD method using metrics like F1 score, sensitivity, accuracy, and specificity. The classification accuracy of the proposed AML-HD model was 96.54%. The proposed model enhanced the total accuracy by 3.81%, 3.71%, and 6.71% better than CNN, FPSOCNN, and ODNN respectively.