A Liver Damage Prediction Using Partial Differential Segmentation with Improved Convolutional Neural Network

Abstract

Background. The liver is one of the most significant and most essential organs in the human body. It is divided into two granular lobes, one on the right and one on the left, connected by a bile duct. The liver is essential in the removal of waste products from human food consumption, the creation of bile, the regulation of metabolic activities, the cleaning of the blood by sensitizing digestive management, and the storage of vitamins and minerals. To perform the classification of liver illnesses using computed tomography (CT scans), two critical phases must first be completed: liver segmentation and categorization. The most difficult challenge in categorizing liver disease is distinguishing the liver from the other organs near it. Methodology. Liver biopsy is a kind of invasive diagnostic procedure, widely regarded as the gold standard for accurately estimating the severity of liver disease. Noninvasive approaches for examining liver illnesses, such as blood serum markers and medical imaging (ultrasound, magnetic resonance MR, and CT) have also been developed. This approach uses the Partial Differential Technique (PDT) to separate the liver from the other organs and Level Set Methodology (LSM) for separating the cancer location from the surrounding tissue based on the projected pictures used as input. With the help of an Improved Convolutional Classifier, the categorization of different phases may be accomplished. Results. Several accuracies, sensitivity, and specificity measurements are produced to assess the categorization of LSM using an Improved Convolutional classifier. Approximately, 97.5% of the performance accuracy of the liver categorization is achieved with a 94.5% continuous interval (CI) of [0.6775 1.0000] and an error rate of 2.1%. The suggested method’s performance is compared to that of two existing algorithms, and the sensitivity and specificity provide an overall average of 96% and 93%, respectively, with 95% Continuous Interval of [0.7513 1.0000] and [0.7126 1.0000] for sensitivity and specificity, respectively.

Random forest and support vector machine based hybrid liver disease detection

“…The major problem in liver disease diagnosis using ML techniques is improving the accuracy of ML algorithms employed in liver disease diagnosis [6]. Improved accuracy results in better diagnosis results reducing the false negatives and finally increasing the precision in the diagnosis of the liver.…”

Self-testing Algorithm of Metro One-button Switch Station Equipment Based on a Neural Network

“…The connectivity patterns between neurons in deep neural networks are inspired by the visual cortical structures found in animal brains, and convolutional neural networks are one of the classical and widely used structures [5]. Sumathy et al used convolutional neural networks to classify medical images and achieved prediction of liver injury [6]. Dong implemented fault diagnosis of rolling bearings based on convolutional neural networks and fast Fourier transform [7].…”

Optimized Deep Learning Model for Disease Prediction in Potato Leaves

“…A kind of machine learning [4][5][6] called deep learning (DL) has been demonstrated to be particularly good at classifying images [5][6][7][8][9]. Convolutional Neural Network (CNN) is used in a variety of tasks, including object identification and image classification [10][11][12][13].…”