Thyroid Detection And Classification Using Dnn Based On Hybrid Meta-Heuristic And Lstm Technique

In the field of medical research, prediction, as well diagnosis thyroid disease, is a major cause that challenging onset axiom. metabolism regulation, hormone secretions play significant role. Two frequent diseases are hypothyroidism and hyperthyroidism release hormones like thyroid, which regulate body’s rate. For analytics, approach data cleansing utilized to analyze enough primitive data, demonstrates patients’ risk. Deep Neural Networks (DNN) most vital efficient technology, predict disorder thyroid. To avoid errors human, evaluation manual process consumes expertise domain time. detect novel Long Short-Term Memory based Convolution Network (LSTM-CNN) with occurrence area Vgg-19. selecting feature, bias correction integrated hybrid optimization technique i.e., Black Widow Optimization Mayfly Approach (HBWO-MOA), also for classifying disease LSTM Vgg-19 Learning (DL) presented. From DDTI dataset image ultrasound, prediction classification efficiency. This analysis shown proposed technology accurate than convolutional methodology. When compared existing techniques AlexNet-LSTM, ResNet-LSTM, Vgg16-LSTM, Vgg-19-LSTM’s precision, sensitivity, accuracy, recalls F1_score effective.