Detection of communicable and non-communicable disease using Lenet-Bi-Lstm model in pathology images

Detection of communicable and non-communicable diseases such as breast cancer and Haemoprotozoan based on the pathology images is important for the survival of the lives of cattle and humans. The existing models related to the segmentation approaches showed good results for the overlapped database but failed to extract the prominent structural and texture features. In the present research work, a deep neural network known as Bi-directional Long Short Term Memory (Bi-LSTM) model performs the process of classification for Haemoprotozoan images into anaplasmosis, babesiosis and theileriosis classes. Additionally, the breast cancer images are also classified as malignant or benign. The proposed research uses LeNet which is an early Convolution Neural Network (CNN) model that possesses the units of CNN for the feature extraction. The obtained features are given as an input for the Bi-LSTM model to classify Haemoprotozoan images and breast cancer images. The results obtained by the proposed LeNet-Bi-LSTM model showed the accuracy of 98.99% better when compared to existing models like Ensemble of Deep Learning Models that obtained the accuracy of 95.29% and Modified ReliefF Algorithm has 95.94% accuracy.