Importance of Preprocessing in Histopathology Image Classification Using Deep Convolutional Neural Network

The aim of this study is to propose an alternative and hybrid solution method for diagnosing the disease from histopathology images taken animals with paratuberculosis intact intestine. In detail, based on using both image processing deep learning better results. Reliable detection known as open problem in medical solutions need be developed. context, 520 were collected a joint Burdur Mehmet Akif Ersoy University, Faculty Veterinary Medicine, Department Pathology. Manually detecting interpreting these requires expertise lot time. For reason, veterinarians, especially newly recruited physicians, have great imaging computer vision systems development treatment methods disease. proposed use CLAHE together. After preprocessing, diagnosis made by classifying convolutional neural network supported VGG-16 architecture. This uses completely original dataset images. Two types applied evaluation parameters. While F1 Score was 93% classified without data it 98% that preprocessed method.