Deep Learning Algorithms for Detection and Classification of Gastrointestinal Diseases

Currently, nearly two million patients die of gastrointestinal diseases worldwide. Video endoscopy is one of the latest technologies in the medical imaging field for the diagnosis of gastrointestinal diseases, such as stomach ulcers, bleeding, and polyps. Medical video endoscopy generates many images, so doctors need considerable time to follow up all the images. This creates a challenge for manual diagnosis and has encouraged investigations into computer-aided techniques to diagnose all the generated images in a short period and with high accuracy. The novelty of the proposed methodology lies in developing a system for diagnosis of gastrointestinal diseases. This paper introduces three networks, GoogleNet, ResNet-50, and AlexNet, which are based on deep learning and evaluates them for their potential in diagnosing a dataset of lower gastrointestinal diseases. All images are enhanced, and the noise is removed before they are inputted into the deep learning networks. The Kvasir dataset contains 5,000 images divided equally into five types of lower gastrointestinal diseases (dyed-lifted polyps, normal cecum, normal pylorus, polyps, and ulcerative colitis). In the classification stage, pretrained convolutional neural network (CNN) models are tuned by transferring learning to perform new tasks. The softmax activation function receives the deep feature vector and classifies the input images into five classes. All CNN models achieved superior results. AlexNet achieved an accuracy of 97%, sensitivity of 96.8%, specificity of 99.20%, and AUC of 99.98%.

Five-year result of microvascular decompression using video endoscopy in the treatment of classic trigeminal neuralgia with paroxysmal pain syndrome

Background. The incidence of trigeminal neuralgia (TN) is 15 per 100,000 people per year. The effectiveness of the existing conservative methods of therapy does not exceed 50%. The use of carbamazepine doubles the frequency of depressive conditions, and 40% of suicidal thoughts. Purpose of the study. To evaluate the long-term results of microvascular decompression using video endoscopy in the treatment of patients with classical trigeminal neuralgia (cNTN) with paroxysmal facial pain. Methods. At the Federal Research and Clinical Center of the FMBA of Russia in the period from 2014 to 2019. 96 patients with cNTN were operated on in 62 (64%) of whom neuralgia was with paroxysmal facial pain, and in 34 (36%) - with constant pain. The average period from the onset of pain syndrome to surgery was 5 years (from 2 months to 15 years). The maximum pain intensity upon admission to the hospital according to the visual analogue scale (VAS) was 10 points, according to the BNI (Barrow Neurological Institute) pain syndrome scale - V. All patients underwent MIA of the trigeminal nerve root using Teflon, and in 9 patients during surgery used video endoscopic assistance. The average follow-up period after surgery was 3.4 1.7 years (from 1 to 5 years).Results. In all (100%) patients, pain was completely relieved after surgery (BNI - I). Excellent and good results after MVD within 5 years were achieved in 98% of patients (BNI - I-II). Facial hypesthesia, which does not bring discomfort and anxiety (BNI-II), developed in 8% (n = 5) of patients. The use of video endoscopy made it possible to identify vessels compressing the trigeminal nerve root with minimal traction of the cerebellum and cranial nerves. The development of cerebellar edema and ischemia occurred in one (1.6%) patient.Conclusion. The MVD method with video endoscopy is effective in the treatment of patients with cNTN with paroxysmal pain syndrome.

Decomposition of color wavelet with higher order statistical texture and convolutional neural network features set based classification of colorectal polyps from video endoscopy

Gastrointestinal cancer is one of the leading causes of death across the world. The gastrointestinal polyps are considered as the precursors of developing this malignant cancer. In order to condense the probability of cancer, early detection and removal of colorectal polyps can be cogitated. The most used diagnostic modality for colorectal polyps is video endoscopy. But the accuracy of diagnosis mostly depends on doctors' experience that is crucial to detect polyps in many cases. Computer-aided polyp detection is promising to reduce the miss detection rate of the polyp and thus improve the accuracy of diagnosis results. The proposed method first detects polyp and non-polyp then illustrates an automatic polyp classification technique from endoscopic video through color wavelet with higher-order statistical texture feature and Convolutional Neural Network (CNN). Gray Level Run Length Matrix (GLRLM) is used for higher-order statistical texture features of different directions (Ɵ = 0o, 45o, 90o, 135o). The features are fed into a linear support vector machine (SVM) to train the classifier. The experimental result demonstrates that the proposed approach is auspicious and operative with residual network architecture, which triumphs the best performance of accuracy, sensitivity, and specificity of 98.83%, 97.87%, and 99.13% respectively for classification of colorectal polyps on standard public endoscopic video databases.