Computer-Aided Colorectal Tumor Classification in NBI Endoscopy Using CNN Features
This work addresses medical image classification for healthcare, but it is incremental as it applies existing CNN methods to a specific domain.
The paper tackled colorectal tumor classification in NBI endoscopy by using CNN features with linear SVM classifiers, achieving a recognition rate of 95% comparable to non-CNN methods.
In this paper we report results for recognizing colorectal NBI endoscopic images by using features extracted from convolutional neural network (CNN). In this comparative study, we extract features from different layers from different CNN models, and then train linear SVM classifiers. Experimental results with 10-fold cross validations show that features from first few convolution layers are enough to achieve similar performance (i.e., recognition rate of 95%) with non-CNN local features such as Bag-of-Visual words, Fisher vector, and VLAD.