In recent years, for reducing diagnostic burdens in stomach screening, a computer aided diagnostic system (CAD system) for endoscopic stomach images is required. In our previous study, an automated polyp detection method from endoscopic images using the SSD (Single Shot MultiBox Detector) has been developed with 93.7% of detection rate. However, the detection target of this method has been limited only to fundic gland polyp. In this paper, we propose a method for automated detection and classification of two different types of polyp; fundic gland polyp (FGP) and hyperplastic polyp (HP) from endoscopic images using the SSD. In the experiment, 71 and 96 practical endoscopic images of FGP and HP were used. For training of SSD, 11210 and 5053 training images of FGP and HP were generated by data augmentation, respectively, and 20% of training images were automatically selected and used as verification images. As a result for test samples including 132 polyps (69 FGPs and 63 HPs), the detection rate for entire polyps was 96.2% (127/132), and the classification rate for two types of polyp was 88.6% (117/132). The number of false positive was only one all through the experiment.
In stomach lesion screening, endoscopic images provide the most effective diagnostic information. However, in the most of lesions at the initial stage, the sign of existence is hard to appear on endoscopic images, and also there is the difference in operations of endoscopes and observation of images in real time among individual medical doctors. Therefore, development of a computer aided diagnostic system (CAD system) for endoscopic images is required. In this study, we propose a method for automated detection of fundic gland polyps from endoscopic images using an object detection algorithm named SSD (Single Shot MultiBox Detector) which is one of CNN (Convolutional Neural Network). SSD used here has 20 of convolution layers and 6 of pooling layers, and the input image size is 300x300. In the experiment, 73 practical fundic gland polyp images were used. To compensate for lack of training images, augmentation was performed using image rotation and edge enhancement. We trained 8751 training images and 2188 verification images. Also, as a preprocessing, highlight areas were removed automatically from all images including both training and test samples. As a result, 94.7% of TP (true positive) rate for 73 fundic gland polyp images was obtained by using our learned SSD.