Dual-Stage U-Shape Convolutional Network for Esophageal Tissue Segmentation in OCT Images

Abstract

Automatic segmentation is the crucial step for esophageal optical coherence tomography (OCT) image processing, which is able to highlight diagnosis-related tissue layers and provide characteristics such as shape and thickness for esophageal disease diagnosis. This study proposes a dual-stage framework using a specifically designed encoder-decoder network configuration for accurate and reliable esophageal layer segmentation, which is named as the dual-stage U-shape convolutional network (D-UCN). The proposed approach utilized one UCN to locate the target tissue region, which is followed by another UCN with similar architecture to achieve the final segmentation. In this way, the proposed strategy effectively solves the problems encountered in our previous studies, such as disturbance from neighboring diagnostically unrelated tissues, probe protection sheaths from the imaging equipment and the inevitable speckle noise. Experimental results on esophageal OCT B-scans from C57BL mice demonstrated that the proposed dual-stage framework achieved performance comparable to manual segmentation. The effectiveness and advantages of the dual-stage strategy are also confirmed in comparison with graph theory dynamic program (GTDP) and U-Net. INDEX TERMS Esophageal layer segmentation, Optical coherence tomography, Fully convolutional network, Medical image analysis. FIGURE 1. Demonstration of (a) a typical esophageal OCT image for mouse and (b) the corresponding manual segmentation result. B. FRAMEWORK AND ARCHITECTURE This study proposes a D-UCN framework to segment esophageal layers from OCT images in a dual-stage strategy. It consists of two cascaded parts with the same U-shape network architecture, UCN-I and UCN-II, as shown in Fig.

Machine Learning Applications for Early Detection of Esophageal Cancer: A Systematic Review

“…However, smaller sample sizes can sometimes result in higher accuracy, which has been observed in the reviewed studies (26,32). Interestingly, studies that reported the highest accuracy did not provide information on the sample size used, which suggests that other factors such as feature processing and model parameter tuning may also play a crucial role (29,37,55). Therefore, future studies should aim to investigate the optimal sample size for issues related to the clinical eld, while also examining the characteristics of ML algorithms, including feature extraction, selection, and optimization, to achieve more accurate and reliable results.…”

“…In one case, an average accuracy of 98% was achieved using Optical coherence tomography (OCT) images (Fig. 5) (37).…”

Machine learning applications for early detection of esophageal cancer: a systematic review

“…In one case, an average accuracy of 98% was achieved using Optical coherence tomography (OCT) images (Fig. 5 ) [ 39 ].…”

“…Furthermore, Faster R-CNN (6445 images), SegNet (6473 images), Neuro_T (5162 images), and YOLO v5 (4447 images) were other ML algorithms that utilized a large sample size for training, testing, and validation [ 24 , 27 , 42 , 49 ]. In addition, in studies focused on early detection of esophageal cancer, U-Net [ 33 , 35 , 36 , 38 , 39 ], Faster R-CNN [ 13 , 26 , 48 , 49 , 51 ] SSD [ 13 , 20 , 25 , 30 , 37 ] algorithms reported in 5 studies had the highest number of uses among all ML algorithms. VGG16 algorithm was also used in 3 studies [ 25 , 28 , 49 ].…”

“…However, smaller sample sizes can sometimes result in higher accuracy, which has been observed in the reviewed studies [ 28 , 34 ]. Interestingly, studies that reported the highest accuracy did not provide information on the sample size used, which suggests that other factors such as feature processing and model parameter tuning may also play a crucial role [ 31 , 39 , 53 ]. Therefore, future studies should aim to investigate the optimal sample size for issues related to the clinical field, while also examining the characteristics of ML algorithms, including feature extraction, selection, and optimization, to achieve more accurate and reliable results.…”