Fang, Yu-Jen, Huang, Chien-Wei, Karmakar, Riya, Mukundan, Arvind, Tsao, Yu-Ming, Yang, Kai-Yao, and Wang, Hsiang-Chen
Simple Summary: Esophageal carcinoma (EC) is a major cause of cancer deaths since it is first undetectable in its early stages. Narrow-band imaging (NBI) detects EC more accurately, sensitively, and specifically than white light imaging (WLI), according to many studies. This work uses a color space connected to décor to change WLIs into NBIs, improving early EC identification. The YOLOv5 algorithm was utilized to train WLI and NBI images separately, demonstrating the versatility of sophisticated object identification approaches in medical image analysis. Based on the confusion matrix and the trained model's precision, recall, specificity, accuracy, and F1-score, the model's performance was assessed. The model was trained to reliably identify dysplasia, squamous cell carcinoma (SCC), and polyps, demonstrating a detailed and focused examination of EC pathology for a better understanding. Dysplasia cancer, a pre-cancerous stage that may increase five-year survival, was detected with higher recall and accuracy by the NBI model. Although the NBI and WLI models recognized the polyp identically, the SCC category lowered its accuracy and recall rate. The NBI model had an accuracy of 0.60, 0.81, and 0.66 in dysplasia, SCC, and polyp categories, and recall rates of 0.40, 0.73, and 0.76. In dysplasia, SCC, and polyp categories, the WLI model was 0.56, 0.99, and 0.65 accurate. Additionally, it had recall rates of 0.39, 0.86, and 0.78 in the same categories. The NBI model performs poorly due to a small collection of training pictures. Increasing the dataset can improve performance. Esophageal carcinoma (EC) is a prominent contributor to cancer-related mortality since it lacks discernible features in its first phases. Multiple studies have shown that narrow-band imaging (NBI) has superior accuracy, sensitivity, and specificity in detecting EC compared to white light imaging (WLI). Thus, this study innovatively employs a color space linked to décor to transform WLIs into NBIs, offering a novel approach to enhance the detection capabilities of EC in its early stages. In this study a total of 3415 WLI along with the corresponding 3415 simulated NBI images were used for analysis combined with the YOLOv5 algorithm to train the WLI images and the NBI images individually showcasing the adaptability of advanced object detection techniques in the context of medical image analysis. The evaluation of the model's performance was based on the produced confusion matrix and five key metrics: precision, recall, specificity, accuracy, and F1-score of the trained model. The model underwent training to accurately identify three specific manifestations of EC, namely dysplasia, squamous cell carcinoma (SCC), and polyps demonstrates a nuanced and targeted analysis, addressing diverse aspects of EC pathology for a more comprehensive understanding. The NBI model effectively enhanced both its recall and accuracy rates in detecting dysplasia cancer, a pre-cancerous stage that might improve the overall five-year survival rate. Conversely, the SCC category decreased its accuracy and recall rate, although the NBI and WLI models performed similarly in recognizing the polyp. The NBI model demonstrated an accuracy of 0.60, 0.81, and 0.66 in the dysplasia, SCC, and polyp categories, respectively. Additionally, it attained a recall rate of 0.40, 0.73, and 0.76 in the same categories. The WLI model demonstrated an accuracy of 0.56, 0.99, and 0.65 in the dysplasia, SCC, and polyp categories, respectively. Additionally, it obtained a recall rate of 0.39, 0.86, and 0.78 in the same categories, respectively. The limited number of training photos is the reason for the suboptimal performance of the NBI model which can be improved by increasing the dataset. [ABSTRACT FROM AUTHOR]