An intelligent location method of key boundary points for assisting the diameter measurement of central serous chorioretinopathy lesion area. (August 2022)

• Record Type: Journal Article
• Authors: Xu, Jianguo; Shen, Jianxin; Yan, Zhipeng; Zhou, Fen; Wan, Cheng; Yang, Weihua
• Is Part Of: Computers in biology and medicine. Volume 147(2022)
• Publication Date: 2022-08
• Subjects: Central serous chorioretinopathy -- Initial location module -- Correction module -- Multi-task learning -- Cascade model; Medicine -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; 610.285
• DOI: 10.1016/j.compbiomed.2022.105730
• Languages: English
• ISSNs: 0010-4825

Abstract

Abstract: The diameter of central serous chorioretinopathy (CSCR) lesion is one of the important indicators to evaluate the severity of CSCR and the efficacy of corresponding treatment schemes. Traditional manual measurement by ophthalmologists is usually based on a single or a small number of optical coherence tomography (OCT) B-scan images. This measurement scheme may not be convincing, vulnerable to subjective factors and lower efficiency. To alleviate the above situation, this paper proposes an intelligent key boundary point location method for all B-scan images of a single patient to assist in the rapid and accurate diameter measurement of the CSCR lesion area. Firstly, an initial location module (ILM) based on the multi-task learning paradigm is appropriately adjusted and introduced into the key boundary point location task, which preliminarily realizes the rapid location of key boundary points. Secondly, to further ameliorate the ILM, a gradient based correction module (GBCM) is designed, followed by the construction of the cascade model (ILM-GBCM) which improves the location accuracy of key boundary points as a whole. Extensive experiments based on five different convolutional neural network (CNN) backbones are carried out, revealing the feasibility of ILM in this task and the effectiveness of ILM-GBCM. On 912 testing images, the maximum correction ratio reaches 83.66%, and the minimum location time at the image level is as low as 0.1754 s, which not only confirms the necessity of correction operation, but also greatly reduce the time cost of ophthalmologists' manual measurement operation in clinic. Graphical abstract: Image 1 Highlights: The IPM is established to realize the separating, denoising and clipping operations, laying a foundation for the dataset construction of the key boundary points in the CSCR lesion area. The multi-task learning based ILM is appropriately adjusted and successfully introduced into the location task of key boundary points, preliminarily realizing their rapid location. The ILM-GBCM is constructed, further improving the location accuracy of ILM and proving the effectiveness and necessity of the designed GBCM. The proposed method can quickly measure the CSCR lesion area diameters, greatly reducing the time cost and providing a potential solution for more rigorous judgment of CSCR severity and the efficacy evaluation of related treatment schemes.