Seo, Hyun seok Yu, Lequan Ren, Hongyi Li, Xiaomeng Shen, Liyue Xing, Lei 2024-01-19T13:04:41Z 2024-01-19T13:04:41Z 2021-10-21 2021-12 0278-0062 https://pubs.kist.re.kr/handle/201004/116029 Deep learning is becoming an indispensable tool for imaging applications, such as image segmentation, classification, and detection. In this work, we reformulate a standard deep learning problem into a new neural network architecture with multi-output channels, which reflects different facets of the objective, and apply the deep neural network to improve the performance of image segmentation. By adding one or more interrelated auxiliary-output channels, we impose an effective consistency regularization for the main task of pixelated classification (i.e., image segmentation). Specifically, multi-output-channel consistency regularization is realized by residual learning via additive paths that connect main-output channel and auxiliary-output channels in the network. The method is evaluated on the detection and delineation of lung and liver tumors with public data. The results clearly show that multi-output-channel consistency implemented by residual learning improves the standard deep neural network. The proposed framework is quite broad and should find widespread applications in various deep learning problems. IEEE English Institute of Electrical and Electronics Engineers Inc. Deep Neural Network with Consistency Regularization of Multi-Output Channels for Improved Tumor Detection and Delineation Article 10.1109/TMI.2021.3084748 1 IEEE Transactions on Medical Imaging, v.40, no.12, pp.3369 - 3378 IEEE Transactions on Medical Imaging 40 12 3369 3378 N scie scopus 000724511900011 2-s2.0-85107193146 Computer Science, Interdisciplinary Applications Engineering, Biomedical Engineering, Electrical & Electronic Imaging Science & Photographic Technology Radiology, Nuclear Medicine & Medical Imaging Computer Science Engineering Imaging Science & Photographic Technology Radiology, Nuclear Medicine & Medical Imaging Article Deep neural networks Image classification Image enhancement Image segmentation Network architecture Neural networks Tumors Auxiliary output Imaging applications Indispensable tools Learning problem Liver tumors Multi-output Output channels Tumor detection Deep learning Artificial intelligence Biomedical imaging cancer detection Feature extraction Image segmentation Neural networks neural networks regularization residual learning segmentation Task analysis Training Tumors