Fangyun Li1, Lingxiao Zhou2, Yunpeng Wang1, Chuan Chen3, Shuyi Yang4,5, Fei Shan3, Lei Liu1,6. 1. Institute of Biomedical Sciences, Fudan University, Shanghai, China. 2. Institute of Microscale Optoelectronics, Shenzhen University, Shenzhen, China. 3. Department of Radiology, Shanghai Public Health Clinical Center, Fudan University, Shanghai, China. 4. Department of Radiology, Zhongshan Hospital, Fudan University, Shanghai, China. 5. Shanghai Institute of Medical Imaging, Shanghai, China. 6. School of Basic Medical Sciences, Fudan University, Shanghai, China.
Abstract
Background: Computer-aided diagnosis based on chest X-ray (CXR) is an exponentially growing field of research owing to the development of deep learning, especially convolutional neural networks (CNNs). However, due to the intrinsic locality of convolution operations, CNNs cannot model long-range dependencies. Although vision transformers (ViTs) have recently been proposed to alleviate this limitation, those trained on patches cannot learn any dependencies for inter-patch pixels and thus, are insufficient for medical image detection. To address this problem, in this paper, we propose a CXR detection method which integrates CNN with a ViT for modeling patch-wise and inter-patch dependencies. Methods: We experimented on the ChestX-ray14 dataset and followed the official training-test set split. Because the training data only had global annotations, the detection network was weakly supervised. A DenseNet with a feature pyramid structure was designed and integrated with an adaptive ViT to model inter-patch and patch-wise long-range dependencies and obtain fine-grained feature maps. We compared the performance using our method with that of other disease detection methods. Results: For disease classification, our method achieved the best result among all the disease detection methods, with a mean area under the curve (AUC) of 0.829. For lesion localization, our method achieved significantly higher intersection of the union (IoU) scores on the test images with bounding box annotations than did the other detection methods. The visualized results showed that our predictions were more accurate and detailed. Furthermore, evaluation of our method in an external validation dataset demonstrated its generalization ability. Conclusions: Our proposed method achieves the new state of the art for thoracic disease classification and weakly supervised localization. It has potential to assist in clinical decision-making. 2022 Quantitative Imaging in Medicine and Surgery. All rights reserved.
Background: Computer-aided diagnosis based on chest X-ray (CXR) is an exponentially growing field of research owing to the development of deep learning, especially convolutional neural networks (CNNs). However, due to the intrinsic locality of convolution operations, CNNs cannot model long-range dependencies. Although vision transformers (ViTs) have recently been proposed to alleviate this limitation, those trained on patches cannot learn any dependencies for inter-patch pixels and thus, are insufficient for medical image detection. To address this problem, in this paper, we propose a CXR detection method which integrates CNN with a ViT for modeling patch-wise and inter-patch dependencies. Methods: We experimented on the ChestX-ray14 dataset and followed the official training-test set split. Because the training data only had global annotations, the detection network was weakly supervised. A DenseNet with a feature pyramid structure was designed and integrated with an adaptive ViT to model inter-patch and patch-wise long-range dependencies and obtain fine-grained feature maps. We compared the performance using our method with that of other disease detection methods. Results: For disease classification, our method achieved the best result among all the disease detection methods, with a mean area under the curve (AUC) of 0.829. For lesion localization, our method achieved significantly higher intersection of the union (IoU) scores on the test images with bounding box annotations than did the other detection methods. The visualized results showed that our predictions were more accurate and detailed. Furthermore, evaluation of our method in an external validation dataset demonstrated its generalization ability. Conclusions: Our proposed method achieves the new state of the art for thoracic disease classification and weakly supervised localization. It has potential to assist in clinical decision-making. 2022 Quantitative Imaging in Medicine and Surgery. All rights reserved.
Authors: Markus Körner; Christof H Weber; Stefan Wirth; Klaus-Jürgen Pfeifer; Maximilian F Reiser; Marcus Treitl Journal: Radiographics Date: 2007 May-Jun Impact factor: 5.333
Authors: Geert Litjens; Thijs Kooi; Babak Ehteshami Bejnordi; Arnaud Arindra Adiyoso Setio; Francesco Ciompi; Mohsen Ghafoorian; Jeroen A W M van der Laak; Bram van Ginneken; Clara I Sánchez Journal: Med Image Anal Date: 2017-07-26 Impact factor: 8.545
Authors: Arnaud Arindra Adiyoso Setio; Alberto Traverso; Thomas de Bel; Moira S N Berens; Cas van den Bogaard; Piergiorgio Cerello; Hao Chen; Qi Dou; Maria Evelina Fantacci; Bram Geurts; Robbert van der Gugten; Pheng Ann Heng; Bart Jansen; Michael M J de Kaste; Valentin Kotov; Jack Yu-Hung Lin; Jeroen T M C Manders; Alexander Sóñora-Mengana; Juan Carlos García-Naranjo; Evgenia Papavasileiou; Mathias Prokop; Marco Saletta; Cornelia M Schaefer-Prokop; Ernst T Scholten; Luuk Scholten; Miranda M Snoeren; Ernesto Lopez Torres; Jef Vandemeulebroucke; Nicole Walasek; Guido C A Zuidhof; Bram van Ginneken; Colin Jacobs Journal: Med Image Anal Date: 2017-07-13 Impact factor: 8.545