File Download
There are no files associated with this item.
Supplementary
-
Citations:
- Appears in Collections:
Conference Paper: Online COVID-19 diagnosis with chest CT images: Lesion-attention deep neural networks
Title | Online COVID-19 diagnosis with chest CT images: Lesion-attention deep neural networks |
---|---|
Other Titles | A Fast Online COVID-19 Diagnostic System with Chest CT Scans |
Authors | |
Keywords | Attention COVID-19 CT image diagnosis deep neural network |
Issue Date | 2020 |
Publisher | ACM SIGKDD. |
Citation | 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2020): SIGKDD Health COVID-19 - 3rd Annual Health Day at the KDD Conference: AI for COVID-19, Virtual Conference. 24 August 2020 How to Cite? |
Abstract | Chest computed tomography (CT) scanning is one of the most important technologies for COVID-19 diagnosis and disease monitoring, particularly for early detection of coronavirus. Recent advancements in computer vision motivate more concerted efforts in developing AI-driven diagnostic tools to accommodate the enormous demands for the COVID-19 diagnostic tests globally. To help alleviate burdens on medical systems, we develop a lesion-attention deep neural network (LA-DNN) to predict COVID-19 positive or negative with a richly annotated chest CT image dataset. Based on the textual radiological report accompanied with each CT image, we extract two types of important information for the annotations: One is the indicator of a positive or negative case of COVID-19, and the other is the description of five lesions on the CT images associated with the positive cases. The proposed data-efficient LA-DNN model focuses on the primary task of binary classification for COVID-19 diagnosis, while an auxiliary multi-label learning task is implemented simultaneously to draw the model’s attention to the five lesions associated with COVID-19. The joint task learning process makes it a highly sample-efficient deep neural network that can learn COVID-19 radiology features more effectively with limited but high-quality, rich-information samples. The experimental results show that the area under the curve (AUC) and sensitivity (recall), precision, and accuracy for COVID-19 diagnosis are 94.0%, 88.8%, 87.9%, and 88.6% respectively, which reach the clinical standards for practical use. A free online system is currently alive for fast diagnosis using CT images at the website https://www.covidct.cn/, and all codes and datasets are freely accessible at our github address. |
Description | Accepted for Poster Presentation title: A Fast Online COVID-19 Diagnostic System with Chest CT Scans |
Persistent Identifier | http://hdl.handle.net/10722/294802 |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Liu, B | - |
dc.contributor.author | Gao, X | - |
dc.contributor.author | He, M | - |
dc.contributor.author | Lv, F | - |
dc.contributor.author | Yin, G | - |
dc.date.accessioned | 2020-12-21T11:48:46Z | - |
dc.date.available | 2020-12-21T11:48:46Z | - |
dc.date.issued | 2020 | - |
dc.identifier.citation | 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2020): SIGKDD Health COVID-19 - 3rd Annual Health Day at the KDD Conference: AI for COVID-19, Virtual Conference. 24 August 2020 | - |
dc.identifier.uri | http://hdl.handle.net/10722/294802 | - |
dc.description | Accepted for Poster Presentation title: A Fast Online COVID-19 Diagnostic System with Chest CT Scans | - |
dc.description.abstract | Chest computed tomography (CT) scanning is one of the most important technologies for COVID-19 diagnosis and disease monitoring, particularly for early detection of coronavirus. Recent advancements in computer vision motivate more concerted efforts in developing AI-driven diagnostic tools to accommodate the enormous demands for the COVID-19 diagnostic tests globally. To help alleviate burdens on medical systems, we develop a lesion-attention deep neural network (LA-DNN) to predict COVID-19 positive or negative with a richly annotated chest CT image dataset. Based on the textual radiological report accompanied with each CT image, we extract two types of important information for the annotations: One is the indicator of a positive or negative case of COVID-19, and the other is the description of five lesions on the CT images associated with the positive cases. The proposed data-efficient LA-DNN model focuses on the primary task of binary classification for COVID-19 diagnosis, while an auxiliary multi-label learning task is implemented simultaneously to draw the model’s attention to the five lesions associated with COVID-19. The joint task learning process makes it a highly sample-efficient deep neural network that can learn COVID-19 radiology features more effectively with limited but high-quality, rich-information samples. The experimental results show that the area under the curve (AUC) and sensitivity (recall), precision, and accuracy for COVID-19 diagnosis are 94.0%, 88.8%, 87.9%, and 88.6% respectively, which reach the clinical standards for practical use. A free online system is currently alive for fast diagnosis using CT images at the website https://www.covidct.cn/, and all codes and datasets are freely accessible at our github address. | - |
dc.language | eng | - |
dc.publisher | ACM SIGKDD. | - |
dc.relation.ispartof | 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2020): SIGKDD Health COVID-19 - 3rd Annual Health Day: AI for COVID-19 | - |
dc.subject | Attention | - |
dc.subject | COVID-19 | - |
dc.subject | CT image | - |
dc.subject | diagnosis | - |
dc.subject | deep neural network | - |
dc.title | Online COVID-19 diagnosis with chest CT images: Lesion-attention deep neural networks | - |
dc.title.alternative | A Fast Online COVID-19 Diagnostic System with Chest CT Scans | - |
dc.type | Conference_Paper | - |
dc.identifier.email | Yin, G: gyin@hku.hk | - |
dc.identifier.authority | Yin, G=rp00831 | - |
dc.identifier.hkuros | 320599 | - |
dc.publisher.place | United States | - |