iCOVID: interpretable deep learning framework for early recovery-time prediction of COVID-19 patients
Abstract Most prior studies focused on developing models for the severity or mortality prediction of COVID-19 patients. However, effective models for recovery-time prediction are still lacking. Here, we present a deep learning solution named iCOVID that can successfully predict the recovery-time of...
| Main Authors: | , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2021-08-01
|
| Series: | npj Digital Medicine |
| Online Access: | https://doi.org/10.1038/s41746-021-00496-3 |
| _version_ | 1797643076268720128 |
|---|---|
| author | Jun Wang Chen Liu Jingwen Li Cheng Yuan Lichi Zhang Cheng Jin Jianwei Xu Yaqi Wang Yaofeng Wen Hongbing Lu Biao Li Chang Chen Xiangdong Li Dinggang Shen Dahong Qian Jian Wang |
| author_facet | Jun Wang Chen Liu Jingwen Li Cheng Yuan Lichi Zhang Cheng Jin Jianwei Xu Yaqi Wang Yaofeng Wen Hongbing Lu Biao Li Chang Chen Xiangdong Li Dinggang Shen Dahong Qian Jian Wang |
| author_sort | Jun Wang |
| collection | DOAJ |
| description | Abstract Most prior studies focused on developing models for the severity or mortality prediction of COVID-19 patients. However, effective models for recovery-time prediction are still lacking. Here, we present a deep learning solution named iCOVID that can successfully predict the recovery-time of COVID-19 patients based on predefined treatment schemes and heterogeneous multimodal patient information collected within 48 hours after admission. Meanwhile, an interpretable mechanism termed FSR is integrated into iCOVID to reveal the features greatly affecting the prediction of each patient. Data from a total of 3008 patients were collected from three hospitals in Wuhan, China, for large-scale verification. The experiments demonstrate that iCOVID can achieve a time-dependent concordance index of 74.9% (95% CI: 73.6-76.3%) and an average day error of 4.4 days (95% CI: 4.2-4.6 days). Our study reveals that treatment schemes, age, symptoms, comorbidities, and biomarkers are highly related to recovery-time predictions. |
| first_indexed | 2024-03-11T14:09:26Z |
| format | Article |
| id | doaj.art-a6e6a3508738431586a63408365d0e7e |
| institution | Directory Open Access Journal |
| issn | 2398-6352 |
| language | English |
| last_indexed | 2024-03-11T14:09:26Z |
| publishDate | 2021-08-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | npj Digital Medicine |
| spelling | doaj.art-a6e6a3508738431586a63408365d0e7e2023-11-02T00:42:06ZengNature Portfolionpj Digital Medicine2398-63522021-08-014111310.1038/s41746-021-00496-3iCOVID: interpretable deep learning framework for early recovery-time prediction of COVID-19 patientsJun Wang0Chen Liu1Jingwen Li2Cheng Yuan3Lichi Zhang4Cheng Jin5Jianwei Xu6Yaqi Wang7Yaofeng Wen8Hongbing Lu9Biao Li10Chang Chen11Xiangdong Li12Dinggang Shen13Dahong Qian14Jian Wang15School of Biomedical Engineering, Shanghai Jiao Tong UniversityDepartment of Radiology, Southwest Hospital, Third Military Medical University (Army Medical University)Department of Gastroenterology, Southwest Hospital, Third Military Medical University (Army Medical University)School of Biomedical Engineering, Shanghai Jiao Tong UniversitySchool of Biomedical Engineering, Shanghai Jiao Tong UniversityDepartment of Radiation Oncology, Stanford University School of MedicineSchool of Biomedical Engineering, Shanghai Jiao Tong UniversityCollege of Media, Communication University of ZhejiangSchool of Biomedical Engineering, Shanghai Jiao Tong UniversityCollege of Computer Science and Technology, Zhejiang UniversityDepartment of Nuclear Medicine, Ruijin HospitalDepartment of Thoracic Surgery, Shanghai Pulmonary Hospital, Tongji University School of MedicineDepartment of Radiology, General Hospital of Southern Theatre Command, PLASchool of Biomedical Engineering, ShanghaiTech UniversitySchool of Biomedical Engineering, Shanghai Jiao Tong UniversityDepartment of Radiology, Southwest Hospital, Third Military Medical University (Army Medical University)Abstract Most prior studies focused on developing models for the severity or mortality prediction of COVID-19 patients. However, effective models for recovery-time prediction are still lacking. Here, we present a deep learning solution named iCOVID that can successfully predict the recovery-time of COVID-19 patients based on predefined treatment schemes and heterogeneous multimodal patient information collected within 48 hours after admission. Meanwhile, an interpretable mechanism termed FSR is integrated into iCOVID to reveal the features greatly affecting the prediction of each patient. Data from a total of 3008 patients were collected from three hospitals in Wuhan, China, for large-scale verification. The experiments demonstrate that iCOVID can achieve a time-dependent concordance index of 74.9% (95% CI: 73.6-76.3%) and an average day error of 4.4 days (95% CI: 4.2-4.6 days). Our study reveals that treatment schemes, age, symptoms, comorbidities, and biomarkers are highly related to recovery-time predictions.https://doi.org/10.1038/s41746-021-00496-3 |
| spellingShingle | Jun Wang Chen Liu Jingwen Li Cheng Yuan Lichi Zhang Cheng Jin Jianwei Xu Yaqi Wang Yaofeng Wen Hongbing Lu Biao Li Chang Chen Xiangdong Li Dinggang Shen Dahong Qian Jian Wang iCOVID: interpretable deep learning framework for early recovery-time prediction of COVID-19 patients npj Digital Medicine |
| title | iCOVID: interpretable deep learning framework for early recovery-time prediction of COVID-19 patients |
| title_full | iCOVID: interpretable deep learning framework for early recovery-time prediction of COVID-19 patients |
| title_fullStr | iCOVID: interpretable deep learning framework for early recovery-time prediction of COVID-19 patients |
| title_full_unstemmed | iCOVID: interpretable deep learning framework for early recovery-time prediction of COVID-19 patients |
| title_short | iCOVID: interpretable deep learning framework for early recovery-time prediction of COVID-19 patients |
| title_sort | icovid interpretable deep learning framework for early recovery time prediction of covid 19 patients |
| url | https://doi.org/10.1038/s41746-021-00496-3 |
| work_keys_str_mv | AT junwang icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT chenliu icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT jingwenli icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT chengyuan icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT lichizhang icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT chengjin icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT jianweixu icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT yaqiwang icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT yaofengwen icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT hongbinglu icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT biaoli icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT changchen icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT xiangdongli icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT dinggangshen icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT dahongqian icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients AT jianwang icovidinterpretabledeeplearningframeworkforearlyrecoverytimepredictionofcovid19patients |