COVID-19 mortality prediction in the intensive care unit with deep learning based on longitudinal chest X-rays and clinical data
Objectives We aimed to develop deep learning models using longitudinal chest X-rays (CXRs) and clinical data to predict in-hospital mortality of COVID-19 patients in the intensive care unit (ICU). Methods Six hundred fifty-four patients (212 deceased, 442 alive, 5645 total CXRs) were identified across two institutions. Imaging and clinical data from one institution were used to train five longitudinal transformer-based networks applying five-fold cross-validation. The models were tested on data from the other institution, and pairwise comparisons were used to determine the best-performing models. Results A higher proportion of deceased patients had elevated white blood cell count, decreased absolute lymphocyte count, elevated creatine concentration, and incidence of cardiovascular and chronic kidney disease. A model based on pre-ICU CXRs achieved an AUC of 0.632 and an accuracy of 0.593, and a model based on ICU CXRs achieved an AUC of 0.697 and an accuracy of 0.657. A model based on all longitudinal CXRs (both pre-ICU and ICU) achieved an AUC of 0.702 and an accuracy of 0.694. A model based on clinical data alone achieved an AUC of 0.653 and an accuracy of 0.657. The addition of longitudinal imaging to clinical data in a combined model significantly improved performance, reaching an AUC of 0.727 (p = 0.039) and an accuracy of 0.732. Conclusions The addition of longitudinal CXRs to clinical data significantly improves mortality prediction with deep learning for COVID-19 patients in the ICU. Key Points • Deep learning was used to predict mortality in COVID-19 ICU patients. • Serial radiographs and clinical data were used. • The models could inform clinical decision-making and resource allocation..
| Medienart: |
E-Artikel |
|---|
| Erscheinungsjahr: |
2022 |
|---|---|
| Erschienen: |
2022 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:32 |
|---|---|
| Enthalten in: |
European radiology - 32(2022), 7 vom: 19. Feb., Seite 4446-4456 |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Cheng, Jianhong [VerfasserIn] |
|---|
| Links: |
Volltext [lizenzpflichtig] |
|---|
| Themen: |
Artificial intelligence |
|---|
| Anmerkungen: |
© The Author(s), under exclusive licence to European Society of Radiology 2022. corrected publication 2022 |
|---|
| doi: |
10.1007/s00330-022-08588-8 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
SPR047357924 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | SPR047357924 | ||
| 003 | DE-627 | ||
| 005 | 20230519200626.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 220622s2022 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1007/s00330-022-08588-8 |2 doi | |
| 035 | |a (DE-627)SPR047357924 | ||
| 035 | |a (SPR)s00330-022-08588-8-e | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Cheng, Jianhong |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a COVID-19 mortality prediction in the intensive care unit with deep learning based on longitudinal chest X-rays and clinical data |
| 264 | 1 | |c 2022 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a © The Author(s), under exclusive licence to European Society of Radiology 2022. corrected publication 2022 | ||
| 520 | |a Objectives We aimed to develop deep learning models using longitudinal chest X-rays (CXRs) and clinical data to predict in-hospital mortality of COVID-19 patients in the intensive care unit (ICU). Methods Six hundred fifty-four patients (212 deceased, 442 alive, 5645 total CXRs) were identified across two institutions. Imaging and clinical data from one institution were used to train five longitudinal transformer-based networks applying five-fold cross-validation. The models were tested on data from the other institution, and pairwise comparisons were used to determine the best-performing models. Results A higher proportion of deceased patients had elevated white blood cell count, decreased absolute lymphocyte count, elevated creatine concentration, and incidence of cardiovascular and chronic kidney disease. A model based on pre-ICU CXRs achieved an AUC of 0.632 and an accuracy of 0.593, and a model based on ICU CXRs achieved an AUC of 0.697 and an accuracy of 0.657. A model based on all longitudinal CXRs (both pre-ICU and ICU) achieved an AUC of 0.702 and an accuracy of 0.694. A model based on clinical data alone achieved an AUC of 0.653 and an accuracy of 0.657. The addition of longitudinal imaging to clinical data in a combined model significantly improved performance, reaching an AUC of 0.727 (p = 0.039) and an accuracy of 0.732. Conclusions The addition of longitudinal CXRs to clinical data significantly improves mortality prediction with deep learning for COVID-19 patients in the ICU. Key Points • Deep learning was used to predict mortality in COVID-19 ICU patients. • Serial radiographs and clinical data were used. • The models could inform clinical decision-making and resource allocation. | ||
| 650 | 4 | |a Artificial intelligence |7 (dpeaa)DE-He213 | |
| 650 | 4 | |a Machine learning |7 (dpeaa)DE-He213 | |
| 650 | 4 | |a Prognosis |7 (dpeaa)DE-He213 | |
| 650 | 4 | |a Hospital mortality |7 (dpeaa)DE-He213 | |
| 650 | 4 | |a Coronavirus |7 (dpeaa)DE-He213 | |
| 700 | 1 | |a Sollee, John |4 aut | |
| 700 | 1 | |a Hsieh, Celina |4 aut | |
| 700 | 1 | |a Yue, Hailin |4 aut | |
| 700 | 1 | |a Vandal, Nicholas |4 aut | |
| 700 | 1 | |a Shanahan, Justin |4 aut | |
| 700 | 1 | |a Choi, Ji Whae |4 aut | |
| 700 | 1 | |a Tran, Thi My Linh |4 aut | |
| 700 | 1 | |a Halsey, Kasey |4 aut | |
| 700 | 1 | |a Iheanacho, Franklin |4 aut | |
| 700 | 1 | |a Warren, James |4 aut | |
| 700 | 1 | |a Ahmed, Abdullah |4 aut | |
| 700 | 1 | |a Eickhoff, Carsten |4 aut | |
| 700 | 1 | |a Feldman, Michael |4 aut | |
| 700 | 1 | |a Mortani Barbosa, Eduardo |4 aut | |
| 700 | 1 | |a Kamel, Ihab |4 aut | |
| 700 | 1 | |a Lin, Cheng Ting |4 aut | |
| 700 | 1 | |a Yi, Thomas |4 aut | |
| 700 | 1 | |a Healey, Terrance |4 aut | |
| 700 | 1 | |a Zhang, Paul |4 aut | |
| 700 | 1 | |a Wu, Jing |4 aut | |
| 700 | 1 | |a Atalay, Michael |4 aut | |
| 700 | 1 | |a Bai, Harrison X. |0 (orcid)0000-0002-9614-2163 |4 aut | |
| 700 | 1 | |a Jiao, Zhicheng |4 aut | |
| 700 | 1 | |a Wang, Jianxin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t European radiology |d Berlin : Springer, 1991 |g 32(2022), 7 vom: 19. Feb., Seite 4446-4456 |w (DE-627)SPR00395112X |w (DE-600)1472718-3 |x 1432-1084 |7 nnns |
| 773 | 1 | 8 | |g volume:32 |g year:2022 |g number:7 |g day:19 |g month:02 |g pages:4446-4456 |
| 856 | 4 | 0 | |u https://dx.doi.org/10.1007/s00330-022-08588-8 |z lizenzpflichtig |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_SPRINGER | ||
| 912 | |a SSG-OLC-PHA | ||
| 951 | |a AR | ||
| 952 | |d 32 |j 2022 |e 7 |b 19 |c 02 |h 4446-4456 | ||