Predicting respiratory failure in patients with COVID-19 pneumonia: a case study from Northern Italy. Ferrari, D., Mandreoli, F., Guaraldi, G., & Missier, P. In The HELPLINE workshop, co-located with the 24th European Conference on AI (ECAI2020), Online!, 2020. CEUR-WS. abstract bibtex 2 downloads The Covid-19 crisis caught health care services around the world by surprise, putting unprecedented pressure on Intensive Care Units (ICU). To help clinical staff to manage the limited ICU capacity, we have developed a Machine Learning model to estimate the probability that a patient admitted to hospital with COVID-19 symptoms would develop severe respiratory failure and require Intensive Care within 48 hours of admission. The model was trained on an initial co-hort of 198 patients admitted to the Infectious Disease ward of Mod-ena University Hospital, in Italy, at the peak of the epidemic, and subsequently refined as more patients were admitted. Using the Light-GBM Decision Tree ensemble approach, we were able to achieve good accuracy (AUC = 0.84) despite a high rate of missing values. Furthermore, we have been able to provide clinicians with explanations in the form of personalised ranked lists of features for each prediction , using only 20 out of more than 90 variables, using Shapley values to describe the importance of each feature.
@inproceedings{ferrari_predicting_2020,
address = {Online!},
title = {Predicting respiratory failure in patients with {COVID}-19 pneumonia: a case study from {Northern} {Italy}},
abstract = {The Covid-19 crisis caught health care services around the world by surprise, putting unprecedented pressure on Intensive Care Units (ICU). To help clinical staff to manage the limited ICU capacity, we have developed a Machine Learning model to estimate the probability that a patient admitted to hospital with COVID-19 symptoms would develop severe respiratory failure and require Intensive Care within 48 hours of admission. The model was trained on an initial co-hort of 198 patients admitted to the Infectious Disease ward of Mod-ena University Hospital, in Italy, at the peak of the epidemic, and subsequently refined as more patients were admitted. Using the Light-GBM Decision Tree ensemble approach, we were able to achieve good accuracy (AUC = 0.84) despite a high rate of missing values. Furthermore, we have been able to provide clinicians with explanations in the form of personalised ranked lists of features for each prediction , using only 20 out of more than 90 variables, using Shapley values to describe the importance of each feature.},
booktitle = {The {HELPLINE} workshop, co-located with the 24th {European} {Conference} on {AI} ({ECAI2020})},
publisher = {CEUR-WS},
author = {Ferrari, Davide and Mandreoli, Federica and Guaraldi, Giovanni and Missier, Paolo},
year = {2020},
keywords = {\#covid, \#machine learning},
}
Downloads: 2
{"_id":"uNoheHksMrPTtdQxS","bibbaseid":"ferrari-mandreoli-guaraldi-missier-predictingrespiratoryfailureinpatientswithcovid19pneumoniaacasestudyfromnorthernitaly-2020","authorIDs":[],"author_short":["Ferrari, D.","Mandreoli, F.","Guaraldi, G.","Missier, P."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","address":"Online!","title":"Predicting respiratory failure in patients with COVID-19 pneumonia: a case study from Northern Italy","abstract":"The Covid-19 crisis caught health care services around the world by surprise, putting unprecedented pressure on Intensive Care Units (ICU). To help clinical staff to manage the limited ICU capacity, we have developed a Machine Learning model to estimate the probability that a patient admitted to hospital with COVID-19 symptoms would develop severe respiratory failure and require Intensive Care within 48 hours of admission. The model was trained on an initial co-hort of 198 patients admitted to the Infectious Disease ward of Mod-ena University Hospital, in Italy, at the peak of the epidemic, and subsequently refined as more patients were admitted. Using the Light-GBM Decision Tree ensemble approach, we were able to achieve good accuracy (AUC = 0.84) despite a high rate of missing values. Furthermore, we have been able to provide clinicians with explanations in the form of personalised ranked lists of features for each prediction , using only 20 out of more than 90 variables, using Shapley values to describe the importance of each feature.","booktitle":"The HELPLINE workshop, co-located with the 24th European Conference on AI (ECAI2020)","publisher":"CEUR-WS","author":[{"propositions":[],"lastnames":["Ferrari"],"firstnames":["Davide"],"suffixes":[]},{"propositions":[],"lastnames":["Mandreoli"],"firstnames":["Federica"],"suffixes":[]},{"propositions":[],"lastnames":["Guaraldi"],"firstnames":["Giovanni"],"suffixes":[]},{"propositions":[],"lastnames":["Missier"],"firstnames":["Paolo"],"suffixes":[]}],"year":"2020","keywords":"#covid, #machine learning","bibtex":"@inproceedings{ferrari_predicting_2020,\n\taddress = {Online!},\n\ttitle = {Predicting respiratory failure in patients with {COVID}-19 pneumonia: a case study from {Northern} {Italy}},\n\tabstract = {The Covid-19 crisis caught health care services around the world by surprise, putting unprecedented pressure on Intensive Care Units (ICU). To help clinical staff to manage the limited ICU capacity, we have developed a Machine Learning model to estimate the probability that a patient admitted to hospital with COVID-19 symptoms would develop severe respiratory failure and require Intensive Care within 48 hours of admission. The model was trained on an initial co-hort of 198 patients admitted to the Infectious Disease ward of Mod-ena University Hospital, in Italy, at the peak of the epidemic, and subsequently refined as more patients were admitted. Using the Light-GBM Decision Tree ensemble approach, we were able to achieve good accuracy (AUC = 0.84) despite a high rate of missing values. Furthermore, we have been able to provide clinicians with explanations in the form of personalised ranked lists of features for each prediction , using only 20 out of more than 90 variables, using Shapley values to describe the importance of each feature.},\n\tbooktitle = {The {HELPLINE} workshop, co-located with the 24th {European} {Conference} on {AI} ({ECAI2020})},\n\tpublisher = {CEUR-WS},\n\tauthor = {Ferrari, Davide and Mandreoli, Federica and Guaraldi, Giovanni and Missier, Paolo},\n\tyear = {2020},\n\tkeywords = {\\#covid, \\#machine learning},\n}\n\n","author_short":["Ferrari, D.","Mandreoli, F.","Guaraldi, G.","Missier, P."],"key":"ferrari_predicting_2020","id":"ferrari_predicting_2020","bibbaseid":"ferrari-mandreoli-guaraldi-missier-predictingrespiratoryfailureinpatientswithcovid19pneumoniaacasestudyfromnorthernitaly-2020","role":"author","urls":{},"keyword":["#covid","#machine learning"],"metadata":{"authorlinks":{}},"downloads":2},"bibtype":"inproceedings","biburl":"https://bibbase.org/network/files/tHMs8ic86gSWoTp44","creationDate":"2020-06-25T08:05:29.160Z","downloads":2,"keywords":["#covid","#machine learning"],"search_terms":["predicting","respiratory","failure","patients","covid","pneumonia","case","study","northern","italy","ferrari","mandreoli","guaraldi","missier"],"title":"Predicting respiratory failure in patients with COVID-19 pneumonia: a case study from Northern Italy","year":2020,"dataSources":["BDjqJntjXzyBmLxhv","oiWqtmpFQ6ZtiMEK2","k75vCTghu54BjX5qH","j9tnaL2u4rifwAc2v","NCorZq2vkXK6BnhLF","ze2X9uz8Dcv2oGipf","afppXLgSuddAzAL9e","wJE4ynGem9MRsXBRn","9zrgMZfGdRkdkNXfZ","qTQGxWDYeue2pHBus"]}