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 cohort of 198 patients admitted to the Infectious Disease ward of Modena University Hospital, in Italy, at the peak of the epidemic, and subsequently refined as more patients were admitted. Using the LightGBM 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.

Predicting Respiratory Failure in Patients with COVID-19 pneumonia: a case study from Northern Italy / Ferrari, Davide; Mandreoli, Federica; Guaraldi, Giovanni; Milić, Jovana; Missier, Paolo. - 2820:(2020), pp. 32-38. (Intervento presentato al convegno 1st International AAI4H - Advances in Artificial Intelligence for Healthcare Workshop, AAI4H 2020 tenutosi a Santiago de Compostela, Spain nel September 4, 2020).

Predicting Respiratory Failure in Patients with COVID-19 pneumonia: a case study from Northern Italy

Davide Ferrari;Federica Mandreoli;Giovanni Guaraldi;Jovana Milić;
2020

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 cohort of 198 patients admitted to the Infectious Disease ward of Modena University Hospital, in Italy, at the peak of the epidemic, and subsequently refined as more patients were admitted. Using the LightGBM 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.
2020
1st International AAI4H - Advances in Artificial Intelligence for Healthcare Workshop, AAI4H 2020
Santiago de Compostela, Spain
September 4, 2020
2820
32
38
Ferrari, Davide; Mandreoli, Federica; Guaraldi, Giovanni; Milić, Jovana; Missier, Paolo
Predicting Respiratory Failure in Patients with COVID-19 pneumonia: a case study from Northern Italy / Ferrari, Davide; Mandreoli, Federica; Guaraldi, Giovanni; Milić, Jovana; Missier, Paolo. - 2820:(2020), pp. 32-38. (Intervento presentato al convegno 1st International AAI4H - Advances in Artificial Intelligence for Healthcare Workshop, AAI4H 2020 tenutosi a Santiago de Compostela, Spain nel September 4, 2020).
File in questo prodotto:
File Dimensione Formato  
AAI4H-7.pdf

Open access

Tipologia: Versione pubblicata dall'editore
Dimensione 414.16 kB
Formato Adobe PDF
414.16 kB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

Licenza Creative Commons
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11380/1236926
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? ND
social impact