Artificial Intelligence Predicts Hospitalization Length in COVID-19 Patients with High Accuracy

Strategic Insights -

Researchers from the Institute for Cancer Research and Treatment (IRCCS) in Ancona, Italy, have developed a machine learning framework using radiomics to predict hospitalization length in COVID-19 patients. The team extracted quantitative imaging features from CT scans and applied a machine learning algorithm to accurately predict hospitalization duration. The study demonstrated that radiomic features can be used to build a machine learning framework for predicting patient hospitalization duration, with the linear support vector machine (LSVM) classifier achieving an accuracy of 86.0% and an area under the receiver operating characteristic curve (AUC-ROC) of 0.93.

Key Takeaways:

  • The study demonstrates the potential of artificial intelligence and radiomics in predicting hospitalization length in COVID-19 patients.
  • The researchers used a machine learning framework to extract quantitative imaging features from CT scans and predict hospitalization duration.
  • The linear support vector machine (LSVM) classifier achieved an accuracy of 86.0% and an area under the receiver operating characteristic curve (AUC-ROC) of 0.93.
  • The study showed that radiomic features can be used to build a machine learning framework for predicting patient hospitalization duration.
  • The research was conducted by a team from the Institute for Cancer Research and Treatment (IRCCS) in Ancona, Italy, and supported by the Italian Ministry of Health.
  • The findings suggest that artificial intelligence and radiomics can be used to improve patient management and resource allocation in hospitals.

Statistics:

  • 168 COVID-19 patients were included in the study.
  • 92 intensity and texture radiomics features were extracted from the CT scans.
  • The LSVM classifier achieved an accuracy of 86.0% and an AUC-ROC of 0.93.
  • The study was published in the journal Bioengineering, with a temporary DOI of 10.3390/bioengineering12040368.

Sources:

  • "Predicting Hospitalization Length in Geriatric Patients Using Artificial Intelligence and Radiomics." Bioengineering, 2025, 12(4):368. (Bioengineering - http://www.mdpi.com/journal/bioengineering).
  • "Research from Institute for Cancer Research and Treatment (IRCCS) Provide New Insights into Artificial Intelligence (Predicting Hospitalization Length in Geriatric Patients Using Artificial Intelligence and Radiomics)." Virus Weekly. May 13, 2025; p 270.