Towards Effective and Reliable Data-driven Prognostication: An Application to COVID-19

Resumo


This study evaluates machine learning methods to predict the prognosis of patients in COVID-19 context. In addition, considering the best-performing machine learning algorithm, we applied the LIME explanation technique for machine learning models to verify how the features correlate with each decision made, in order to assist an expert regarding the groundings of the decision made by the model. The results reveal that the model developed was able to predict the patient’s prognosis with an ROC-AUC = 0.8524. The prediction explanations allowed us to understand how each feature contributes to the decision made by the model, thus bringing transparency to the developed model.
Palavras-chave: COVID-19, Machine Learning, Computer aided prognosis, Mortality prediction, Explainable AI

Referências

Figuerêdo, J. et al. Machine learning for prognosis of patients with covid-19: An early days analysis. In Anais do XVIII ENIAC. SBC, Porto Alegre, RS, Brasil, pp. 59–70, 2021.

Kumar, A. et al. A review of modern technologies for tackling covid-19 pandemic. Diabetes & Metabolic Syndrome: Clinical Research & Reviews 14 (4): 569 – 573, 2020.

Lu, R. et al. Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding. The Lancet 395 (10224): 565–574, 2020.

Mattos, J. et al. Clinical risk factors of icu & fatal covid-19 cases in brazil. In Anais do VIII KDMiLe. SBC, Porto Alegre, RS, Brasil, pp. 33–40, 2020.

Mittelstadt, B. et al. Explaining explanations in ai. In Proceedings of the Conference on FAT. ACM, New York, NY, USA, pp. 279–288, 2019.

Pan, D. et al. A predicting nomogram for mortality in patients with covid-19. Frontiers in Public Health vol. 8, pp. 461, 2020.

Ribeiro, M. T. et al. “why should i trust you”: Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD. ACM, New York, USA, pp. 1135–1144, 2016.

Soares, F. et al. Analysis and prediction of childhood pneumonia deaths using machine learning algorithms. In Anais do IX KDMiLe. SBC, Porto Alegre, RS, Brasil, pp. 16–23, 2021.

Souza, F. S. H. et al. Predicting the disease outcome in covid-19 positive patients through machine learning: a retrospective cohort study with brazilian data. medRxiv, 2020.

White, D. B. and Lo, B. A Framework for Rationing Ventilators and Critical Care Beds During the COVID-19 Pandemic. JAMA 323 (18): 1773–1774, 05, 2020.

WHO. Coronavirus disease 2019 Situation Report. https://covid19.who.int/, 2021. Accessed 01 April 2021.

Xie, J. et al. Association Between Hypoxemia and Mortality in Patients With COVID-19. Mayo Clinic Proceedings 95 (6): 1138–1147, jun, 2020.

Yan, L. et al. An interpretable mortality prediction model for COVID-19 patients. Nat. Mach. Intell. 2 (5): 283–288, may, 2020.

Yu, K.-H. et al. Artificial intelligence in healthcare. Nature Biomedical Engineering 2 (10): 719–731, Oct, 2018.
Publicado
26/09/2023
FIGUERÊDO, José Solenir Lima; ARAUJO-CALUMBY, Renata Freitas; CALUMBY, Rodrigo Tripodi. Towards Effective and Reliable Data-driven Prognostication: An Application to COVID-19. In: SYMPOSIUM ON KNOWLEDGE DISCOVERY, MINING AND LEARNING (KDMILE), 11. , 2023, Belo Horizonte/MG. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2023 . p. 81-88. ISSN 2763-8944. DOI: https://doi.org/10.5753/kdmile.2023.232894.