Machine Learning Based Automated Detection of Sepsis: A Study
Keywords:
Decision Tree, K Nearest Neighbors, Machine Learning, Random Forest, sepsisAbstract
Sepsis is a serious condition that occurs when the body's response to infection leads to tissue damage, organ failure, or death. Human body generally releases chemicals into the blood vessels to fight any kind of infection. Approximately 30 million people suffer from sepsis and 6 million people die from sepsis every year according to WHO. Around 4 million 200 thousand new-born and children are affected by sepsis all over the world. Early prediction of sepsis and treatment are necessary for improving sepsis results where every hour of deferred treatment has been related with approximately a 4–8% expansion in mortality. In response to the PhysioNet/CinC Challenge 2019, many researchers developed several early detection techniques for sepsis. In this work, the various early detection techniques are discussed and how they work is explained. We have studied and compared techniques developed using Machine Learning in response to PhysioNet/CinC Challenge 2019 for early prediction of sepsis using clinical data. Apart from that we have presented three different methods to detect sepsis early, using the data taken from PhysioNet/CinC Challenge 2019 based on Random Forest, Decision Tree and K Nearest Neighbors.
References
Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Hotchkiss RS, et al. The 3rd international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA. 2016; 315(8): 801–810.
WHO. (2020 Aug 26). Sepsis. [Online]. https://www.who.int/news-room/fact-sheets/detail/sepsis. (Accessed on 06/03/2021).
Reyna Matthew A, et al. Early prediction of sepsis from clinical data: the PhysioNet/Computing in Cardiology Challenge 2019. Crit Care Med. 2019; 48(2): 210–217.
Fu Mengsha, et al. An Ensemble Machine Learning Model for the Early Detection of Sepsis From Clinical Data. 2019 Computing in Cardiology (CinC), IEEE. 2019; 1–4. 5. Vadim Smolyakov. (2017 Sep). Ensemble Learning to Improve Machine Learning Results. [Online]. https://blog.statsbot.co/ensemble-learning-d1dcd548e936. last accessed 2020/04/19.
Firoozabadi Reza, Saeed Babaeizadeh. An Ensemble of Bagged Decision Trees for Early Prediction of Sepsis. 2019 Computing in Cardiology (CinC), IEEE. 2019; 1–4.
Zabihi Morteza, Serkan Kiranyaz, Moncef Gabbouj. Sepsis Prediction in Intensive Care Unit Using Ensemble of XGboost Models. 2019 Computing in Cardiology (CinC), IEEE. 2019; 1–4.
GitHub. Chasedehan/BoostARoota: A fast xgboost feature selection algorithm. https://github.com/chasedehan/BoostARoota. [Online]. last accessed 2020/04/20.
Chen Tianqi, Carlos Guestrin. Xgboost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. 2016; 785–794.
Wang Yongchao, et al. Prediction of Sepsis from Clinical Data Using LSTM and XGBoost. 2019 Computing in Cardiology Conference. 2019.
Liu Luchen, et al. Early Prediction of Sepsis from Clinical Datavia Heterogeneous Event Aggregation. arXiv preprint arXiv:1910.06792. 2019.
Hammoud Ibrahim, Ramakrishnan IV, Mark Henry. Early Prediction of Sepsis using Gradient Boosting Decision Trees with Custom Label Weighting. 2019 Computing in Cardiology Conference. 2019.
Ke Guolin, et al. Lightgbm: A highly efficient gradient boosting decision tree. Advances in neural information processing systems. 2017; 3149–3157.
Gilbertson Erik H, et al. Early Detection of Sepsis Using Feature Selection, Feature Extraction, and Neural Network Classification. 2019 Computing in Cardiology (CinC), IEEE. 2019; 1–4.
Pimentel Marco AF, et al. Uncertainty-Aware Model for Reliable Prediction of Sepsis in the ICU. 2019 Computing in Cardiology (CinC), IEEE. 2019; 1–4.
