A correlation matrix-based tensor decomposition method for early prediction of sepsis from clinical data

• Autorzy: Nesaragi Naimahmed , Patidar Shivnarayan , Thangaraj Veerakumar

Identyfikatory

DOI

10.1016/j.bbe.2021.06.009

• Języki publikacji: EN

Abstrakty

EN

Early detection of sepsis can assist in clinical triage and decision-making, resulting in early intervention with improved outcomes. This study aims to develop a machine learning framework to predict the onset of sepsis through EHR data by applying tensor decomposition on correlation matrices of clinical covariates for every record, arranged on an hourly basis for the length of stay (LOS) in intensive care unit. A third-order tensor [...] representing a clinical correlation among selected 24 covariates for a considered time frame of sepsis onset duration of 6 h, with a stride of 1 h is formed for each record. Such a fused tensor with dimensions [...] for every record undergoes Tucker decomposition with an optimal choice of rank. The factor matrices U₁; U₂; U₃ thus obtained after decomposition are excluded and only the core tensor r with a dimension [...] has been retained, and used to provide latent features for prediction of sepsis onset. A five-fold cross-validation scheme is employed wherein the obtained 100 latent features from the reshaped core tensor, are fed to Light Gradient Boosting Machine Learning models (LightGBM) for binary classification, further alleviating the involved class imbalance. The machine-learning framework is designed via Bayesian optimization, yielding an average normalized utility score of 0.4314 on publicly available PhysioNet/Computing in Cardiology Challenge 2019 training data. The proposed tensor decomposition deciphers the higher-order interrelations among the considered clinical covariates for early prediction of sepsis and the results obtained are on par with existing state-of-the-art performances.

Słowa kluczowe

EN

sepsis; machine learning; tensor factorization; early prediction; model-based diagnosis; PhysioNet

PL

sepsa; uczenie maszynowe; PhysioNet

• Wydawca: Nałęcz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences ; Elsevier
• Czasopismo: Biocybernetics and Biomedical Engineering
• Rocznik: 2021
• Tom: Vol. 41, no. 3
• Strony: 1013--1024
• Opis fizyczny: Bibliogr. 74 poz., rys., tab.

Twórcy

autor

Nesaragi Naimahmed

• Dept. of Electronics and Communication Engineering, National Institute of Technology Goa, Goa, India

autor

Patidar Shivnarayan

• National Institute of Technology Goa, Farmagudi, Ponda, Goa 403401, India

autor

Thangaraj Veerakumar

• Dept. of Electronics and Communication Engineering, National Institute of Technology Goa, Goa, India

Bibliografia

• [1] Papalexakis EE, Faloutsos C, Sidiropoulos ND. Tensors for data mining and data fusion: Models, applications, and scalable algorithms. ACM Trans Intell Syst Technol (TIST) 2016;8:1–44.
• [2] Luo Y, Xin Y, Hochberg E, Joshi R, Uzuner O, Szolovits P. Subgraph augmented non-negative tensor factorization (santf) for modeling clinical narrative text. J Am Med Inf Assoc 2015;22:1009–19.
• [3] Rabanser S, Shchur O, Günnemann S. Introduction to tensor decompositions and their applications in machine learning; 2017. arXiv preprint arXiv:1711.10781.
• [4] Chi EC, Kolda TG. On tensors, sparsity, and nonnegative factorizations. SIAM J Matrix Anal Appl 2012;33:1272–99.
• [5] Kolda TG, Bader BW. Tensor decompositions and applications. SIAM Rev 2009;51:455–500.
• [6] Dunlavy DM, Kolda TG, Acar E. Temporal link prediction using matrix and tensor factorizations. ACM Trans Knowl Discovery Data (TKDD) 2011;5:1–27.
• [7] Kim Y, El-Kareh R, Sun J, Yu H, Jiang X. Discriminative and distinct phenotyping by constrained tensor factorization. Sci Rep 2017;7:1–12.
• [8] Ho JC, Ghosh J, Sun J. Marble: high-throughput phenotyping from electronic health records via sparse nonnegative tensor factorization. In: Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. pp. 115–24.
• [9] Wang H, Zhang Q, Yuan J. Semantically enhanced medical information retrieval system: a tensor factorization based approach. IEEE Access 2017;5:7584–93.
• [10] Luo Y, Wang F, Szolovits P. Tensor factorization toward precision medicine. Brief Bioinf 2017;18:511–4.
• [11] Yang Y, Dunson DB. Bayesian conditional tensor factorizations for high-dimensional classification. J Am Stat Assoc 2016;111:656–69.
• [12] Wang H, Zhang Q, So H-Y, Kwok A, Wong ZS-Y. Temporal prediction of in-hospital falls using tensor factorisation. BMJ Innov 2018;4.
• [13] Goldfarb D, Qin Z. Robust low-rank tensor recovery: Models and algorithms. SIAM J Matrix Anal Appl 2014;35:225–53.
• [14] Yach D, Hawkes C, Gould CL, Hofman KJ. The global burden of chronic diseases: overcoming impediments to prevention and control. Jama 2004;291:2616–22.
• [15] Vasconcelos CN, Vasconcelos BN. Experiments using deep learning for dermoscopy image analysis. Pattern Recognit Lett 2017.
• [16] Kohli R, Tan SS-L. Electronic health records: how can is researchers contribute to transforming healthcare? Mis Quart 2016;40:553–73.
• [17] Shmueli G, Koppius OR. Predictive analytics in information systems research. MIS Quart 2011:553–72.
• [18] He K, Huang S, Qian X. Early detection and risk assessment for chronic disease with irregular longitudinal data analysis. J Biomed Inf 2019;96 103231.
• [19] Ashley EA. The precision medicine initiative: a new national effort. Jama 2015;313:2119–20.
• [20] Collins FS, Varmus H. A new initiative on precision medicine. New Engl J Med 2015;372:793–5.
• [21] Mirnezami R, Nicholson J, Darzi A. Preparing for precision medicine. N Engl J Med 2012;366:489–91.
• [22] Finkelstein J, cheol Jeong I. Machine learning approaches to personalize early prediction of asthma exacerbations. Ann NY Acad Sci 2017;1387:153.
• [23] Leung RK, Wang Y, Ma RC, Luk AO, Lam V, Ng M, So WY, Tsui SK, Chan JC. Using a multi-staged strategy based on machine learning and mathematical modeling to predict genotype-phenotype risk patterns in diabetic kidney disease: a prospective case-control cohort analysis. BMC Nephrol 2013;14:1–9.
• [24] Giesen L, Singer M. What is sepsis? In: Wiersinga WJ, Seymour CW, editors. Handbook of Sepsis. Cham: Springer International Publishing; 2018. p. 3–14.
• [25] Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Bellomo R, Bernard GR, Chiche J-D, Coopersmith CM, Hotchkiss RS, Levy MM, Marshall JC, Martin GS, Opal SM, Rubenfeld GD, van der Poll T, Vincent J-L, Angus DC. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016;315:801.
• [26] Potey PM, Rossi AG, Lucas CD, Dorward DA. Neutrophils in the initiation and resolution of acute pulmonary inflammation: understanding biological function and therapeutic potential. J Pathol 2019;247:672–85.
• [27] Bergmann CB, Salyer CE, Beckmann N, Caldwell CC. Intraperitoneal neutrophil il-10 production is promoted by interferon c) in a murine model of sepsis model in the acute phase of sepsis. Biochem Biophys Res Commun 2020;530:278–84.
• [28] Ahn SY, Maeng Y-S, Kim YR, Choe YH, Hwang HS, Hyun Y-M. In vivo monitoring of dynamic interaction between neutrophil and human umbilical cord blood-derived mesenchymal stem cell in mouse liver during sepsis. Stem Cell Res Therapy 2020;11:44.
• [29] Seree-Aphinan C, Vichitkunakorn P, Navakanitworakul R, Khwannimit B. Distinguishing sepsis from infection by neutrophil dysfunction: A promising role of cxcr2 surface level. Front Immunol 2020;11:3339.
• [30] Shen X-F, Cao K, Jiang J-P, Guan W-X, Du J-F. Neutrophil dysregulation during sepsis: an overview and update. J Cell Mol Med 2017;21:1687–97.
• [31] Brown K, Brain S, Pearson J, Edgeworth J, Lewis S, Treacher D. Neutrophils in development of multiple organ failure in sepsis. The Lancet 2006;368:157–69.
• [32] Groeneveld KM, Koenderman L, Warren BL, Jol S, Leenen LP, Hietbrink F. Early decreased neutrophil responsiveness is related to late onset sepsis in multitrauma patients: an international cohort study. PLoS One 2017;12 e0180145.
• [33] Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock. Intensive Care Med 2012;39(2013):165–228.
• [34] Seymour CW, Kennedy JN, Wang S, Chang C-CH, Elliott CF, Xu Z, Berry S, Clermont G, Cooper G, Gomez H, Huang DT, Kellum JA, Mi Q, Opal SM, Talisa V, van der Poll T, Visweswaran S, Vodovotz Y, Weiss JC, Yealy DM, Yende S, Angus DC. Derivation, Validation, and Potential Treatment Implications of Novel Clinical Phenotypes for Sepsis. JAMA 2019;321:2003.
• [35] Vincent J-L, Moreno R, Takala J, Willatts S, De Mendonc¸a A, Bruining H, Reinhart C, Suter P, Thijs LG. The sofa (sepsis-related organ failure assessment) score to describe organ dysfunction/failure; 1996.
• [36] Henry KE, Hager DN, Pronovost PJ, Saria S. A targeted realtime early warning score (trewscore) for septic shock. Sci Translat Med 2015;7:299ra122–299ra122.
• [37] Horng S, Sontag DA, Halpern Y, Jernite Y, Shapiro NI, Nathanson LA. Creating an automated trigger for sepsis clinical decision support at emergency department triage using machine learning. PLOS One 2017;12 e0174708.
• [38] Desautels T, Calvert J, Hoffman J, Jay M, Kerem Y, Shieh L, Shimabukuro D, Chettipally U, Feldman MD, Barton C, Wales DJ, Das R. Prediction of sepsis in the intensive care unit with minimal electronic health record data: a machine learning approach. JMIR Med Inf 2016;4 e28.
• [39] Kawamoto K, Houlihan CA, Balas EA, Lobach DF. Improving clinical practice using clinical decision support systems: a systematic review of trials to identify features critical to success. BMJ 2005;330:765.
• [40] Oh SM, Stefani KM, Kim HC. Development and application of chronic disease risk prediction models. Yonsei Med J 2014;55:853.
• [41] de Castilho FM, Ribeiro ALP, Nobre V, Barros G, de Sousa MR. Heart rate variability as predictor of mortality in sepsis: A systematic review. PloS One 2018;13 e0203487.
• [42] Thakur J, Pahuja SK, Pahuja R. Performance comparison of systemic inflammatory response syndrome with logistic regression models to predict sepsis in neonates. Children 2017;4:111.
• [43] Wardi G, Carlile M, Holder A, Shashikumar S, Hayden SR, Nemati S. Predicting progression to septic shock in the emergency department using an externally generalizable machine-learning algorithm. Ann Emerg Med 2021;77:395–406.
• [44] Wardi G, Shashikumar S, Carlile M, Krak M, Hayden S, Holder A, Nemati S. 301 use of transfer learning to improve external validity of a machine-learning algorithm to predict septic shock in the emergency department. Ann Emerg Med 2020;76:S116.
• [45] Nemati S, Holder A, Razmi F, Stanley MD, Clifford GD, Buchman TG. An interpretable machine learning model for accurate prediction of sepsis in the ICU. Critical Care Med 2018;46:547–53.
• [46] Fleuren LM, Klausch TL, Zwager CL, Schoonmade LJ, Guo T, Roggeveen LF, et al. Machine learning for the prediction of sepsis: a systematic review and meta-analysis of diagnostic test accuracy. Intensive Care Med 2020;46:383–400.
• [47] Lauritsen SM, Kalør ME, Kongsgaard EL, Lauritsen KM, Jørgensen MJ, Lange J, Thiesson B. Early detection of sepsis utilizing deep learning on electronic health record event sequences. Artif Intell Med 2020;104 101820.
• [48] Svenson P, Haralabopoulos G, Torres MT. Sepsis deterioration prediction using channelled long short-term memory networks. In: International Conference on Artificial Intelligence in Medicine, Springer. pp. 359–70.
• [49] Moor M, Horn M, Rieck B, Roqueiro D, Borgwardt K. Early recognition of sepsis with gaussian process temporal convolutional networks and dynamic time warping. In: Machine Learning for Healthcare Conference, PMLR. pp. 2–26.
• [50] Moor M, Horn M, Rieck B, Roqueiro D, Borgwardt K. Temporal convolutional networks and dynamic time warping can drastically improve the early prediction of sepsis; 2019. arXiv preprint arXiv:1902.01659.
• [51] Reyna MA, Josef CS, Jeter R, Shashikumar SP, Westover MB, Nemati S, Clifford GD, Sharma A. Early prediction of sepsis from clinical data: the physionet/computing in cardiology challenge 2019. Critical Care Med 2020;48:210–7 [Publisher: Lippincott Williams & Wilkins].
• [52] Morrill JH, Kormilitzin A, Nevado-Holgado AJ, Swaminathan S, Howison SD, Lyons TJ. Utilization of the signature method to identify the early onset of sepsis from multivariate physiological time series in critical care monitoring. Critical Care Med 2020;48:e976–81.
• [53] Li X, Xu X, Xie F, Xu X, Sun Y, Liu X, et al. A time-phased machine learning model for real-time prediction of sepsis in critical care. Critical Care Med 2020;48:e884–8.
• [54] Zabihi M, Kiranyaz S, Gabbouj M. Sepsis prediction in intensive care unit using ensemble of xgboost models. In: 2019 Computing in Cardiology (CinC). pp. Page 1–Page 4.
• [55] Yang M, Liu C, Wang X, Li Y, Gao H, Liu X, Li J. An explainable artificial intelligence predictor for early detection of sepsis. Critical Care Med 2020;48:e1091–6.
• [56] Nesaragi N, Patidar S. Early prediction of sepsis from clinical data using ratio and power-based features. Critical Care Med 2020;48:e1343–9.
• [57] Goldstein BA, Navar AM, Pencina MJ, Ioannidis J. Opportunities and challenges in developing risk prediction models with electronic health records data: a systematic review. J Am Med Inf Assoc 2017;24:198–208.
• [58] Brochu E, Cora VM, de Freitas N. A tutorial on Bayesian optimization of expensive cost functions, with application to active user modeling and hierarchical reinforcement learning; 2010. arXiv:1012.2599 [cs]. ArXiv: 1012.2599.
• [59] Snoek J, Larochelle H, Adams RP. Practical bayesian optimization of machine learning algorithms. Advances in Neural Information Processing Systems 2012;25:2951–9.
• [60] Biglarbeigi P, McLaughlin D, Rjoob K, Abdullah A, McCallan N, Jasinska-Piadlo A, et al., Early prediction of sepsis considering early warning scoring systems. In: 2019 Computing in Cardiology (CinC). pp. Page 1–Page 4.
• [61] Hillar CJ, Lim L-H. Most tensor problems are np-hard. J ACM (JACM) 2013;60:1–39.
• [62] Kossaifi J, Panagakis Y, Anandkumar A, Pantic M. Tensorly: Tensor learning in python. J Mach Learn Res 2019;20:925–30.
• [63] Delahanty RJ, Alvarez J, Flynn LM, Sherwin RL, Jones SS. Development and evaluation of a machine learning model for the early identification of patients at risk for sepsis. Ann Emerg Med 2019;73:334–44.
• [64] Lyra S, Leonhardt S, Antink CH. Early prediction of sepsis using random forest classification for imbalanced clinical data. In: 2019 Computing in Cardiology (CinC). pp. 1–4.
• [65] Kok C, Jahmunah V, Oh SL, Zhou X, Gururajan R, Tao X, Cheong KH, Gururajan R, Molinari F, Acharya UR. Automated prediction of sepsis using temporal convolutional network. Comput Biol Med 2020;127 103957.
• [66] Chang Y, Rubin J, Boverman G, Vij S, Rahman A, Natarajan A, Parvaneh S. A multi-task imputation and classification neural architecture for early prediction of sepsis from multivariate clinical time series. In: 2019 Computing in Cardiology (CinC). IEEE. pp. Page–1.
• [67] Lee BT, Kwon O-Y, Park H, Cho K-J, Kwon J-M, Lee Y, et al. Graph convolutional networks-based noisy data imputation in electronic health record. Critical Care Med 2020;48: e1106–11.
• [68] Du JA, Sadr N, Chazal Pd. Automated prediction of sepsis onset using gradient boosted decision trees. In: 2019 Computing in Cardiology (CinC). pp. Page 1–Page 4.
• [69] He Z, Du L, Zhang P, Zhao R, Chen X, Fang Z. Early sepsis prediction using ensemble learning with deep features and artificial features extracted from clinical electronic health records. Critical Care Med 2020;48:e1337–42.
• [70] Rafiei A, Rezaee A, Hajati F, Gheisari S, Golzan M. Ssp: Early prediction of sepsis using fully connected lstm-cnn model. Comput Biol Med 2021;128 104110.
• [71] Wang Y, Xiao B, Bi X, Li W, Zhang J, Ma X. Prediction of sepsis from clinical data using long short-term memory and extreme gradient boosting. In: 2019 Computing in Cardiology (CinC). IEEE. pp. Page–1.
• [72] Sweely B, Park A, Winter L, Liu L, Zhao X. Time-padded random forest ensemble to capture changes in physiology leading to sepsis development. In: 2019 Computing in Cardiology (CinC). IEEE. pp. 1–4.
• [73] Bailey E, Aeron S. Word embeddings via tensor factorization; 2017. arXiv preprint arXiv:1704.02686.
• [74] Anandkumar A, Ge R, Hsu D, Kakade SM, Telgarsky M. Tensor decompositions for learning latent variable models (a survey for alt). In: International Conference on Algorithmic Learning Theory. Springer. pp. 19–38.

Uwagi

Opracowanie rekordu ze środków MNiSW, umowa Nr 461252 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2021).