Sepsis remains a costly and prevalent syndrome in hospitals; however, machine learning systems together with signal processing technique can increase sepsis detection and prediction. This study validates artificial neural networks (ANN) machine­learning tool and nonlinear signal analysis using empirical mode decomposition to extract features (EMD) for sepsis detection and early semi­prediction.
A dataset was drawn from the database at National Taiwan University of Science and Technology in join research with Taipei Medical University representing pig object measurement from February 2012 to May 2014 in all experiments. This database contains medical measurement data of forty pigs and includes the given sepsis diagnosis, and with at least one recording of each of six vital signs (heart rate, systolic blood pressure, diastolic blood pressure, mean arterial pressure, body temperature, and systemic vascular resistance) were included.
Two separate models were constructed using the six vital signs. First, the complete model which consist of six vital signs. Second, the hypertension model which consist of SVR and blood pressure. The area under the receiver operating characteristic (AUC) curve was our primary measure of accuracy. These two models achieved high performance in sepsis detection that seen from the area under the receiver operating characteristic (AUC) curves were 0.92 (0.88­0.95) and 0.86 (0.80­0.92), respectively. Furthermore, we constructed the performance of the machine learning algorithm(ANN) to early semi­prediction task of sepsis. ANN performance was measured at 0.5 hours to 2 hours prior to sepsis onset and achieved an AUC in range 0.95 (0.93­0.97) to 0.83 (0.72­0.93). The ANN predicts sepsis up to 2 hours in advance and identifies sepsis onset had competed accurately with well­known previous sepsis prediction studies, also maintaining high performance for both sepsis detection and early semi­prediction.

[1] D. C. Angus, W. T. Linde­Zwirble, J. Lidicker, G. Clermont, J. Carcillo, and M. R. Pinsky, “Epidemiology of severe sepsis in the united states: analysis of incidence, outcome, and associated costs of care,” Critical care medicine, vol. 29, no. 7, pp. 1303–1310, 2001.
[2] Y.­J. Chen, F.­L. Chen, J.­H. Chen, M.­T. M. Wu, Y.­L. Chen, D.­S. Chien, and Y. Ko, “Epidemiology of sepsis in taiwan,” Medicine, vol. 98, no. 20, 2019.
[3] H.­N. Shen, C.­L. Lu, and H.­H. Yang, “Epidemiologic trend of severe sepsis in taiwan from 1997 through 2006,” Chest, vol. 138, no. 2, pp. 298–304, 2010.
[4] R. S. Hotchkiss, L. L. Moldawer, S. M. Opal, K. Reinhart, I. R. Turnbull, and J.­L. Vincent, “Sepsis
and septic shock,” Nature reviews Disease primers, vol. 2, no. 1, pp. 1–21, 2016.
[5] N. Dey, A. S. Ashour, W. S. Mohamed, and N. G. Nguyen, “Biomedical signals,” in Acoustic Sensors
for Biomedical Applications, pp. 7–20, Springer, 2019.
[6] G. Bianchi and R. Sorrentino, Electronic filter simulation & design. McGraw Hill Professional, 2007.
[7] Q. Mao, M. Jay, J. L. Hoffman, J. Calvert, C. Barton, D. Shimabukuro, L. Shieh, U. Chettipally,G. Fletcher, Y. Kerem, et al., “Multicentre validation of a sepsis prediction algorithm using only vital sign data in the emergency department, general ward and icu,” BMJ open, vol. 8, no. 1, p. e017833,
2018.
[8] R. Liu, J. L. Greenstein, S. J. Granite, J. C. Fackler, M. M. Bembea, S. V. Sarma, and R. L. Winslow,
“Data­driven discovery of a novel sepsis pre­shock state predicts impending septic shock in the icu,”
Scientific reports, vol. 9, no. 1, pp. 1–9, 2019.
[9] E. Alickovic, J. Kevric, and A. Subasi, “Performance evaluation of empirical mode decomposition, discrete wavelet transform, and wavelet packed decomposition for automated epileptic seizure detection and prediction,” Biomedical signal processing and control, vol. 39, pp. 94–102, 2018.
[10] J. Cohen, “The immunopathogenesis of sepsis,” Nature, vol. 420, no. 6917, pp. 885–891, 2002.
[11] Y. Yamamoto, A. Harashima, H. Saito, K. Tsuneyama, S. Munesue, S. Motoyoshi, D. Han, T. Watanabe, M. Asano, S. Takasawa, et al., “Septic shock is associated with receptor for advanced glycation end products ligation of lps,” The Journal of Immunology, vol. 186, no. 5, pp. 3248–3257, 2011.
[12] D. J. Funk, J. E. Parrillo, and A. Kumar, “Sepsis and septic shock: a history,” Critical care clinics,vol. 25, no. 1, pp. 83–101, 2009.
[13] F. Gül, M. K. Arslantaş, İ. Cinel, and A. Kumar, “Changing definitions of sepsis,” Turkish journal of anaesthesiology and reanimation, vol. 45, no. 3, p. 129, 2017.
[14] R. C. Bone, R. A. Balk, F. B. Cerra, R. P. Dellinger, A. M. Fein, W. A. Knaus, R. M. Schein, and W. J. Sibbald, “Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis,” Chest, vol. 101, no. 6, pp. 1644–1655, 1992.
[15] M. M. Levy, M. P. Fink, J. C. Marshall, E. Abraham, D. Angus, D. Cook, J. Cohen, S. M. Opal, J.­L. Vincent, G. Ramsay, et al., “2001 sccm/esicm/accp/ats/sis nternational sepsis definitions conference,”Intensive care medicine, vol. 29, no. 4, pp. 530–538, 2003.
[16] M. Singer, C. S. Deutschman, C. W. Seymour, M. Shankar­Hari, D. Annane, M. Bauer, R. Bellomo,G. R. Bernard, J.­D. Chiche, C. M. Coopersmith, et al., “The third international consensus definitions
for sepsis and septic shock (sepsis­3),” Jama, vol. 315, no. 8, pp. 801–810, 2016.
[17] G. Majno, “The ancient riddle of σñψις (sepsis),” The Journal of infectious diseases, vol. 163, no. 5, pp. 937–945, 1991.
[18] C. J. Schuh, “Sepsis and septic shock analysis using neural networks,” in NAFIPS 2007­2007 Annual Meeting of the North American Fuzzy Information Processing Society, pp. 650–654, IEEE, 2007.
[19] J. C. Ho, C. H. Lee, and J. Ghosh, “Imputation­enhanced prediction of septic shock in icu patients,” in Proceedings of the ACM SIGKDD workshop on health informatics (HI­KDD12), p. 18, 2012.
[20] J. Guillén, J. Liu, M. Furr, T. Wang, S. Strong, C. C. Moore, A. Flower, and L. E. Barnes, “Predictive models for severe sepsis in adult icu patients,” in 2015 Systems and Information Engineering Design Symposium, pp. 182–187, IEEE, 2015.
[21] K. E. Henry, D. N. Hager, P. J. Pronovost, and S. Saria, “A targeted real­time early warning score(trewscore) for septic shock,” Science translational medicine, vol. 7, no. 299, pp. 299ra122–299ra122,2015.
[22] T. Desautels, J. Calvert, J. Hoffman, M. Jay, Y. Kerem, L. Shieh, D. Shimabukuro, U. Chettipally, M. D. Feldman, C. Barton, et al., “Prediction of sepsis in the intensive care unit with minimal electronic health record data: a machine learning approach,” JMIR medical informatics, vol. 4, no. 3, p. e28,2016.
[23] L. Spelt, B. Andersson, J. Nilsson, and R. Andersson, “Prognostic models for outcome following liver resection for colorectal cancer metastases: a systematic review,” European Journal of Surgical Oncology (EJSO), vol. 38, no. 1, pp. 16–24, 2012.
[24] D. Mortazavi, A. Z. Kouzani, and H. Soltanian­Zadeh, “Segmentation of multiple sclerosis lesions in mr images: a review,” Neuroradiology, vol. 54, no. 4, pp. 299–320, 2012.
[25] F. E. Ahmed, “Artificial neural networks for diagnosis and survival prediction in colon cancer,” Molecular cancer, vol. 4, no. 1, p. 29, 2005.
[26] A. Bartosch­Härlid, B. Andersson, U. Aho, J. Nilsson, and R. Andersson, “Artificial neural networks in pancreatic disease,” British Journal of Surgery: Incorporating European Journal of Surgery and Swiss Surgery, vol. 95, no. 7, pp. 817–826, 2008.
[27] C. S. Siristatidis, C. Chrelias, A. Pouliakis, E. Katsimanis, and D. Kassanos, “Artificial neural networks in gynaecological diseases: Current and potential future applications,” Medical Science Monitor, vol. 16, no. 10, pp. RA231–RA236, 2010.
[28] D. O. Shankaracharya, S. Samanta, and A. S. Vidyarthi, “Computational intelligence in early diabetes diagnosis: a review,” The review of diabetic studies: RDS, vol. 7, no. 4, p. 252, 2010.
[29] M. MATLAB, “Matlab r2018b,” The MathWorks: Natick, MA, USA, 2018.
[30] F. Amato, A. López, E. M. Peña­Méndez, P. Vaňhara, A. Hampl, and J. Havel, “Artificial neural networks in medical diagnosis,” Journal of Applied Biomadicine, vol. 11, no. 2, pp. 47–58, 2013.
[31] S. R. Wisniewski, A. C. Leon, M. W. Otto, and M. H. Trivedi, “Prevention of missing data in clinical research studies,” Biological psychiatry, vol. 59, no. 11, pp. 997–1000, 2006.
[32] H. Kang, “The prevention and handling of the missing data,” Korean journal of anesthesiology, vol. 64,no. 5, p. 402, 2013.
[33] R. M. Hamer and P. M. Simpson, “Last observation carried forward versus mixed models in the anal-
ysis of psychiatric clinical trials,” American Journal of Psychiatry, vol. 166, no. 6, pp. 639–641, 2009.
[34] V. López, A. Fernández, S. García, V. Palade, and F. Herrera, “An insight into classification with imbalanced data: Empirical results and current trends on using data intrinsic characteristics,” Information sciences, vol. 250, pp. 113–141, 2013.
[35] A. Ghasemi and S. Zahediasl, “Normality tests for statistical analysis: a guide for non­statisticians,”International journal of endocrinology and metabolism, vol. 10, no. 2, p. 486, 2012.
[36] J. Peat and B. Barton, Medical statistics: A guide to data analysis and critical appraisal. John Wiley & Sons, 2008.
[37] A. BenSaïda, “Shapiro­wilk and shapiro­francia normality tests
(https://www.mathworks.com/matlabcentral/fileexchange/13964­shapiro­wilk­and­shapiro­francia-normality­tests),” MATLAB Central File Exchange. Retrieved May 15, 2020.
[38] T. K. Kim, “T test as a parametric statistic,” Korean journal of anesthesiology, vol. 68, no. 6, p. 540,2015.
[39] L. M. Fleuren, T. L. Klausch, C. L. Zwager, L. J. Schoonmade, T. Guo, L. F. Roggeveen, E. L. Swart,A. R. Girbes, P. Thoral, A. Ercole, et al., “Machine learning for the prediction of sepsis, a systematic review and meta­analysis of diagnostic test accuracy,” Intensive Care Medicine, vol. 46, no. 3, pp. 383–400, 2020.
[40] D. W. Shimabukuro, C. W. Barton, M. D. Feldman, S. J. Mataraso, and R. Das, “Effect of a machine learning­based severe sepsis prediction algorithm on patient survival and hospital length of stay: a randomised clinical trial,” BMJ open respiratory research, vol. 4, no. 1, p. e000234, 2017.
[41] S. Nemati, A. Holder, F. Razmi, M. D. Stanley, G. D. Clifford, and T. G. Buchman, “An interpretable machine learning model for accurate prediction of sepsis in the icu,” Critical care medicine, vol. 46,no. 4, pp. 547–553, 2018.
[42] J. Young, “The heart and circulation in severe sepsis,” British journal of anaesthesia, vol. 93, no. 1,pp. 114–120, 2004.
[43] N. E. Huang, Z. Shen, S. R. Long, M. C. Wu, H. H. Shih, Q. Zheng, N.­C. Yen, C. C. Tung, and H. H.Liu, “The empirical mode decomposition and the hilbert spectrum for nonlinear and non­stationary time series analysis,” Proceedings of the Royal Society of London. Series A: mathematical, physical and engineering sciences, vol. 454, no. 1971, pp. 903–995, 1998.
[44] P. Flandrin, G. Rilling, and P. Goncalves, “Empirical mode decomposition as a filter bank,” IEEE signal processing letters, vol. 11, no. 2, pp. 112–114, 2004.
[45] S. Li, W. Zhou, Q. Yuan, S. Geng, and D. Cai, “Feature extraction and recognition of ictal eeg using emd and svm,” Computers in biology and medicine, vol. 43, no. 7, pp. 807–816, 2013.
[46] U. R. Acharya, F. Molinari, S. V. Sree, S. Chattopadhyay, K.­H. Ng, and J. S. Suri, “Automated diagnosis of epileptic eeg using entropies,” Biomedical Signal Processing and Control, vol. 7, no. 4,pp. 401–408, 2012.
[47] M. Lozano, J. A. Fiz, and R. Jané, “Estimation of instantaneous frequency from empirical mode decomposition on respiratory sounds analysis,” in 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 981–984, IEEE, 2013.
[48] N. Zhuang, Y. Zeng, L. Tong, C. Zhang, H. Zhang, and B. Yan, “Emotion recognition from eeg signals using multidimensional information in emd domain,” BioMed research international, vol. 2017, 2017.
[49] M. T. Hagan and M. B. Menhaj, “Training feedforward networks with the marquardt algorithm,” IEEE transactions on Neural Networks, vol. 5, no. 6, pp. 989–993, 1994.
[50] H. Yu and B. M. Wilamowski, “Levenberg­marquardt training,” Industrial electronics handbook,vol. 5, no. 12, p. 1, 2011.
[51] S. Roweis, “Levenberg­marquardt optimization,” Notes, University Of Toronto, 1996.
[52] D. W. Marquardt, “An algorithm for least­squares estimation of nonlinear parameters,” Journal of the society for Industrial and Applied Mathematics, vol. 11, no. 2, pp. 431–441, 1963.
[53] M. H. Zweig and G. Campbell, “Receiver­operating characteristic (roc) plots: a fundamental evaluation tool in clinical medicine,” Clinical chemistry, vol. 39, no. 4, pp. 561–577, 1993.
[54] H. J. Kam and H. Y. Kim, “Learning representations for the early detection of sepsis with deep neural networks,” Computers in biology and medicine, vol. 89, pp. 248–255, 2017.
[55] C. Barton, U. Chettipally, Y. Zhou, Z. Jiang, A. Lynn­Palevsky, S. Le, J. Calvert, and R. Das, “Eval-
uation of a machine learning algorithm for up to 48­hour advance prediction of sepsis using six vital signs,” Computers in biology and medicine, vol. 109, pp. 79–84, 2019.
[56] J. S. Calvert, D. A. Price, U. K. Chettipally, C. W. Barton, M. D. Feldman, J. L. Hoffman, M. Jay, and R. Das, “A computational approach to early sepsis detection,” Computers in biology and medicine,vol. 74, pp. 69–73, 2016.
[57] M. Scherpf, F. Gräßer, H. Malberg, and S. Zaunseder, “Predicting sepsis with a recurrent neural network using the mimic iii database,” Computers in biology and medicine, vol. 113, p. 103395, 2019.