Hakan Ayyıldız, Seda Arslan Tuncer, Mehmet Kalaycı, Rojda Aslan

İdrar Yolu Enfeksiyonu olan Yoğun Bakım Hastalarında Biyokimya Parametreleri Kullanarak Makine Öğrenmesi ile Gram boyama tahmini

Amaç: Bu çalışma, yoğun bakımda yatan ve idrar yolu enfeksiyonu bulunan hastalarda başlangıç tedavi seçimine rehberlik etmek için Gram boyamaya alternatif olabilecek, biyokimya tetkikleri ile oluşturulmuş makine öğrenmesi algoritmalarının etkililiğine odaklanmıştır. Yöntem: Çalışmaya alınan örnekler laboratuvardan Gram boyama isteği yapılmış ve Gram boyamaya eş zamanlı TİT, CBC ve CRP isteği yapılmış 203 yoğun bakım hastasından geriye yönelik oluşturulmuştur. Çalışmada, biyokimya laboratuvar parametreleri kullanarak Gram boyamaya alternatif, Gram-negatif ve Gram-pozitif bakteri tahmini için sınıflandırma yapılmıştır. Bulgular: Sınıflandırmada ANN (Artificial Neural Networks), SVM (Support vector machine), KNN (K-Nearest Neighbors), DTL (Decision Tree Learning) gibi birçok sınıflandırıcı kullanılmış olup, en yüksek sınıflandırma başarısına ANN ile ulaşılmıştır. Sınıflandırma sonucu ANN ile %84.6 Acc, %88.5 Sn, %73.5 Sp değerlerine ulaşılmıştır. Sonuç: Özellikle algoritmanın yüksek özgüllüğü ve doğruluğu, yoğun bakım hastalarında idrar yolu enfeksiyonu ampirik antibiyotik tedavisi seçiminde kullanılabileceğini, laboratuvar testlerinin makine öğrenmesi metotları ile harmanlanarak daha gelişmiş ve teknolojik imkanlar sağlayabileceğini göstermiştir.

Gram Stain Prediction with Machine Learning Techniques Using Biochemical Parameters in ICU Patients with Urinary Tract Infection

Purpose: The aim of this study was to develop a useful algorithm based on complete blood count (CBC), urinalysis, and biochemical parameters that could be an alternative to Gram staining in the prediction of UTI and the determination of initial antibiotic treatment in ICU patients. Methods: All the specimens included in the study were obtained from ICU patients and were subjected to Gram staining in the laboratory. Simultaneously, CBC, urinalysis, and biochemical tests were performed for each specimen. A classification based on biochemical parameters was performed for the estimation of Gram-negative and Gram-positive bacteria, as an alternative to Gram staining. Results: Classification was achieved using multiple classification systems including Artificial Neural Networks (ANN), Support Vector Machine (SVM), the K-Nearest Neighbors (KNN), and Decision Tree Language (DTL) and the best classification performance was achieved by ANN, with an accuracy of 84.6%, a sensitivity of 88.5%, and a specificity of 73.5%. Conclusion: The high specificity and accuracy of the algorithm indicated that this algorithm can be effectively used in the selection of empirical antibiotic treatment for ICU patients with UTI and can provide more advanced and technological opportunities by combining laboratory parameters with machine learning techniques.

PDF

___

1. Saint S, Chenoweth CE. Biofi lms and catheter-associated urinary tract infections. Infect Dis Clin North Am. 2003; 17: 411-32.
2. Warren JW. The catheter and urinary tract infection. Med Clin North Am. 1991; 75: 481-93
3. Maki DG, Tambyah PA. Engineering out the risk for infection with urinary catheters. Emerg Infect Dis. 2001; 7: 342-7.
4. Liu Y, Xiao D, Shi XH. Urinary tract infection control in intensive care patients. Medicine (Baltimore). 2018; 97(38): e12195.
5. Richards MJ, Edwards JR, Culver DH, et al. Nosocomial infections in combined medical-surgical care units in the United States. Infect Control Hosp Epidemiol. 2000; 21: 510-15.
6. Liu VX, Fielding-Singh V, Greene JD, et al. The Timing of Early Antibiotics and Hospital Mortality in Sepsis. Am J Respir Crit Care Med. 2017; 196(7): 856-63.
7. Khilnani GC, Zirpe K, Hadda V, et al. Guidelines for Antibiotic Prescription in Intensive Care Unit. Indian J Crit Care Med. 2019; 23(Suppl 1): 1-63
8. Shlaes DM, Gerding DN, John Jr JF, et al. Society for Healthcare Epidemiology of America and Infectious Diseases Society of America Joint Committee on the Prevention of Antimicrobial Resistance: guidelines for the prevention of antimicrobial resistance in hospitals. Infect Control Hosp Epidemiol. 1997; 18:275-91.
9. Kaki R, Elligsen M, Walker S, et al. Impact of antimicrobial stewardship in critical care: a systematic review. J Antimicrob Chemother. 2011; 66:1223–30.
10. WHO. Global action plan on antimicrobial resistance. 2015. http:// apps.who.int/iris/bitstream/10665/193736/1/9789241509763_ eng.pdf. Acsessed 1 Nov 2018.
11. Kavvas ES, Catoiu E, Mih N, et al. Machine learning and structural analysis of Mycobacterium tuberculosis pan-genome identifies genetic signatures of antibiotic resistance. Nat Commun. 2018; 9(1): 4306.
12. Richardson A, Signor BM, Lidbury BA, et al. Clinical chemistry in higher dimensions: Machine-learning and enhanced prediction from routine clinical chemistry data. Clin. Biochem. 2016; 49; 1213-20.
13. Wildenhain J, Spitzer M, Dolma S, et al. Prediction of Synergism from Chemical-Genetic Interactions by Machine Learning. Cell Syst. 2015;1(6):383-95
14. Congjie He, Meng Ma, Ping Wang. Extract InterpretabilityAccuracy balanced Rules from Artificial Neural Networks: A Review. Neurocomputing. 2020; 387: 346-58.
15. Xu M, Papageorgiou DP, Abidi SZ, et al. A deep convolutional neural network for classification of red blood cells in sickle cell anemia. PLoS Comput Biol. 2017; 13(10): e1005746.
16. H Ayyıldız, SA Tuncer. Determination of the effect of red blood cell parameters in the discrimination of iron deficiency anemia and beta thalassemia via Neighborhood Component Analysis Feature Selection-Based machine learning Chemometrics and Intelligent Laboratory Systems 196, 103886,2020
17. Munita JM, Arias CA. Mechanisms of Antibiotic Resistance. Microbiol Spectr. 2016 Apr;4(2):10.1128/microbiolspec.VMBF-0016-2015.
18. Yoshimura J, Kinoshita T, Yamakawa K, et al. Impact of Gram stain results on initial treatment selection in patients with ventilatorassociated pneumonia: a retrospective analysis of two treatment algorithms. Crit Care. 2017; 21(1):156.
19. O’Horo JC, Thompson D, Safdar N. Is the Gram stain useful in the microbiologic diagnosis of VAP? A meta-analysis. Clin Infect Dis. 2012; 55:551–61.
20. Tetenta S, Metersky ML. Tracheal aspirate Gram stain has limited sensitivity and specificity for detecting Staphylococcus aureus. Respirology. 2011; 16:86–9.
21. Sartika IN, Suarta K, Ardhani P. Diagnostic value of urine Gram staining for urinary tract infection in children. Paediatrica Indonesiana 2009; 49(4): 205-8.
22. Díaz-Martín A, Martínez-González ML, et al. Antibiotic prescription patterns in the empiric therapy of severe sepsis: combination of antimicrobials with different mechanisms of action reduces mortality. Crit Care. 2012; 16(6): R223.
23. Tseng CC, Huang KT, Chen YC, et al. Factors predicting ventilator dependence in patients with ventilator-associated pneumonia. Scientific World Journal. 2012; 2012: 547241
24. Çelikel T. Sepsis: Genel Bakış. Yoğun Bakım Dergisi 2005;5(2):73-74.
25. Wilson ML, Gaido L. Laboratory diagnosis of urinary tract infections in adult patients. Clin Infect Dis. 2004; 38(8): 1150-8
26. Urmi U L, Jahan N, Nahar S, et al. Gram-positive uropathogens: Empirical treatment and emerging antimicrobial resistance. Biomed Res Clin Prac 2019; 4: 1-4.
27. Sever C, Kulahci Y, Duman H. Prediction of mortality and causes of death in a burn centre: a retrospective clinical study. J Clin Anal Med 2011;2(3): p. 24-6
28. Ak O, Batirel A, Ozer S, Çolakoğlu S. Nosocomial infections and risk factors in the intensive care unit of a teaching and research hospital: a prospective cohort study. Med Sci Monit. 2011; 17(5): PH29-34
29. https://www.newscientist.com/article/2216418-hans-christiangram-the-biologist-who-helped-investigate-bacteria/