ANALİTİK HİYERARŞİ SÜRECİ VE VERİ MADENCİLİĞİ TEKNİKLERİYLE HİBRİT BİR KARAR DESTEK SİSTEMİ UYGULAMASI: TAM KAN SAYIMI DEĞERLERİ İLE KOVİD19 TANISI

Öz

Amaç: Veri madenciliği teknikleri, yapay zeka temelli tanı doğruluğunu artırmada önemli bir etkiye sahiptir. Bu araştırmada, hastaneye Kovid-19 şüphesiyle gelen bir kişinin, görüntüleme ve PCR testi sonuçları elde edilene dek, tam kan sayımı sonuçları kullanılarak, Kovid-19 olma durumu hakkında tahminde bulunan bir web tabanlı karar desteği geliştirilmesi amaçlanmıştır. Yöntem: Bu çalışmada öncelikle veri seti üzerinde veri ön işleme teknikleri uygulanmış, daha sonra veri madenciliği yaklaşımları kullanılarak özellik seçimi yapılmıştır. Değişken sayısı azaltıldıktan sonra çok kriterli karar verme yaklaşımının önde gelenlerinden analitik hiyerarşi süreci yöntemi (AHP) kullanılmıştır. Uzman görüşleri ile birleştirilen AHP yöntemiyle makine öğrenmesiyle elde edilen değişkenlerin öncelikleri belirlenmiş ve kamuya açık veriler kullanılarak bir karar modeli geliştirilmiştir. Bu karar modelinin bir web tabanlı uygulaması, daha sonra son kullanıcılara karar destek sistemi sağlamak üzere hazırlanmıştır. Ayrıca, karar destek sisteminin kullanılabilirliğini ve kullanıcı memnuniyetini ölçmek için bir değerlendirme yapılmıştır. Bulgular: RFE-SVM özellik seçim algoritması yedi önemli değişkeni tanımlamıştır: Bazofil, Eozinofil, Lenfosit, Lökosit, Nötrofil, Trombosit ve Monosit. AHP yöntemi ile Kovid-19 tanısına karar vermeyle ilgili farklı uzmanlık alanlarından altı uzman hekim ile görüşülmüştür. 42 uzman kullanıcı (%57.1'i erkek, yaş ortalaması 37.30±10.56) sistemi değerlendirdi. Sistem Kullanılabilirlik Ölçeği (SUS) puanının ortalaması 81.43±15.64 olup, yüksek kullanılabilirliği göstermektedir. Sonuç: Sonuç olarak, bu sistem hastanın daha hızlı izole edilmesini ve ilk tedavisinin başlatılmasını sağlayabilir.

Anahtar Kelimeler

• Kovid-19
• Makine Öğrenimi
• Dengesiz Veri
• Öznitelik Seçimi
• Karar Destek Sistemi
• Analitik Hiyerarşi Süreci (AHP) Yöntemi

A HYBRID DECISION SUPPORT SYSTEM APPLICATION WITH THE ANALYTIC HIERARCHY PROCESS AND DATA MINING TECHNIQUES: DIAGNOSIS OF COVID19 WITH COMPLETE BLOOD COUNT VALUES

Öz

Objective: Data mining techniques have a significant impact on enhancing the precision of diagnostics based on artificial intelligence. In this research, it was aimed to develop a web-based decision support that predicts the status of a person who comes to the hospital with Covid-19 suspicion by using complete blood count results until the imaging and PCR test results are obtained. Method: In this study, firstly data pre-processing techniques on the data set were applied, then feature selection was made using data mining approaches. After reducing the number of variables, the analytical hierarchy process method (AHP), a prominent multi-criteria decision-making approach, was utilized. Through the AHP method combined with expert opinions, the priorities of the variables determined by machine learning were ascertained, leading to the development of a decision model using publicly accessible data. A web-based application of this decision model was subsequently crafted to provide the decision support system to the end-users. Furthermore, an evaluation was conducted to gauge the usability of the decision support system and the satisfaction of its users. Results: RFE-SVM feature selection algorithm identified seven pivotal variables: Basophil, Eosinophil, Lymphocyte, Leukocyte, Neutrophil, Platelet, and Monocyte. Consultations were held with six expert physicians spanning diverse specialties relevant to COVID-19 diagnosis decision-making with the AHP method. Out of the 42 expert users (57.1% were male, with an average age of 37.30±10.56) were evaluated the system. The System Usability Scale (SUS) score averaged 81.43±15.64, indicating high usability. Conclusion: Consequently, this system might enable faster isolation of the patient and the commencement of preliminary treatment.

Anahtar Kelimeler

• Covid-19
• Machine Learning
• Imbalance Data
• Feature Selection
• Decision Support System
• Analytic Hierarchy Process (AHP) Method

Kaynakça

1. Dorn M, Grisci BI, Narloch PH, et al. Comparison of machine learning techniques to handle imbalanced COVID-19 CBC datasets. PeerJ Comput Sci. 2021;7:1-34.
2. Nicola M, Alsafi Z, Sohrabi C, et al. The socio-economic implications of the coronavirus pandemic (COVID-19): A review. Int J Surg. 2020;78(3):185-193.
3. Ge H, Wang X, Yuan X, et al. The epidemiology and clinical information about COVID-19. Eur J Clin Microbiol Infect Dis. 2020;39(6):1011-1019.
4. Bernheim A, Mei X, Huang M, et al. Chest CT Findings in Coronavirus Disease-19 (COVID-19): Relationship to Duration of Infection. Radiology. 2020;295(3):200463.
5. Hope MD, Raptis CA, Shah A, Hammer MM, Henry TS. A role for CT in COVID-19? What data really tell us so far. Lancet. 2020;395(10231):1189-1190.
6. Hadaya J, Schumm M, Livingston EH. Testing Individuals for Coronavirus Disease 2019 (COVID-19). JAMA. 2019;2020.
7. Vogels CBF, Brito AF, Wyllie AL, et al. Grubaugh ND. Analytical sensitivity and efficiency comparisons of SARS-CoV-2 RT–qPCR primer–probe sets. Nat Microbiol. 2020;5(10):1299-1305.
8. Zame WR, Bica I, Shen C, et al. M. Machine learning for clinical trials in the era of COVID-19. Stat Biopharm Res. 2020;12(4):506-517.

9. Ferrari D, Motta A, Strollo M, Banfi G, Locatelli M. Routine blood tests as a potential diagnostic tool for COVID-19. Clin Chem Lab Med. 2020;58(7):1095-1099.
10. Yan L, Zhang HT, Goncalves J, et al. An interpretable mortality prediction model for COVID-19 patients. Nat Mach Intell. 2020;2(5):283-288.
11. Alballa N, Al-Turaiki I. Machine learning approaches in COVID-19 diagnosis. mortality. and severity risk prediction: A review. Informatics in Medicine Unlocked. 2021;24:100564.
12. Zhou F, Yu T, Du R, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan. China: a retrospective cohort study. Lancet. 2020;395(10229):1054-1062.
13. Feltes BC, Vieira IA, Parraga-Alava J, et al. Feature selection reveal peripheral blood parameter’s changes between COVID-19 infections patients from Brazil and Ecuador. Infect Genet Evol. 2022;98:105228.
14. Saaty RW. The analytic hierarchy process-what it is and how it is used. Math Modelling. 1987;9(3-5):161-176.
15. Ho W. Integrated analytic hierarchy process and its applications - A literature review. Eur J Oper Res. 2008;186(1):211-228.
16. Brooke J. SUS-A quick and dirty usability scale. Usability Eval Ind. 1996;189-194.
17. Krawczyk B. Learning from imbalanced data: open challenges and future directions. Prog Artif Intell. 2016;5(4):221-232.
18. Bibicu D, Moraru L, Biswas A. Thyroid nodule recognition based on feature selection and pixel classification methods. J Digit Imaging. 2013;26(1):119-128.
19. Chen CH. A hybrid intelligent model of analyzing clinical breast cancer data using clustering techniques with feature selection. Appl Soft Comput J. 2014;20:4-14.
20. Kazemi Y, Mirroshandel SA. A novel method for predicting kidney stone type using ensemble learning. Artif Intell Med. 2018;84:117-126.
21. Chen CW, Tsai YH, Chang FR, Lin WC. Ensemble feature selection in medical datasets: Combining filter. wrapper. and embedded feature selection results. Expert Syst. 2020;37(5):1-10.
22. Adunlin G, Diaby V. Xiao H. Application of multicriteria decision analysis in health care: A systematic review and bibliometric analysis. Health Expect. 2015;18(6):1894-1905.
23. Suner A, Karakülah G, Dicle O, Sökmen S, Çelikoğlu CC. CorrecTreatment: A web-based decision support tool for rectal cancer treatment that uses the analytic hierarchy process and decision tree. Appl Clin Inform. 2015;6(1):56-74.
24. Suner A, Çelikoğlu CC, Dicle O. Sökmen S. Sequential decision tree using the analytic hierarchy process for decision support in rectal cancer. Artif Intell Med. 2012;56(1):59-68.
25. Silveira EC. Prediction of COVID-19 from hemogram results and age using machine learning. Front Health Inform. 2020;9(1):39.
26. Wyatt JC. Decision support systems. J R Soc Med. 2000;93(12):629-633.