Wyniki wyszukiwania - BazTech

Ograniczanie wyników

2 Biocybernetics and Biomedical Engineering

2 Akben S. B.

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Sortuj według:

Ogranicz wyniki do:

Predicting the success of wart treatment methods using decision tree based fuzzy informative images

Akben S. B.

Biocybernetics and Biomedical Engineering

2018

Vol. 38, no. 4

819--827

Warts are small, rough, benign tumours caused by human papillomavirus (HPV). A challenge is predicting the success of wart treatment methods because success may vary depending on the patient and the features of disease. Recently, a machine learning based expert prediction system and related prediction rules were proposed. However, the success of this system is not satisfactory and should be improved. Furthermore, medical experts find it difficult to interpret the suggested rules of this system. The decision tree-based method was accordingly used in this study to determine the rules of predicting the success of wart treatment methods. According to findings, the success rate varied from 90 to 95% according to the treatment method; these rates are higher than previously reported. Furthermore, the decision tree rules that were determined can be transformed into images to visually interpret the success rates of treatment methods as a function of patient age and the time elapsed since disease appearance. This study provides a method for simple and more accurate interpretation of rules for medical experts. The success of treatment methods is now predictable as a percentage.

Early stage of chronic kidney disease by using statistical evaluation of the previous measurement results

Akben S. B.

Biocybernetics and Biomedical Engineering

2016

Vol. 36, no. 4

626--631

Chronic kidney disease (CKD) that causes the progressive losses in kidney functions is one of the major public health problems. Expert medical doctors can diagnose the CKD through symptoms, blood and urine tests in its early stages. However, the diagnosis of CKD might be difficult for expert medical doctors in case of the questionable measurement result. Therefore a new mathematical method that would be helpful to the expert medical doctors is required. It can be said that there is no studies related with automatic diagnosis of CKD in the literature. This study aims to remedy this shortcoming in the literature. In this study, for each of test and symptom values, averages of measurement results were calculated separately for healthy subjects and patients. Then the measured values were marked as ‘‘0’’ or ‘‘1’’ (healthy or patient) according to closeness to average values. Finally, the classification was performed by averaging the values marked for each subject. The success rate of the proposed method is between 96% and 98% according to the age ranges. In conclusion section of the study, how to implement the proposed method in real life is offered.