Research on Multidrug Resistant Tuberculosis Discussed by Researchers at Shahid Beheshti University of Medical Sciences (Prediction of multidrug-resistant tuberculosis in tuberculosis patients using perceptron artificial neural networks model).

Researchers at Shahid Beheshti University of Medical Sciences in Tehran, Iran have conducted a study on multidrug-resistant tuberculosis (MDR-TB) and developed a predictive model using perceptron artificial neural networks (ANNs). The study analyzed data from 1,050 TB patients treated at Masih Daneshvari Hospital from 2005 to 2015. The model achieved a predictive accuracy of 85% for the training group and 80% for the testing group. Factors such as marital status, education, drug use, comorbidities, and age were identified as influential in predicting MDR-TB. The researchers believe that the application of this model can improve the diagnosis of MDR-TB due to its high accuracy. [Extracted from the article]