Comparative Analysis of Machine Learning Algorithms for Mycobacterium Tuberculosis Protein Sequences on the Basis of Physicochemical Parameters

Artificial Intelligence on Diagnostic Aid of Leprosy: A Systematic Literature Review

“…Studies by [ 14 , 87 , 94 ] classified leprosy as either paucibacillary or multibacillary. Binary classification occurred in 43% of the selected articles [ 12 , 14 , 76 , 82 , 83 , 87 , 94 , 95 , 96 ]. The remaining 57% of the articles used multiclass classification tasks, in which the models classified different skin diseases and, among them, leprosy [ 84 , 85 , 86 , 88 , 89 , 90 , 91 , 92 , 93 , 97 , 98 , 99 ].…”

“…The studies [ 14 , 76 , 87 , 94 , 96 ] utilized the outcomes of tests as input data to AI models, such as the RNA sequencing technique (RNA-Seq) and real-time reverse transcription polymerase chain reaction (RT-qPCR) used in molecular and cellular biology. The RNA-Seq technique extracts total RNA from a biological sample, converts it into complementary DNA (cDNA), and performs next-generation sequencing (NGS) [ 100 ].…”

“…The study by Tió-Coma et al [ 76 ] used the results of gene expression analyses from RNA-Seq and RT-qPCR as input to develop an AI model. The study by Pillai and Chouhan [ 96 ] analyzed the H37Rv strain to study the immunology and pathogenesis of tuberculosis. H37Rv is a strain of Mycobacterium tuberculosis [ 102 ] and share characteristics similar to M. leprae .…”

“…The review revealed that accuracy is the metric used by the authors in 90% of the studies. In addition, accuracy was the only metric to measure model performance in 52% of articles [ 14 , 82 , 85 , 89 , 90 , 93 , 95 , 96 , 97 , 98 , 99 ]. Accuracy can be misleading in multiclass classification tasks with imbalanced datasets, in which one class may have more samples than others.…”

“…Evaluating the quality of the selected studies allowed us to qualitatively analyze the methodological rigor of the ML studies in the medical area, their contributions, and the reproducibility of the results (see detailed quality assessment per study in Supplementary File S2 ). In 11 articles [ 83 , 86 , 90 , 92 , 93 , 94 , 95 , 96 , 97 , 98 , 99 ], the scores reached less than 40% of the tool score. Other studies attained results above 40% of the quality criteria [ 76 , 82 , 87 ], and the study in [ 12 ] achieved 63.3%, which was the best evaluation.…”