AUTHOR=Li Chengye , Hou Lingxian , Pan Jingye , Chen Huiling , Cai Xueding , Liang Guoxi 

TITLE=Tuberculous pleural effusion prediction using ant colony optimizer with grade-based search assisted support vector machine

JOURNAL=Frontiers in Neuroinformatics

VOLUME=Volume 16 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/neuroinformatics/articles/10.3389/fninf.2022.1078685

DOI=10.3389/fninf.2022.1078685

ISSN=1662-5196

ABSTRACT=Despite the fact that tuberculous pleural effusion (TBPE) is only an inflammatory response of the pleura caused by tuberculosis infection, it leads to pleural adhesions and creates sequelae of pleural thickening, which may significantly impair thoracic mobility. Therefore, quick diagnosis and treatment of TBPE are essential; however, at now, the diagnosis of TBPE is still mostly reliant on certain traditional diagnostic procedures, so the early diagnosis of TBPE still has some limitations. In this study, we present bGACO-SVM, a model with good diagnostic capabilities, for the supplementary diagnosis of TBPE. This model is a feature selection wrapper based on an enhanced continuous ant colony optimization with grade-based search technique (GACO) and support vector machine (SVM). In GACO, grade-based search significantly improves the convergence performance of the algorithm and the capacity to avoid being caught in local optima, hence improving the classification capability of bGACO-SVM. In order to test the performance of GACO, this work performs comparison experiments between GACO and nine fundamental algorithms and nine state-of-the-art variant algorithms; the findings give compelling proof of the fundamental benefits of GACO. The accuracy of bGACO-predictions SVM's was evaluated using available datasets from UCI and the TBPE dataset. In the TBPE dataset trials, 147 TBPE patients were evaluated using the created bGACO-SVM model, demonstrating that the bGACO-SVM approach is an effective technique to predict TBPE accurately.