AUTHOR=Parveen Rahamathulla Mohamudha , Sam Emmanuel W. R. , Bindhu A. , Mustaq Ahmed Mohamed TITLE=YOLOv8's advancements in tuberculosis identification from chest images JOURNAL=Frontiers in Big Data VOLUME=Volume 7 - 2024 YEAR=2024 URL=https://www.frontiersin.org/journals/big-data/articles/10.3389/fdata.2024.1401981 DOI=10.3389/fdata.2024.1401981 ISSN=2624-909X ABSTRACT=Tuberculosis (TB) is chronic and pathogenic disease which lead to life threatening situations like death. Many people have affected by TB owing to inaccuracy or late diagnosis and deficiency of treatment. The early detection of TB is important to protect people from severity of the disease and threatening consequences. Traditionally, different manual methods have utilised for the TB prediction such as chest X-rays and CT scans. Nevertheless, these approaches are identified to be time-consuming and ineffective for acquiring optimal results. To resolve this problem, several researches have focused on TB prediction. Conversely, it results with lacks in accuracy, over fitting of data, and speed. For improving the TB prediction, the proposed research employs SFF (Selection Focal Fusion) block in YoloV8 (You Look Only Once) object detection model with Attention Mechanism through the Kaggle TBX-11k dataset. The YoloV8 is used for its ability to detect multiple objects at single pass. However, it struggles with small object and finds inability to perform fine grained classifications. To evade this problem, the proposed research incorporates SFF technique to improve the detection performance and decreases the small objects missed detection rates. Correspondingly, the efficacy of the projected mechanism is calculated utilising various performance metrics such as recall, precision, F1Score and mAP (Mean Average Precision) to estimate the performance of the proposed framework. Furthermore, the comparison of existing models discloses the efficiency of the proposed research. The present research is envisioned to contribute to the medical world and assists radiologists for identifying Tuberculosis using YoloV8 model to obtain optimal outcome.