Is the use of Adam optimiser and label smoothing adequate for optimizing YOLOv7 and YOLOv7-E6E to attain a high-quality automated analysis and differential diagnostic evaluation of bronchoalveolar lavage fluid?

Abstract

Background: In a world where lower respiratory tract infections rank among the leading causes of death and disability-adjusted life years (DALYs), precise and timely diagnosis is crucial. Bronchoalveolar lavage (BAL) fluid analysis is a pivotal diagnostic tool in pneumology and intensive care medicine, but its effectiveness relies on individual expertise. Our research focuses on the "You Only Look Once" (YOLO) algorithm, aiming to improve the precision and efficiency of BAL cell detection. Methods: We assess various YOLOv7 iterations, including YOLOv7, YOLOv7 with Adam and label smoothing, YOLOv7-E6E, and YOLOv7-E6E with Adam and label smoothing focusing on the detection of four key cell types of diagnostic importance in BAL fluid: macrophages, lymphocytes, neutrophils, and eosinophils. This study utilized cytospin preparations of BAL fluid, employing May-Grunwald-Giemsa staining, and analyzed a dataset comprising 2,032 images with 42,221 annotations. Classification performance was evaluated using recall, precision, F1 score, mAP@.5 and mAP@.5;.95 along with a confusion matrix. Results: The comparison of four algorithmic approaches revealed minor distinctions in mean results, falling short of statistical significance (p < 0.01; p < 0.05). YOLOv7, with an inference time of 13.5 ms for 640 x 640 px images, achieved commendable performance across all cell types, boasting an average F1 metric of 0.922, precision of 0.916, recall of 0.928, and mAP@.5 of 0.966. Remarkably, all cell classifications exhibited consistent outcomes, with no significant disparities among classes. Notably, YOLOv7 demonstrated marginally superior class value dispersion when compared to YOLOv7-adam-label-smoothing, YOLOv7-E6E, and YOLOv7-adam-label-smoothing, albeit without statistical significance. Conclusion: Consequently, there is limited justification for deploying the more computationally intensive YOLOv7-E6E and YOLOv7-E6E-adam-label-smoothing models. This investigation indicates that the default YOLOv7 variant is the preferred choice for differential cytology due to its accessibility, lower computational demands, and overall more consistent results than comparative studies.