Is there a simple and good way to validate the predicted box and class by yolov8?

Question

I'm predicting 4 classes with yolov8n in chest x-ray which are "ICD, Necklace, Shoulder Endoprosthesis, Vascular Port"; I'm getting bounding boxes from yolo in python and sometimes yolo predicts incorrectly. I want to somehow check if the predicted box and class is really correct or not, like it predicted a noise as an object; I thought that template matching would be good for "ICD" and "Vascular Port" since most of them are similar with each other in the photos, but not for "Necklace" and "Shoulder Endoprosthesis" cause there are lots of different shapes of them. Could you please tell me other solutions that they could help?