Transdermal Patch Inspection
Confirm the appropriate amount of liquid medication on adhesive skin patches
Graphical programming environment for deep learning-based industrial image analysis
A variety of antibiotics, antiseptics, antifungals, and other medications can be administered by an adhesive patch worn on the skin. During manufacture, the active ingredient, in a liquid vehicle, is dropped onto the patches.
It is difficult to get the droplets to be consistent in size. FDA regulations mandate a narrow range of acceptable therapeutic dosages for each product. If the droplet is too large or too small, misplaced on the patch, or oddly shaped, the dosage will be incorrect, which could lead to compliance problems and potential recalls.
Conventional machine vision has difficulty analyzing the presence and exact size of these droplets against the background of the patch. The human eye is poor at volume judgments from observation, so manual inspection is error prone.
Cognex Deep Learning is a highly reliable technology for medical adhesive patch inspection. It trains on a set of images over a range of acceptable sizes and shapes of droplets on patches and a set of images of unacceptable droplets, based on measured medication dosage and desired placement. The classification tool then creates a category of acceptable droplet shapes and sizes and puts all other variations in the rejected category.