Automatize skin prick test with a low cost Machine vision system

Abstract

Skin prick testing (SPT) is a reliable method to diagnose IgE-mediated allergic disease in patients with rhino conjunctivitis, asthma, urticaria, anaphylaxis, atopic eczema and suspected food and drug allergy. Due to the the lack of wheals public available dataset and related diameters and due to expen-sive acquisition system, in this work we investigate the feasibility to have high performance and low cost device for wheals detection and measurements. Having cheap acquisition device, it would be easier to have more acquisition stations per clinic, also speeding up the process of building the dataset. Proposed paper aims to automate the wheals measuring process, by detecting multiple wheals and by isolating them from the rest of the image, which is an essential step for the dataset building. We also proposed a cheap acquisition system made by a consumer type depth camera, Intel Realsense d435, a 3D printed arm support, a touchscreen panel displaying a GUI for the acquisitions, a Raspberry Pi 4 computer. Preliminary results showing promising performance even if they are acquired with a very cheap acquisition station proposed.