Towards the Scalable Cloud Platform for Non-Invasive Skin Cancer Diagnostics

According to the EU statistics¹ skin cancer - melanoma is the one of most deadly cancers. At the same time, it has up to 95% chance² for curing if detected on the early stage. The early diagnostics is obstructed by low availability of dermatology specialists and low percentage of patients that are taking regular diagnostics. To deal with the early diagnostics unavailability we propose implementing portable automated diagnostic device, available to wide range of medical institutions. Automated diagnostic systems are not yet capable for providing final diagnosis, but are enough for detecting threats of skin cancer. Therefore, by bringing automatic tools to primary care physicians and other medical personal a lot persons could save their lives by detecting cancer on the early stage. To provide automated diagnostic tool, it is needed to solve multiple tasks including: skin images acquisition, image processing, calculating disease probability and delivering diagnosis result. These tasks have existing solutions as a commercial tool or as scientific methods, but there is a lack of ready to use service. Each paper author has specialization in one the fields, that will be used to implement proposed system in ongoing EU project.