Biomedical Applications of Cloud Computing’, ( skin cancer detection using smart phone)
Keywords: Embedded Vision, Low Energy Processing, Machine Learning
particularly if your research proposal could focus on how we could use the cloud to upload and rapidly analyse images obtained from a skin cancer clinic
Smartphone apps that diagnose skin cancers have been available to the public for the past 5 years . There are however some concerns about the reliability and public safety of these devices. Medical researches have recently reported high error rates of up to 30% when compared to the diagnosis of an expert dermatologist. A pressing task is to understand the limitations of these systems and propose improved diagnostic models. This project will seek to identify applications, devices and software most relevant for clinical use particularly for the most difficult to diagnose cases.
Skin cancer can often be detected by monitoring for changes in moles over time. One of the limitations of smart phone devices is their limited storage and computational abilities. If you could use the power of cloud computing you could monitor the skin cancer image across time (lots of storage on the cloud for images) and you could also use much more advanced image processing algorithms (lots more processing power on the cloud). You could then use the cloud to monitor for changes in the each individual mole by comparing the images stored of it on the cloud across time.