Community Skin Health
Created a image classifying application to enable users to interact with a pre-trained model which gave them guidance on possible skin conditions. The application is designed to supplement existing government data sets to get valuable insights on community health trends. It also allows the community to engage more actively with the government using a medium they are already very familiar with.
Using IOT devices is a smart way to gather data from users to build up a accurate data set. Combining this with AI and other pre-existing data-sets allows for all sorts of possible discoveries regarding data correlations. For example combining our data with sets from the bureau of meteorology or ABS occupation data could lead to new insights regarding environmental factors and community skin health.
Evidence of Work
Description of Use: The device interacts with the user and collects data to be used and compared to other data sources.
Historical weather observations and statistics, Bureau of Meteorology
Description of Use: We used the UV statistics from these weather observations to find correlations between reported cases of skin conditions and the recorded weather observations at the same time.
International Skin Imaging Collaboration (“ISIC”) Public Image Dataset
ACT Govt - ACT Selected Cancer incidence and mortalityHealth Sep 2018
Description of Use: Compares reported skin condition cases to other environmental factors yearly over the last 20+ years.
Healthy Communities Challenge
Bounty: Making open data more open.
Region: Australian Capital TerritoryChallenge
Save Lives With Data