A tourist has told of her “life-changing” visit to the Camera Obscura in Edinburgh after one of its thermal cameras detected she had breast cancer.
This raises the possibility of creating a very low cost breast screening programme using smartphones. Add-on infrared (thermal) cameras for smartphones already exist and it would be simple to add the sensor to the actual phone. This would reduce the additional cost to a few dollars or potentially a few cents.
At the moment women in the UK receive an NHS scan every 5 years. This is highly accurate but 2% of people who get cancer will get it in the month after their scan. The 5 year interval means that cancer could potentially remain undetected for some time. A monthly IR scan might save lives.
In the middle-income countries like China or Brazil good screening programmes may not exist for middle-class women and regular IR scans using smartphones might promote early detection and treatment. Success in middle-income countries could prompt a wider roll-out.
Analysis of millions of scan images is a natural application of artificial intelligence (AI). The huge data set that would arise from monthly scans is likely to drive up detection and reduce false positives.
Primary research needs to be conducted to see if this will work;
- Clinical Research
- longitudinal studies on patients with breast cancers and creating IR signatures
- Crowd-sourced Image Library
- collect thousands of thermal images of breasts by volunteers appropriately licenced for use in AI algorithms.
Whereas clinical research should follow medical ethics norms, it is possible that the crowd sourced images could be collected simply by contract.