Although websites such as Drugs.com and WebMD currently have pill identification tools — it’s the time required to enter detailed descriptions of drugs which can make such services too time-consuming for use in a clinical setting.
Jesus Caban and colleagues are working on a handheld image-analysis software that can distinguish the shape, color and imprint of most widely prescribed pills from a simple smartphone snapshot, according to a report from New Scientist.
Currently, the software has performed with a 91% accuracy in identifying 568 of the most prescribed pills from images taken in a variety of angles and lighting conditions. The biggest challenge is the app’s inability to identify pills captured from awkward angles — something that researchers expect will be improved upon going forward.
So far, this mobile pill identifier shows huge potential as a useful tool — both at home and for health care professionals.