Andrew Johnson is CTO for AIMed with responsibility for database management, web development along with client management. A highly experienced publishing executive with a passion for technology.
Apart from walking into a pharmacy and asking a Pharmacist, current methods for identifying medications involve using prescription medication recognition portals such as PillBox, WebMD and others by chain pharmacies such as CVS, Walgreens and Rite Aid. They all use drop down menus where the user manually inputs the pill characteristics, none have the ability to utilize an image to identify pills. ‘Googling’ pill information can lead to inaccurate information as the information is not curated or verified by healthcare professionals.
The lack of dependable models that can safely and accurately identify prescription medications using images is inhibiting the ability to improve accuracy (through automated medication reconciliation), help fully automate the dispensing process using a ‘second’ pair of eyes ($5.3Billion market) or help patients with visual impairment (try using an app for the visually impaired to identify pills – the results are un-impressive). 40% of medication identification requests are from people ages 65-84 and there is a growing trend where technology is starting to play a role in helping people stay longer in their homes. In a recent study, it was found that 91% of respondents (aged between 50-80) said they wanted to live in their own home rather than alternatives such as assisted living communities, a trend that is known as aging in place.
We have been working on creating a pharmacist curated dataset and experimenting with advanced machine learning methods in order to identify pills using any device. This problem cannot be simply solved using regular techniques. There are challenges in identifying pills in different lighting environments, backgrounds, camera resolutions, camera angels. The goal is to curate a worldwide database of prescription pills that can easily be identified using images.
Pharmacist with a background in machine learning