Software-as-a-medical-device is classified as 'software which is designed and developed to carry out a specific medical function'. So whether you are taking an input from a blood glucose meter and adjusting insulin dosages based on the input, or you are creating a dermatology assist tool, using an image of a skin condition to identify the melanoma risk. You can improve patient benefits through its function alone.
Many interested parties confuse software-as-a-medical-device (SaMD) with companion software. Companion software can provide data tracking, record keeping and risk reduction, but will fall short on being categorised as SaMD due to the lack of a direct medical function. Companion software is something that we can help deliver, but it is crucial that we correctly identify the differences.
The role of machine learning in software-as-a-medical-device is in processing, qualifying, analysing and correlating large data sets, which can be trained to support medical decision making across a range of disciplines from radiology to oncology, ophthalmology to cardiology. The value of SaMD development can, in many cases, be correlated with an increase in available datasets, with which to train these algorithms.
We're a bit guilty of hiding our light under a bushel. Find out from our partners, why collaborating with Waracle has delivered so much value for their organisation.