AI Imaging Medical Device Software
This project relates to the development of medical device software for the purposes of automatically classifying features found in CT scan images. The Client had developed proprietary algorithms utilising artificial intelligence algorithms, and had generated data to prove their efficacy. They then approached Boost Design to take their proprietary formulas and develop them into functional medical device software, as well as develop the design documentation necessary to be used in their upcoming medical device regulatory submissions.
- Working closely with the Client in order to establish an optimal plan for the project, in consideration of functional, regulatory, and commercial goals and constraints
- Researching the AI image classification system developed by the Client, and translating it into simple medical device software that can be verified and validated, whilst maintaining algorithm performance
- Developing a DIACOM software interface to extract new images from medical device databases as well as to push processed information back into the database, creating a system that is simple for Clinicians to use
- Working closely with the Client to consider usability requirements as part of the development process
- Developing a software architecture compliant with medical device requirements
- Developing software unit tests and user tests, and conducting verification testing of the software
- Developing risk assessments, and considering software design and features to appropriately mitigate risks
- Completing all development and documentation in compliance with IEC62304, ISO13485, IEC14971, and FDA
The project was completed with all features implemented, verified, and compliant documentation generated.