Our client is developing a new ophthalmic imaging device for at-home monitoring of retinal disease and needed a software architecture design. The design needed to consider how to transfer the volumetric DICOM OCT images captured from the device to the cloud for QA, processing and analysis. The client had licensed a deep learning model from an academic institution and needed to design the infrastructure to host and serve the inference results from the model. Finally, our client also required a path to manage the imaging process for patients and display results to providers.

A software architecture roadmap was developed to enable cloud connectivity. Application and data processing logic components were architected across firmware, gateway, mobile application, and cloud components. API layers and QA processes were defined. A companion mobile application was recommended to manage user engagement with the device and capture patient outcomes. Longer term, this dataset will enable personalized treatment recommendations and influence reimbursement decisions.