A user friendly frontend for an AI cancer-detection model
Aidence is a med-tech start-up, providing AI-powered clinical applications for the oncology pathway. Their team consists of data science, medical and regulatory experts, supported by experienced advisors and critical users. Together they have built Veye Chest and Veye Reporting, their first AI clinical applications for early lung cancer diagnosis and reporting.
Creating a user-friendly frontend for Aidence’s application that can be easily used by radiologists in hospitals across multiple countries to check and label early-stage lung cancer.
What started as a three-month contract turned into long-lasting collaboration with Aidence.
Improving the application used by radiologists to train an AI model that detects early- stage cancer, by focusing both on the test process and on the frontend for the annotation tool.
Our colleague and Frontend Engineer, Mettin Parzinski, started working at Aidence in November 2020.
His initial task was extending their automated tests using React Testing-Library. Subsequently, because Aidence’s initial MVP was highly appreciated by the clients, the prototype went directly into production.
However, one of the challenges faced by Mettin and his team was that adjustments of the application, as requested by clients, were fairly difficult to manage. As such, Mettin’s current project focuses on rewriting the application, with the aim of making it flexible when adding new features and adaptable to multiple country regulations when writing CT reports. To achieve this, Mettin switched from using Vue to React, TypeScript and React testing libraries.