Web application for cancer diagnosis using photonics and machine learning
Client
Canostix
Country
Germany
Industry
Diagnostics, Laboratory Software, SaMD
Collaboration
February 2022 - May 2022
Duration
4 months
Scope
Backend, Project Management, QA, Documentation and Risk Analysis
Technologies
Python, Docker, FastAPI
About the project
Canostix is a group of scientists working on a revolutionary method of cancer diagnosis using photonics and machine learning. The technique will enable doctors to detect various cancer types at an early stage, allowing patients to benefit from less invasive and more effective treatments.
The client invented a new way of blood testing using spectroscopy to differentiate between patients with and without cancer with excellent accuracy.
The end product - CancerScan - doesn’t need a specialist to interpret the outcomes, as the device produces a reliable result with high accuracy.
What did our client need?
Canostix came to us with very thought-through, innovative project ideas in the field of diagnostics / laboratory software. The client needed a healthcare-experienced software development partner to make their idea a ready product.
There was a need for an external company to redevelop the MatLab script to a ready-to-use solution. The whole process had to be compliant with ISO 13485, IEC 62304 and IVDR risk class C, that includes software documentation.
The work packages were executed in a very fast, professional, and flexible manner. They finished within a shorter time frame than initially assumed. Price, speed and quality.
How did we help to approach it?
We provided a service of Managed Project with a team of two developers, Michał and Przemek, and Marta, a Project Leader with a PhD and academic background in the field of BioTech.
Our work was focused on:
setting business goals
review of client’s materials - Matlab scripts and documentation
setting technical goals for project including translating Matlab scripts into Python
translation of Matlab scripts to Python console application according to technical and medical requirements
code review - a standard procedure of our ISO process
software documentation - quality assurance and risk assessment needed for regulatory reasons
Responding to our client's needs, we organised the development process in accordance with ISO/IEC 62304 within the client's Quality Management System (QMS). We created a simplified lean process tailored to the size of the project. It followed the requirements of our client and medical regulations/standards.
We coordinated the software development process in a Kanban project management framework. Together with the client, we chose Jira, Slack and emails as communication channels on which we were available daily.
We also regularly uploaded and reviewed MDR & ISO 13485 implementation documentation in the Confluence documentation system (by Atlassian).
Project’s scope in numbers
0%
10%
Meetings
0%
10%
ISO/MDR Documentation
0%
80%
Development
What have we delivered?
We developed a Python-based SiMD, including automated testing and documentation, that could receive data from multiple types of input devices and process raw data to provide a binary diagnostic result.
Fully automated and stable end-product
Solution applicable for diverse brands of spectrometers
Ready-to-use software documentation compliant with ISO & MDR requirements
Technologies used in the project
Feel free to ask about any technology of your interest that may not be mentioned below, as these are only the main ones focused on.
Are you interested in the development of software for your medical device?
We're happy to answer all your questions at the Discovery Call