AI-based medical web app for OB/GYN doctors

About the project

Innitius is an innovative Spanish-based medical device startup. It was founded due to cooperation between the University of Granada and the Andalusian Health Service. Their goal is to create medical devices that enhance patient care.

Nowadays, Innitius aims to support women by developing a medical device that helps diagnose preterm labour using pioneering machine learning solutions.

What did our client need?

Innitius developed the Cervisense – a portable diagnostic device that helps OB/GYN doctors differentiate between false and true preterm labour threats. To achieve this goal, our client developed a custom algorithm to analyse data captured by Cervisense but encountered three challenges:

1. Designing an MDR-compliant web app which would allow the input and storage of the results received from the device.

2. Integrating the machine learning algorithm into the web app to analyse the data.

3. Requiring USB connectivity with Cervisense to enable automatic result readouts.

How did we approach the challenge?

Innitius needed a web app to complement Cervisense that would comply with Medical Device Regulation requirements while integrating their original machine learning algorithm.

We started our collaboration with Innitius by preparing the Product Design including UX (User Experience) and UI (User Interface) of the web application. We then moved on to the backend development and testing. We continuously consulted our progress with the client to guarantee a design that met their expectations. This way, we prepared an intuitive and effortless app for OB/GYN doctors.

Recognising the importance of risk analysis in medical devices and its influence on patients, we designed the app to provide information about the risk range generated by the algorithm's analysis.

The Cervisense is a device that could revolutionise the care of pregnant women. However, transcribing data from the display by the OB/GYN can be time-consuming and lead to some errors. When Innitius partnered with hospitals to develop new innovative solutions, we got involved in the task of transferring the Cervisense data to a web application via USB as a separate commissioned project.

Technology stack and integration

Our team developed the web app for Innitius with TypeScript, React, Python, and Django, deploying it to the cloud using AWS.

The project was conducted and tested under OWASP guidelines to strengthen the app's security. We've also integrated an industry-standard token-based authentication scheme extended with a mandatory second factor in the form of a time-based one-time password (TOTP MFA). It was supplemented with an MFA devices management system adapted to meet the day-to-day reality of medical personnel and IT management of hospitals.

One of our tasks was to integrate the machine learning algorithm, developed by Innitius for analysing data collected from patients, into the web application we were designing. We collaborated closely with the Innitius team throughout this process, ensuring the algorithm's integrity.

An interesting challenge involved connecting the app to a Cervisense device via USB, which we addressed by leveraging the Web Serial API to utilise the UART interface already available in the device. We carried out manual and automated tests to ensure smooth data transfer. This approach enabled seamless data transfer from the device to the application without troubling the OB/GYN doctors.

Wanting to make sure that the data transfer worked flawlessly, we worked with Cervisense hardware provided to us by the client.

Ensuring compliance with regulatory requirements

Innitius aims to have Cervisense and the web app recognised as a medical device under the Medical Device Regulation (2017/745). Thus, when designing the application, we drew on the expertise of our software developers and engaged our regulatory affairs specialists to ensure compliance with the MDR.

Aiming for the highest level of MDR compliance, we made all design stages align with ISO 13485, ISO 27001, and IEC 62304 standards. Additionally, to prioritise data security, we complied with both GDPR.

The client is planning to launch its product in the US market. That’s why we are now helping them prepare their cybersecurity documentation under the requirements of the FDA (Federal Drug Administration), providing comprehensive assistance throughout the procedure.

Project’s Team & Communication

Our team included three skilled medical software developers – Michał, Kamil, and Jarosław – supported by Wanda, our regulatory affairs specialist for MDR compliance. Our design expert, Natalia, crafted user-friendly interfaces while Michał ensured rigorous testing. Kuba, an experienced MDR project leader at Revolve, led the project, providing our client with a seamless, compliant solution.

We used Slack and Microsoft Teams to maintain smooth and continuous communication with Innitius. A client’s representative also joined the daily standups, actively engaging in the web app design process. As a result, we could adapt immediately to the most minor project changes.

Our solution

Collaborating with Innitius allowed us to contribute to an innovative solution supporting women's health. Over more than a year of partnership, we developed a medical web app that:

• integrates a machine learning algorithm,

• ensures a positive user experience,

• meets the requirements of the Medical Device Regulation, ISO 13485, and IEC 62304,

• incorporates security and data privacy measures that comply with OWASP and GDPR standards.

We continue working with Innitius to prepare cybersecurity documentation according to FDA requirements, offering support to enter the US market.