Intelligent assistance for monitoring surgical activity

The main objective of the APTRIC project is to design and develop an artificial intelligence-based solution capable of automatically analysing overhead surgical videos. The purpose of this tool is to extract structured clinical information from the images, demonstrating the technical feasibility of applying deep learning models in an environment as complex as the operating theatre. The aim is to automate key tasks such as classifying the type of surgery and assisted report generation, thereby optimising the medical workflow.

This initiative is the result of collaboration between the company Zerintia Healthtech and Vicomtech, which contributes its experience and track record in the application of data analysis and deep learning technologies, essential knowledge for training the algorithms needed to correctly interpret surgical activity.

The project has been developed using an iterative working methodology based on experimentation and continuous improvement. Starting with the selection and organisation of surgical images, the team has worked on training artificial intelligence algorithms to learn to visually interpret what happens during an operation, laying the foundations for the system's knowledge.

Rather than seeking a closed and immutable solution, the focus has been on exploring the real capacity of these algorithms to assist medical staff, moving towards the automatic generation of reports and verifying the reliability of the technology step by step. Thanks to this approach, APTRIC makes it possible to validate the enormous potential of these tools, opening the door to new ways of streamlining documentation and management in the operating theatres of the future.