Open-CAS - Open datasets for validating and benchmarking computer assisted surgery

The validation of accuracy is a very important step towards a thorough evaluation for computer assisted surgery systems (CAS). It is especially important for algorithms, modules and systems that are in a research or prototype state. In order to provide navigational information to the surgeon CAS systems rely on suitable intraoperative sensors such as tracking systems, endoscopic video or ultrasound data. Typically, complex algorithms for sensor processing and registration techniques are used in order to align pre-operative planning data to the patient based on the sensor information. In this context, the first validations step is to assess the accuracy of each individual component before proceeding to validate the whole system.

Common data repository

Our goal for this website is to establish a repository of common data sets that allow all CAS researchers to spend more time on development and less time on validation. We will continuously update the website with suitable validation data from numerical simulations, phantom experiments and animal interventions. We hope that this website can contribute to a larger repository of such datasets: Either by building a link cloud between different repositories or by establishing a common one.

Benchmarking CAS components

In the future we envision a repository that can serve not only as a central access point for data, but also for benchmark results. In this context, our goal is to facilitate the development of standardized data formats and corresponding processing tools in order to make the data easily accessible.

Would you like to publish you validation data or do you have a website that contains such datasets? Then we are happy if you would like to contribute, either by sending us a link to your related dataset page or by uploading your dataset together with a reference to your paper and/or website. Please get in touch with us.