To champion the creation of sustainable, robust, and equitable digital healthcare systems that prevent the perpetuation of healthcare inequalities, ARC researchers took the lead in organising the second workshop on Open-Source Software for Surgical Technologies at the Hamlyn Symposium on Medical Robotics on June 28th, 2024.

The workshop focused on a key question: how can we transform open-source software libraries into sustainable, long-term supported tools that are translatable to clinical practice? To address this, the event brought together engineers, researchers, and clinicians from academia and industry to present their work, discuss current progress, challenges, and trends, and lay the foundation for building a collaborative community around Open-Source Software Innovations in Surgical, Medical and AI Technologies.

In this post, we are excited to share recordings of our exceptional lineup of speakers and celebrate the poster awardees from the workshop, along with Zenodo links to other posters. The talks and posters spanned a variety of topics, including certification, commercialisation, and case studies of open-source software in research and industry scenarios. This workshop highlighted the profound impact of open-source software in advancing surgical technologies and medical innovation.

Speakers

• Introduction to the workshop by Miguel Xochicale (UCL) and Eva C. Herbst (ETH) [Talk]
• Xenophon Papademetris (Yale) discussed Medical Software, Regulation, and Related Challenges [Talk]
• Kevin Cleary (Children’s National) provided a Historical Perspective on Open-Source Software for Medical Robotics and Imaging, highlighting lessons learned for managing open-source projects [Talk]
• Sonia Pujol (Harvard Medical School)  and Colton Barr (Queen’s University) presented on Open-Source Neuronavigation for Low-Resource Healthcare Settings [Sonia’s talk] [Colton’s talk]
• Andres Diaz-Pinto (NVIDIA) introduced MONAI Label, an open-source tool for image labelling and learning [Talk]
• Mikael Brudfors (NVIDIA) showcased Accelerating AI Medical Device Streaming Workflows with NVIDIA Holoscan [Talk]
• Shashank Sharma (KARL STORZ VentureONE) spoke on Enabling Software-Defined Medical Robotics Using Open-Source Software [Talk]
• Andrew Wichmann (Johns Hopkins Technology Ventures) explored Commercial Open-Source Software [Talk]
• Micha Pfeiffer (NCT Dresden)  delved into Software-Driven Computer-Assisted Navigation in the Age of Deep Learning [Talk]
• Tamás Haidegger (Óbuda University) discussed Open-Source Software Supporting Product Development: Sustainability versus Certification [Talk]

Watch all the recorded talks on this YouTube Playlist.

️Poster awardees
Congratulations to all the awardees for their outstanding contributions to advancing innovation in surgical technologies!

Best Poster Award
Martin Huber et al. from King’s College London “LBR-Stack: ROS 2 and Python Integration of KUKA FRI for Med and IIWA Robots”
GitHub: https://github.com/lbr-stack/lbr_fri_ros2_stack/
arXiv: https://arxiv.org/abs/2311.12709

Runner-Up Awards (Three-Way Tie)

• Keisuke Ueda et al. from Medical DATAWAY “Automated Surgical Report Generation Using In-context Learning with Scene Labels from Surgical Videos” Poster in Zenodo https://zenodo.org/records/12518729
• Mikel De Iturrate Reyzabal et al. from King’s College London “PyModalSurgical. An image-space modal analysis library for surgical videos: generating haptic and visual feedback” Poster in Zenodo: https://zenodo.org/records/12204075​
• Ewald Ury et al. from KU LEUVEN “Markerless Augmented Reality Guidance System for Maxillofacial Surgery”

See other posters Peter Kazanzides et al., dVRK-Si: The Next Generation da Vinci Research Kit, Reza Haqshenas et al., OptimUS: an open-source fast full-wave solver for calculating acoustic wave propagation with applications in biomedical ultrasound.

Get in touch

• Join our Discord server: https://discord.gg/P6wB44Ftft ​
• Star our GitHub repo: https://github.com/oss-for-surg-med-ai-tech/workshop-hamlyn2024
• Follow our GitHub organisation https://github.com/oss-for-surg-med-ai-tech/
• If you are still interested to contribute as co-author of a coming white-paper on “Open-Source Software for Surgical, Medical and AI Technologies: Challenges and opportunities” please send an email to m.xochicale@ucl.ac.uk or raise a new issue.
• If you are interested in co-organising or participating in next year’s workshop, we would love to hear from you!

We can’t wait to see you again next year!
Warm regards, Eva, Stephen, & Miguel

 

 

 

#HSMR24 #HamlynSymposium2024 #Healthcare #OpenSource #ArtificialIntelligence #SurgTech #MedTech #AITech

What’s in a name?

Electronic Healthcare Records are a vital source of data for a wide range of research, ranging from the implementation of new medical devices to the investigation of cures for disease. However, this data cannot just be made available to anyone, a written request must be vetted and approved, appropriate data must be extracted and anonymised and finally the anonymised data must be deposited in a secure location i.e. the whole process and the data itself must be SAFE. 

What is SAFEHR?

SAFEHR (pronounced simply as safer) is a cross institutional collaboration between University College London Hospital and ARC whose primary goal is to support healthcare researchers at UCL and the hospital.

Data for research

We facilitate researchers requiring data by providing a single point of contact for their requests for multimodal data from the hospital. In order to achieve this we have developed 

• An online form which captures all the necessary information for an application for data
• OMOP-ES – a tool that given a cohort extracts the data in the The OMOP Common Data Model
• PIXL – a pipeline for extracting images from the hospital systems 
• A pipeline to anonymize free text notes using Cogstack
• Code that links all these together and delivers them to the Data Safe Haven
• • • •  

Data for education

We have also been instrumental in establishing a process and set of guidelines for the use of this data in Education. Students doing courses that may require healthcare data will in future be able to use data that is realistic.

Data for anyone

Working with a project developed by the Turing Institute we are working on a pipeline that will produce a synthetic set of any data we extract that can be made publicly available for anybody to use.

Data we can provide

Records	Images	Reports
Demographics	X-rays	Radiology
Admissions	MRI	MS Clinicians notes
Measurements	CT (coming soon)
much more…

Have you noticed it’s all about the data!!

The SAFEHR Launch

SAFEHR was officially launched on December 4th 2024 to an audience of about 50 people including some prestigious names from both UCL and UCLH. We presented all aspects of our pipeline from making an inquiry, browsing our data catalogue, getting approval to actually receiving data. The slides from the presentation are available and you can find more information on our website

 

Acknowledgements

We acknowledge funding support from UCL, the UCLH Biomedical Research Centre, the Institute of Neurology, Dementia Research Institute and projects led by Prof. Gary Royle and Dr Arman Eshaghi.             

Research by Nick Youngson CC BY-SA 3.0 Pix4free

Team (u)CLI finish 3rd in the national Computing Insights UK student Cluster Challenge!

Team members from left to right:

• Zak Morgan, Dept of Computer Science
• Rozenn Raffaut, Dept of Med Phys & Biomedical Eng
• Tom Bickley, Dept of Chemistry
• Yuliang Huang, Dept of Med Phys & Biomedical Eng
• Qi Li,  Dept of Med Phys & Biomedical Eng

 

 

 

 

The CIUK Cluster Challenge is a national inter-collegiate competition. This year there were 16 student teams taking part from across the UK.

The competition was formed of three challenges during October and November, before culminating in an additional three challenges at Computing Insights UK (CIUK) in Manchester in December.

The team performed well in the initial pre-conference challenges, and were at the top of the leaderboard heading into the onsite challenges.

They were supported throughout by the team mascot, Archie ‘ARC’vark.

Competition was fierce and overall team (u)CLI finished 3rd out of 15 teams who completed all challenges. Narrowly missing out on 2nd place by 2 points. Full results are available on the CIUK Cluster Challenge website.

Overall the team performed very well and received some very positive feedback from some of the challenge setters.

We look forward to supporting future teams in new challenges, and building a community of students who want to participate through our termly Cluster Club sessions.