Surgical Data Science Collective (SDSC) brought their computer-vision-based analytics to life at the recent American Association of Neurological Surgeons (AANS) annual meeting with their first-ever live demonstration of surgical tool detection using machine learning models.
Using an anatomical replica of a head provided by UpSurgeOn, AANS attendees were able to perform a simulated pituitary tumor removal procedure. As the "surgery" unfolded, SDSC's AI models sprang into action, detecting and highlighting key surgical tools in real-time, including the curette, dissector, and rongeur.
From the live video feed streamed from a Storz endoscope , SDSC's innovative ML models generated vibrant heatmaps pinpointing the tools' positions and their proximity to the pituitary gland region. This cutting-edge capability represents just one facet of SDSC's comprehensive surgical video analytics suite, which is designed to provide surgeons with invaluable data for skills assessment and procedure optimization.
Typically, SDSC conducts such analyses post-operation through the Surgical Video Platform, generating insightful metrics like total distance covered by each tool and frequency of tool usage. However, the live demo at AANS showcased the potential for real-time guidance and feedback during intricate neurosurgical procedures.
To streamline the demonstration, the SDSC team had pre-prepared the UpSurgeOn box, expediting access to the simulated pituitary gland region. This enabled conference attendees to jump straight into the hands-on surgical experience while witnessing the AI's capabilities firsthand.
The live demo was a cornerstone of the AI Immersion section at AANS, underscoring the growing prominence of machine learning and computer vision in modern surgical practice.
Two of SDSC’s machine learning engineers, Jack Cook and Ayesha Syeda, were there to answer questions and collect feedback from surgeons. They were accompanied by Founder and President Dr. Daniel Donoho, and Head of Engineering Ahmed Amin, who also presented about SDSC’s transformative work.
In a parallel collaboration with Twelve Labs, SDSC also showcased its ability to automatically generate comprehensive reports from surgical video footage, further highlighting the company's dedication to enhancing surgical workflows through innovative technological solutions.
As surgical data continues to accumulate at an astonishing rate, pioneers like SDSC are poised to play a pivotal role in harnessing its potential, driving improved patient outcomes and advancing the frontiers of medical AI.
.png)
