The Revolution of Data in Neurosurgery · Associate Professor of Neurological Surgery, Vanderbilt...
Transcript of The Revolution of Data in Neurosurgery · Associate Professor of Neurological Surgery, Vanderbilt...
The Revolution of Data in Neurosurgery
Lola B. Chambless, MD, FAANS
Associate Professor of Neurological Surgery, Vanderbilt University
Residency Program Director, Vanderbilt University Dept of Neurological Surgery
Chair, Committee on Data Science, Congress of Neurological Surgeons
NeuroSafe 2019
Disclosures
• Stryker (consultant)
• Digital Reasoning (ownership interest)
“The sexiest job of the 21st
century” !!!!!
Data Science in healthcare/neurosurgery•Present state•Near horizon
How will AI augment physician performance inthe future?
“Artificial Intelligence is the field of computer science dedicated
to solving cognitive problems commonly associated with
human intelligence, such as learning, problem solving,
and pattern recognition.” Amazon.com
MachineLearning Deep
Learning
Others
Strengths of AI
• Use of data streams beyond text• Images, sounds, locations
• Computational Power• Process vast quantities of information while testing numerous algorithms
• Learning and adaptation
Computer vision
• Uses deep learning to allow machine to detect “features”
• Visual pattern software estimated to be 5-10% more accurate than human eye… and increasing
What about neurologic disease?
Pitfalls
• While computer vision can be highly predictive, it can have a basis in flawed associations.
• Major “misses” are more likely when the possibilities are non-binary.
Pitfalls
• Algorithms can amplify healthcare disparities.
• Issues of ethics and privacy may be underestimated.
• Merging data from multiple sources can be vexing.• Research data is likely of higher quality then typical clinical data
How can AI augment neurosurgeon performance?
• Natural language processing
• Robotic assisted procedures
• Point of care predictive models
Is the future of AI-integrated medical practice personalized?
Robotics
• Placement of instrumentation enhanced by use of patient-specific anatomic features
• Reduction of complication rates compared to free hand techniques
• $40 billion value (savings) by 2026
Bringing AI to the end user (the clinician)
• https://los.insds.org/los • https://tlif-disp.insds.org/app#
• https://tlif-disp.insds.org/app#
•AI enhanced reports of digital images
•Robotic no-fly zones to minimize human error
•Point of care models to predict key outcomes which are personalized to the physician
What’s new with the CNS Data Science Committee?
Thank [email protected]