SURGICAL:AI AI Democratise Computer Aided Surgery WHY NOT GOOGLE - PowerPoint PPT Presentation

SURGICAL:AI AI Democratise Computer Aided Surgery

WHY NOT GOOGLE MAP AMAZON UBER FACEBOOK ALPHAGO FOR SURGERY

SURGERY CRISIS OA Affects 1 in 5 (US), Healthcare Cost $30 Trillions 1 in 6 (China) (US 23% GDP) OA cost 1%-2.5% GDP, Surgery Cost 40% ($12 Trillion) Second cause of disability 5 Billion have no Access to Surgery serves Only 5% Surgery patients (US), 0.5% (China)

SURGICAL AUTONOMY AI and Robotics 50% , worth $6 Trillion Narrow AI Highly Controlled Environment Human Support Cost Insensitive Ethical , Risk and Regulation

DO SURGEON REALLY NEEDS AUTONOMY ACRobot/Mako/Stryker ProBot Full Autonomous, 1997 Sedasys Cooperative, 2012 Da Vinci, No Autonomy, 2000 BrainLab, Navigation Robotic Drill

THE BARRIERS (T,C,S,A) Time Time Capabilit Safety Affordabi (prepare) (OR) y lity Manual 10 minutes 2 hour 100 50% +0 experts success Robot +1 Day 4 hour 1,000 90% +$500,00- experts success $2,00,000 Navigation +1 Day 3 hour 300 75% +$100,000 experts success -300,00 PSI +1 Month 1 hour 1,000 90% +$1,000 experts success SurgicalAI 10 minutes 1 hour 1000 95% +$500 SURGICAL:AI experts success

AUTONOMOUS PSI DESIGN BY AI Web base Mixture Deep Morphology Engine Planning And Design Engine Reality Interface TensorLab: Big Data Warehouse And Autonomous Machine Learning System Design in Less in 10 seconds, Deliver in 4 hours

SHOULDER PSI EXAMPLE (CAOS 2016) (FULLY AUTONOMOUS)

Neural Poincare Engine Deep Morphology Engine = + + Deep Poincare Map (https://arxiv.org/pdf/1703.09200.pdf) Speed : 0.2 Seconds for Segmentation one bone from a volume Data Efficiently : Traditional Deep Learning trained 5 volume we achieve accuracy close to human Memory Efficient : for Medical Images processing Wishing A baby will becoming a great Needs about 500MB memory, It can run on your mobile phone doctor by showing him lots of scans Flexible: without going to school Parts scan or arbitrary posture. Error Estimation: 1 Data Scarcity: We know when it fails. 30 Volumes vs 2 Millions 2 Memory Consumption It works like Surgeon, 3D volume vs 1 D signal It has memory, knows where to attend its focus 3 No Explanation and No Confidence 4 Flexible Workflow….. 5 Not Cooperative

SURGICAL DB Medical Image and Processed Images PSI Models Implants The Surgery Plan NoSQL and Map Reduce Search Engine Solution Find me all the pre and after surgery CT scans of all the patients using Company A implants on the third Spine Joint The only Question Surgeons Care but No existing PACS or EHR system can Answer

TensorLab for Autonomous Continuous Learning The Principal: 1) Everything is Data. 2) Everything is Identified by Query. 3) Pulling based Stream Processing Build Job Data: Acc Heart with Tumour Heart IoU Arch: Failure Best Segmenation Model Time Left Input Version: Most Accurate Version Volume Model Versions and Model Architecture Performance Metrics Output Version: Images and Sematic Labels Heart Tumour Detection Search Engine is used Recruitment by Search Engine as the Router for Data Analyser Validator Validator Builder Builder Analyser Analyser Builder Job Queue Job Query as Load Balancer Job1 Job6 Job 2 Job5 Job3 Job4

CLINICAL RESULTS: Save 40 minutes in OR,1000GBP for 1 procedure

PATIENT SPECIFIC OSTEOTOMY Guide was designed, printed and Guide was fixed to the unique Bone was resected following the sterilized position on the bone guidance of the guide Wedge was closed and bone was 15° wedge was resected and fixed removed

OTHER PROCEDURES

APPLICATIONS 1) Surgery Planning 2) Patient Specific Guide Device 3) Surgical Workflow Management 4) Implants Registration and Management

A TEAM OF WORLD LEADING EXPERTS Professor Anan Shetty Dr Fangde Liu (CEO &CRO) (Chief Medical Scientist) Chief Architecture of Several The best Surgeon in the UK Flagship Surgical Robots, Shetty-Kim Technique Inventor Pharmacovigilance for multination The Surgeon of 2017 Autonomous CMR Analysis System by Royal College of Surgeon and Diagnosis Proudest Award for UK, Senior Associate of the Hunterian Professor Royal Society of Medicine Advisor of NICE ( National Institute of Clinical Excellence) Dr Hao Dong (Engineering VP) Yinna Bu PhD at Imperial College, (Client and BD Manager in China) Data Science Institute MSc from UCL Deep Learning Engineering Expert CEO of two start-ups The Creator of TensorLayer Most Popular DL Framework in China

Mr Saif Ahmed Consultant Surgeon Prof Ferdinando Rodriguez y Baena (Advisor) Professor of Anatomy Expert on 3D Printing and Full Professor of Medical Robots at IC, Surgical Robots World First Orthopaedics Surgical Robots Executive Comminute of Sold For Stryker 1.67 Billion Dollars Computer Aided Team Leader of Biggest Surgical Robots Project Orthopaedics' UK In European Union Chair of World Computer Aided Orthopaedics Dr Jianmo Li (Advisor) PhD Prof Andrew Amis Postdoc at Imperial College Inventor of Knee PSI Prof Yike Guo (Advisor) Full Professor of Data Science at Imperial College Chairman of TranSmart Foundation Dr Joshua Giles (Advisor) World Standard Platform for PostDoc at Imperial College Translation and Clinical Medicine Inventor of Should PSI Adopted by all the major Pham Company in the world Including J&J, Merkek, Roche, Sanophie.

Kent Institute of Medicine & Surgery The hospital is equipped with state- • • KIMS is the largest Independent of-the-art technology and innovative hospital in SouthEast England and diagnostic services Kent. • Wide range of treatment – covering No.1 patient satisfaction (92%) • all the orthopedic area and joint against other 92 independent hospital replacement in the UK

AI Democratise Computer Aided Surgery Autonomous Surgery is Feasible The barrier is Human Machine Cooperation For Surgery, Not Just AI, Vehicle Matters Provides the tool, not solution GPU plays crucial by breaking the time barrier

IDEALondon, 2016 CAOS 2016, Japan Osaka Daily Mail, 2016 GTC 2017 Re.Work Deep Learning in Healthcare Summit, 2017

Surgical AI: Bring Big Data and AI into Operating Theatre Thank You