The COVID-19 pandemic is running its apocalyptic course around the world. One thing is certain in the midst of total uncertainty and utter chaos: SARS-CoV-2 is a very daunting enemy and we remain totally subjugated by these viral overlords. It is easy, therefore, for we humans to lose patience and become careless as well as to decrease capacity for resilience. We need to regain our composure and to be defiant in order to for us to prevail this human vs virus struggle.
A major part of this resilience and defiance is continue our work in adoption and deployment of artificial intelligence concepts and projects in clinical medicine. Here are key takeaways from today’s AIMed Clinician Series/Surgery webinar in collaboration with the International Pediatric Endosurgery Group.
My top takeaways:
1. The data gathered from all the disparate sources is the foundation for artificial intelligence work in decision support and medical imaging.
2. Data needs to have security to protect against ransomware and other breaches in security as it is a valuable commodity.
3. Hypotension prediction index to prevent hypotension during surgery can be a powerful tool to support the anesthesiologist decision making process.
4. Automation of fluid management can expedite appropriate fluid resuscitation in an effort to improve outcome.
5. Collaboration is key for surgeons to work with data scientists on projects to solve problems that traditional tools have not been able to solve.
6. Videos need a memory (long short-term memory or LSTM) along with convolutional neural network (CNN).
7. Collective surgical consciousness from recording and collecting surgical videos is a very useful tool in the future for any surgeon, especially with difficult or unusual cases.
8. Machine learning has a myriad of uses for surgeons including intraoperative imaging and perioperative risk assessment and management.
9. The steps for a data science project: 1. Frame the question; 2. Gather the data; 3. Curate the data and select features; 4. Divide the dataset into training and test sets; 5) Test the model; and 6) Iterate.
10. Data science can help neutralize human cognitive biases and heuristics that can distract the clinician from better decision making.
11. Important aspects of data science projects include: framing the question appropriately and collecting relevant data from all sources as well as deciding on the right model to deploy.
12. Natural language processing, under leveraged in its use in healthcare, can be used for assessment and communication tools such as the one for COVID-19 for children and parents.
13. NLP can also be effectively used for quality efforts with labelled data that is already available for other uses in the healthcare systems.
14. Advances in NLP is exponential: transformers bidirectional encoder representation from transformers (BERT) and generative pre-trained transformer 3 (GPT-3) are here; Big Bird is also.
15. Artificial intelligence can be embedded into surgery so that it can learn and be smarter in the future (20 years of experience can simply be the same one year experience x 20).
16. Intelligence information to guide surgery could be real-time or post-procedure so this aspect should be determined perhaps individually.
17. Artificial intelligence can allow the surgeons to have a better means of communicating with both team players and families.
18. It will be of utmost importance for surgeons to work with AI tools to achieve a human-machine synergy to improve outcome and patient experience.
19. The progress of AI in surgery may be a hybrid between slow and methodical as well as disruptive, and we can adapt according to pace of innovation in AI.
Thank you faculty and attendees for your knowledge and expertise as we all learned a great deal today at AIMed Surgery!
Anthony Chang, MD, MBA, MPH, MS
Chief Intelligence and Innovation Officer
Medical Director, The Sharon Disney Lund
Medical Intelligence and Innovation Institute (mi3)
Children’s Hospital of Orange County