If there is a hierarchy of need for artificial intelligence in medicine, the lowest tier will probably be data. Medical information coming in the forms of written notes, test results, images and historical records etc. are all parked under which. These protected health information (PHI) captured by electronic health or medical records train machines to learn and make decisions beyond human’s capabilities.
Data, safety & ethics and diversity as the essentials
However, most PHI are not structured; there is no centralized system or universal method to keep a record of them all. Coupled with the lack of rules and regulation over who officially owns these information and insufficient trained specialists, the data which we get hold onto at the moment, may not be competent enoughor fully utilized to its maximum potential. Data’s importance and complexity, have thus, turned it to be the most crucial essential of AI in medicine.
If the vexed question of data can be overcome, safety and ethics would be the following tier. It’s tricky to access the safety of AI as all healthcare professionals undergo rigorous training. Some experts believe the real risks will only emerge when we give AI too much autonomy or we over-adhere them. As in the case of training surgeons with virtual reality, which some take it as undermining dexterity. Others believe popularity of AI will pose new Trolley problem, as patients are unsure if they should give the AI or human doctor, the rights to treat them.
Should essentials of artificial intelligence in medicine in hierarchical form
Diversity may be the third tier of this hierarchy. New technologies like artificial intelligence are neutral but human, who are in charge of designing and using AI, are not. If AI fails to cater to the many needs, cultures and systems of a spectrum of patients, employing it into medicine will only benefit those that the creators have in mind when a particular AI is being made. Likewise, the same medical condition has been found to express differently in male and female patients. If the developed algorithm is not able to sift out the differences, it defeat the purpose of having an innovation which believe to make healthcare more efficient and astute.
Let’s hear from four experts of technology, healthcare and academic backgrounds on their definitions of what is essential to AI medicine and if it should ever be ranked in a hierarchy.
Session Focus: Essential issues in AI Medicine
When: Friday, December 14th 2018 (14:00-15:00)
A session to explore the important subjects of AI in medicine; with four professionals sharing their respective insights from various fields.
Attendees will gain the following knowledge:
Discover the many facets of artificial intelligence in medicine and each of their merits and inferiority.
Be informed of the field’s diversity; how institutions from hospitals, research centers to private companies are tapping into the industry.
Review the different opportunities each essential has given to artificial intelligence in medicine; if the present practice can be altered to provide more benefits.
Benefit from an array of input provided by experts with a vast background.
John Mattison, Chief Health Information Officer, Kaiser Permanente, USA
Kevin Seals, Resident Physician, Diagnostic Radiology, UCLA Health, USA
Kathy Jenkins, Senior Associate in Cardiology, Department of Cardiology, Professor of Pediatrics, Harvard Medical School, USA
Gavin Bogle, Chief Business & Legal Officer, Systems Oncology, USA
Sina Bari, Solutions Architect, iMerit Technology Services, USA
You can sign up for this session here.