Artificial Intelligence in Healthcare - The Need for Ethics | Varoon Mathur | TEDxUBC

The advent of artificial intelligence (AI) promises to revolutionize the way we think about medicine and healthcare, but whom do we hold accountable when diagnoses fail or when automated procedures go awry? Drawing on experiences in global health and machine learning can help highlight the need for a sound ethical framework to be put in place to deal with emerging challenges and threats posed by AI, so that we can learn to build technology that puts patients and their health first. Varoon is a senior Computer Science student and a global health activist. Focused intently on the lack of affordable medicines for the most vulnerable and marginalized, Varoon serves on the national leadership team of the student-led NGO Universities Allied for Essential Medicines (UAEM). He has led delegations as part of the World Health Organization and led the publication of the first Global Health report card for U15 Canadian Universities to evaluate their impact on global equity in biomedical research. This talk was given at a TEDx event using the TED conference format but independently organized by a local community. Learn more at www.ted.com/tedx

[@yourhealthmatterswithAminat]

Thank you Varoon for this intellectual piece. I am a nurse and I am looking at Application of AI in nursing for my PhD study. This is thought provoking .

[@varoon3]

Thank you so much!

[@BlueBunny-qy8re]

Varoon, I admire your efforts to raise awareness for the equal access and cost of healthcare for people of all economic and social classes. I commend you for pursuing something that you are passionate about, even if it may not have been the most traditional career path. AI in healthcare is an interesting concept and has much potential in helping patients. While I believe it can positively impact patients and their health, I am still wary to trust AI over a physician’s knowledge and years of experience. After you explained that AI algorithms amplify implicit biases, my mind immediately jumped to a concept that was taught in my ethics course in medical school: justice. For those who aren’t aware, justice can be defined as the fair and equitable distribution of health resources for all, independent from gender, ethnicity, place of birth or residence, social status, political beliefs, and birthplace, according to the American Medical Association. It’s outrageous that 10 million people every year die from diseases that are 100% treatable, but if AI is going to worsen this gap in healthcare, do we really want to continue trying to force it into medicine? I completely support your belief that radiologists, as well as physicians of other specialties, won’t be replaced by an AI algorithm and I know that I personally, as well as many of my family members and friends, value transparency and accountability in healthcare. I don’t feel that I could completely trust an AI algorithm in the decision-making process for my health, nor would I feel comfortable trusting an algorithm that displays bias towards certain populations, such as African Americans, the elderly, pregnant patients, and children. Until we can prove that AI algorithms ensure justice and accuracy in patient care, I don’t feel that anyone should trust AI with such important tasks. Moving forward, as you have already stated, a task force for ethical AI in medicine would absolutely be a step in the right direction in putting every patient’s health first. Members for this task force would need to be chosen very strategically to ensure that the AI algorithm was ethical, free of any implicit bias, and could guarantee justice for all patients. Maybe you could be the one to head up this task force?!

[@sushmaramswami5722]

A very intellectual and informative talk..........it really helped me understand what is AI.

[@amymjmurphy]

Wonderful piece -- thank you for sharing your thoughts on innovation and particularly the imperative for addressing the failures within our educational structure.

[@MedschoolMom]

Bravo Mr. Mathur for taking on this important and complicated issue. I think there is huge potential for the use of AI to improve healthcare. The right algorithms can help diagnose symptoms and identify risk factors better than any human. AI could do so much good. But there are also huge risks. As Mr. Mathur illustrated, if the algorithms aren’t programmed correctly, they could just perpetuate and amplify the biases we’re already struggling to overcome. If we don’t do it right, AI will just make things worse. Can we justify using AI for the convenience with the possibility of knowing it could cause harm? Non maleficence is an important part of healthcare ethics, and we should think real hard before letting AI takeover. I also think AI should never fully replace human beings from the healthcare equation. AI should be a tool that helps physicians make decisions, but we always need a thinking and compassionate person who can objectively consider whether the AI’s conclusions make sense given the situation. And of course we also want to make sure patients are still getting kind and considerate care.

[@user-sd2kq6ol6s]

Varoon, thank you so much for your time and expertise within this subject matter. This what very informative and insightful for the future of medicine. I really appreciated your comment regarding trust with your Physician and trust with AI or an algorithm. It comes to medical ethics. I think that this is a conversation that we need to have. One of the principles of medical ethics is nonmaleficence, which essentially means, do no harm. The AI or will solely look at what is best for the patient based off of the algorithm, and what it can go off of may not what is best for the patient, nor what the patient is saying, nor respecting the autonomy of the patient but purely makes its decision based off of the algorithm that it is set with. This could further do harm to the patients trust or even not give the best options that the patient wishes. The lack of autonomy and increase of implicit bias that could be expressed with the AI would cause more harm than good to the patients and their relationship with the medical field/staff. In terms of relationship, this would also affect the loved ones of the family. How would the AI be able to empathize, express compassion, or console a loved one? Or if there is there is an issue with the power of attorney or who makes the final decision, how would an algorithm be able to resolve the problem while being able to maintain the focus on the patient? There are tons of questions that need to be answered and I believe that the answer lies in the Physicians and future Physicians that are handling these patients right now. The rapport, teamwork, and collaboration that humans have to collective move forward for the well-being of the patient when we get in the “grey zone” with ethics is unparalleled. I will say that I have less hope than Varoon about the potential for AI in medical ethics. Transparency and accountability are two core values when it comes to relationships that we have. Think about your parents, spouse, or even Primary Care Physician: you expect them to be transparent and hold themselves accountable for your well-being and your vulnerability in some way, shape, or form. If AI can not uphold that standard, it will be difficult to have them at the forefront of medical decisions. You can also put this in the perspective of someone who comes in and is seeking drugs, how would AI be able to tell if someone is able to mess with the algorithm to get what they want? Lack of accountability if no excuse in the medical field, there is usually a very fine line with treating a patient. With that, I don’t think AI or algorithms are something that can be implemented in medicine at the moment.

[@hershdhillon]

Very nice talk Varoon. Great to have contemporaries like yourself in the field. TC

[@garrettraiden1229]

you prolly dont care but does any of you know of a trick to log back into an Instagram account..? I stupidly lost the account password. I love any tips you can offer me

[@callanrafael1658]

@Garrett Raiden Instablaster :)

[@garrettraiden1229]

@Callan Rafael i really appreciate your reply. I got to the site on google and Im in the hacking process atm. Seems to take a while so I will reply here later with my results.

[@garrettraiden1229]

@Callan Rafael It did the trick and I actually got access to my account again. I am so happy:D Thank you so much, you saved my account !

[@callanrafael1658]

@Garrett Raiden glad I could help xD

[@joncarrasco962]

SEMA4'S Centrallis AI platform will change personalized medical care. As well as receive a huge amount of data for AI advancements and market share gains.

[@BlueBunny-qy8re]

Varoon, I admire your efforts to raise awareness for the equal access and cost of healthcare for people of all economic and social classes. I commend you for pursuing something that you are passionate about, even if it may not have been the most traditional career path. AI in healthcare is an interesting concept and has much potential in helping patients. While I believe it can positively impact patients and their health, I am still wary to trust AI over a physician’s knowledge and years of experience. After you explained that AI algorithms amplify implicit biases, my mind immediately jumped to a concept that was taught in my ethics course in medical school: justice. For those who aren’t aware, justice can be defined as the fair and equitable distribution of health resources for all, independent from gender, ethnicity, place of birth or residence, social status, political beliefs, and birthplace, according to the American Medical Association. It’s outrageous that 10 million people every year die from diseases that are 100% treatable, but if AI is going to worsen this gap in healthcare, do we really want to continue trying to force it into medicine? I completely support your belief that radiologists, as well as physicians of other specialties, won’t be replaced by an AI algorithm and I know that I personally, as well as many of my family members and friends, value transparency and accountability in healthcare. I don’t feel that I could completely trust an AI algorithm in the decision-making process for my health, nor would I feel comfortable trusting an algorithm that displays bias towards certain populations, such as African Americans, the elderly, pregnant patients, and children. Until we can prove that AI algorithms ensure justice and accuracy in patient care, I don’t feel that anyone should trust AI with such important tasks. Moving forward, as you have already stated, a task force for ethical AI in medicine would absolutely be a step in the right direction in putting every patient’s health first. Members for this task force would need to be chosen very strategically to ensure that the AI algorithm was ethical, free of any implicit bias, and could guarantee justice for all patients. Maybe you could be the one to head up this task force?!