AI in health care: the burning question that will only be answered with time
By Amit Yadav*, CEO of GE Healthcare, Australia and New Zealand
Wednesday, 29 November, 2023
We’re not in Kansas anymore. Artificial intelligence (AI) is thrusting the world forward at an unprecedented speed, and health care is at the coalface of this monumental shift.
Rightly so, many of us are still comprehending the possibilities — or should I say, impossibilities — of AI: how it’s going to impact daily life, change clinical practice and improve patient outcomes, and how on earth we’re going to manage its lightning-fast expansion.
The global sense of caution is palpable — and in many cases, it is warranted. Tech industry CEOs and country leaders have in recent months come out with firm warnings against allowing AI to evolve unregulated and unchecked.
But I firmly believe the pros far outweigh the potential cons at this point.
AI is often seen as merely enhancing the capabilities of everything around us. However, in health care, the vision is very real and immediate — quicker access to highly personalised protocols and a significantly enhanced quality of care.
We already know the importance of data to the patient care pathway. But AI has the power to seamlessly carry that data across in-home, primary and tertiary settings, as well as clinical disciplines.
Look at cardiovascular disease (CVD) — Australia’s leading driver of ill health and premature death and arguably the world’s most pressing health concern today following the COVID-19 pandemic. CVD contributes to about 5.1% (600,000) of all Australian hospitalisations every year and costs the healthcare system billions upon billions of dollars. Unfortunately, these figures are not going down: they’re rising.
Enter AI-powered cardiology diagnostic and monitoring devices, which are already revolutionising early detection and prevention of this devastating disease group.
By 2033, health care’s dynamics are poised to shift significantly, with systems coming to individuals’ homes. In the home, many wearable devices equipped with AI algorithms can provide users with valuable insights into their health and fitness, continuously monitoring vital signs, activity levels and sleep patterns.
These in-home devices can transmit data to healthcare professionals via applications, massively reducing the time lag between symptoms arising and treatment.
Looking at how AI is improving diagnostics in primary and tertiary settings and the evidence is equally, if not more astounding. Perhaps the most common outpatient diagnostic procedure, electrocardiograms (ECG), are undergoing their own AI revolution. From automated interpretation of complex results to detecting arrhythmia, abnormalities or even drug-induced or congenital long QT syndrome (LQTS) — all of which would have taken days or weeks in the past. This saves lives and the invaluable time of clinicians, who can spend more time with more patients. The more efficient and accurate the diagnostic procedures, the better.
But AI is not only helping doctors treat those who already foresee a risk of CVD, it is also forecasting it. Using patient data, including medical history, lifestyle factors and real-time monitoring data, AI assesses individualised and population-level cardiovascular risk and is assisting with early interventions and future modelling.
Clearly, AI is changing our world in an almost endless number of ways. Combined, these major changes will have a positive impact on health system access and equity, performance and the patient pathway.
The issue, however, is whether we can collectively keep pace with that limitless future while holding a few sacred and steadfast principles.
For one, medical technology manufacturers, including us at GE HealthCare, should not seek progress for progress’s sake. That’s to say, we must be honest enough to say “when” on AI. Clinical decision-making must be protected, and we should view this evolution as another opportunity to empower our healthcare professionals, not the other way around. This is precisely why companies like GE HealthCare are creating a set of ethics around how AI will and won’t be used in the healthcare industry.
Patient consent, privacy and security are also just as much the responsibility of technology manufacturers as for the clinicians. Now, with the ability to share patient data across settings in the blink of an eye, it falls on the healthcare technology industry to ensure systems collect interoperable data securely and prioritise patient privacy above all else.
Likewise, AI algorithms are only as good as the data they are trained on. If the data is biased, the algorithms will also be biased. When it comes to empowering healthcare and diagnostic technologies with AI, it is critical that we seek to achieve and re-evaluate algorithmic fairness and address potentially problematic biases as they arise. The consequences of not enshrining this standard in future practice could be disastrous for patients the world over.
Nevertheless, we are at an exciting juncture in our global healthcare journey, and AI’s arrival and expansion into medical technologies promises more benefits than not. Perhaps most obviously is in the world of diagnostics, where preventing people from suffering ill health and even death is becoming more possible every single day.
But the challenges are evident and healthcare delivery is increasingly dependent on accurate patient data and precision technologies to ensure informed clinical decision-making, safe and unbiased therapeutic interventions and positive outcomes.
Entering this exciting period of change, I encourage my fellow healthcare stakeholders (leaders, organisations and policymakers) to think carefully and devote time to these challenges, so society can overcome the sense of caution and mistrust for AI that is permeating across industries. Only then will we be able to truly answer the question underlying all these issues.
Can we come to trust artificial intelligence in what is our most sacred human asset — our health?
 |
