This post is by guest blogger Matthew Savage
Previously, the 4-year mumps vaccination held the crown for the fastest generation of a vaccination. But, with the announcement that a new vaccination for Covid-19 may be available in the coming months, just 1 year on from detection, it is interesting to consider how the vaccine has been created in such a short amount of time. Of course, billions of pounds have been thrown at the research, along with thousands of the finest scientific brains across the world. However, artificial intelligence has also played a huge role in speeding up the process. Vaccine development in the past used machine learning as a “minor sidekick” (Chen, 2020) but more recently AI has been used heavily in the quest for a vaccination, helping researchers understand the virus, its structure and to predict which pieces of the virus the immune system is most likely to recognize. With such a fantastic contribution from AI in the area of research and development, the question remains; will AI replace medical professionals anytime soon?
What can AI do better than a human?
- Speed and accuracy – It is well documented that AI is fantastic at object detection (a computer vision technique that allows us to identify and locate objects in an image or video), and research suggests that AI has improved imaging exponentially. For example, AI was found to analyse specific lung nodules between 62% and 97% faster than radiologists. In addition to this, an AI-assisted robotic technique created by Mazor Robotics resulted in a five-fold reduction in surgical complications vs. surgery completed by a human surgeon alone (Kalis, 2018). Such research provides compelling evidence for AI’s superiority in speed and accuracy and, with further future developments, which could indicate the replacement of surgeons, radiologists and other imaging professionals.
- Reduced costs – Robot assisted surgery is predicted to save the US health service $40 billon by 2026 (Kalis, 2018) and other types of software enable doctors and nurses to have reliable, precise patient information at the touch of a button, saving time and money. For example, the introduction of the da Vinci Surgical system, a minimally invasive robotic surgical system, allows surgical procedures to take place with increased accuracy and decreased costs (Yu, 2020).
- Reducing invasive and risky procedures – AI has the ability to identify certain diseases with far more accuracy than medical professionals, based on scans and dynamic modelling. For example, HeatFlow Analysis is a cloud-based service that enables the detection of coronary artery disease based on CT scans of a patient’s heart. AI can then run a model of blood running through the vessels and can identify abnormalities from the model. Without this, surgeons would have to perform invasive angiograms (Hurtado, 2020). Robots can also provide more precision and less pain through feedback systems which check pressure and grip (Kakar, 2017). These advancements have changed the role of a surgeon significantly already and could change the role even further in the future.
What makes it unlikely that AI will not completely replace health professionals?
Despite superior accuracy, speed and a reduction in costs, it seems unlikely that AI can ever fully replace human doctors. There are several reasons as to why this seems to be the case.
- Patient trust – An argument can be made to patients that they are “safer” in the hands of AI, using the points above. However, patient trust in AI is a real issue when it comes to patient care. Many patients see themselves as individual and unique and value personalisation in their care. When 200 students were offered a free assessment to provide diagnosis of stress levels, 40% signed up when a human doctor was to perform the diagnosis but only 26% when a computer was to offer the diagnosis (Longoni and Morewedge, 2019). Humans trust humans and this is certainly not a bad thing. Automation bias can lead to an overreliance on technology and this can lead to errors by an AI system (Coiera, 2019). That is why we need a trusted doctor who can check the information to fully trust it.
- Creative and critical thought – Although AI can spot objects and is great at voice recognition, it is not so good at question answering, can be overwhelmed by multiple voices in the same room and can also lack critical thought. Patients are not pre-selected data sets and obtaining a diagnosis is not a linear process.
- Empathy - Krittanawong (2018) points out that AI “cannot engage in high-level conversation or interaction with patients to gain their trust, reassure them, or express empathy, all important parts of the doctor–patient relationship.” Empathy is reflective and patients need to trust a physician to really engage in the therapeutic process. This is particularly important in mental health treatment.
Conclusion
In a data-led world, there is no doubt that AI will continue to increase its influence in the medical field. From increased accuracy and responsibility in surgical procedures, to helping overworked physicians complete assessments and other routine tasks with greater accuracy, AI will certainly support physicians in increasing their ability to provide important face-time to patients and to focus more time on complex cases. AI will provide faster research and development in the search for vaccinations, scouring the literature, experimental data and running multiple, complex algorithms in the search for answers. But AI will not replace the human touch, the listening ear or the trusted professional patients require.
Sparta Health provide their services with this in mind, including mental health triage practitioners, counsellors, psychologists and other healthcare services providers. Please contact us for more information.
About Matt Savage
Matthew Savage has an MSc in Psychology, is a qualified personal trainer, and has worked within the field of cognitive rehabilitation for 5 years. He is an FA qualified football coach, with a keen interest in moral behaviour and wellbeing within team sports.