Generally speaking, artificial intelligence (AI) refers to computer systems that have the potential to perform complex human tasks. When applied to healthcare, novel AI technology, and debate surrounding its usage, have been met with both excitement and fear. Any technology which rivals human ability has the potential to revolutionise the way healthcare is provided globally, significantly improving efficiency and accuracy. However, many are still unsure if the current technology is safe enough to handle human health. This poses the question, is AI in medical care still a distant dream? To answer this question, let’s first take a look at some of the impacts AI has recently had around the world, and the recognisable potential it has for the near future.
Improving the efficiency of…
… research centres. Machine learning technologies have already been employed to accurately sieve through vast medical databases and huge quantities of scientific papers, at a much greater speed and lower cost than humans are able to do.
… finding new drug formulas. With improved research capabilities, AI technology also has the potential to come up with new ideas, testable hypothesis and theories which would usually take humans multiple lifetimes. Additionally, it is often thought that humans are more likely to become doggedly attached to the theories they create, even if those theories encounter serious drawbacks. AI technology reduces the likelihood of any emotional attachment to medical research which hinders innovation and change.
… clinical care. AI technology is already able to delineate the edges of particular tumours in the body in minutes; a task often done by hand over up to four hours with limited consistency and reduced accuracy. Similar improvements have been seen in the provision of burn care and limb transplants.
… of skin cancer. A recent study conducted at the University of Heidelberg found that AI computers were able to identity 95% of skin cancers, where humans were only able to identify 86.6%. Minute yet crucially important indicators are often invisible to the human eye.
… of breast cancer. Kheiron Medical Technologies in London recently found that AI software has exceeded the required ability level for radiologists screening for breast cancer. Interestingly, a recent Economist article argued that over-treatment and over-diagnosis of this particular form of cancer are common in the U.K., and second opinions are often required. AI technology, therefore, already has the potential to rapidly reduce the amount of time and money spent on checking scans in a country where 1 in 8 women will likely be diagnosed with breast cancer during their lifetime.
… of head injury severity. A Kings College surgeon, Chris Mansi, recently founded Viz.ai., a firm which used machine learning technologies to analyse which brain injury patients needed urgent care and which patients are able to safely wait. Based on personal experience as a brain surgeon, Mansi believes that numerous deaths could be saved annually by prioritising the severity of injuries, especially in road accidents.
… of atrial fibrillation. In 2017, Apple revealed an app which uses the built-in heart rate monitor in Apple watches to flag to patients with irregular heartbeats that they may be experiencing atrial fibrillation and should seek doctor’s advice.
Improved patient care, including…
… the provision of medicine. Artificial pancreases have already been created that are able to measure glucose levels in the blood using highly accurate sensors, before delivering appropriate insulin levels accordingly.
… the accuracy of prescriptions. AI technology is already able to tailor suggested medical care in line with the predicted response which a particular patient may have, based on factors such as their genetics.
… the empathy levels of care provided. Interestingly, some machines have even been trained to detect particular emotions of patients to encourage doctors to engage more emphatically during treatments and appointments.
Undeniably then, AI technology has already had noteworthy impacts on the way healthcare is provided in the U.K, and has visible potential to produce further benefits in the very near future. According to the evidence, these benefits are most often found in enhanced research, diagnostics and predictions. Technology news source, VentureBeat, even found that 55 out of 218 healthcare start-ups were focused on improving predictive resources. However, high barriers to entry exist. Exceptionally high regulation in the healthcare industry means that clear enough accountability and justification for the use of such complex and relatively untested technology is currently very difficult to provide. Whilst evidence suggests that AI has the potential to solve healthcare’s most prominent mysteries, such as the development of a HIV vaccine, medical AI is unlikely to ever make human experts entirely redundant in the fields they invade. In fact, the think tank, Medium, suggested that even the most intelligent technology is unlikely to ever fully be able to rival the strength of the human doctor-patient relationship fostered by face-to- face interaction, especially in primary appointments.
If in doubt, see a doctor face-to-face for a thorough diagnosis. Download the Qured app today for a doctor appointment in comfort of your own home within 2 hours: onelink.to/jtygac.