Artificial intelligence is set to transform healthcare by accelerating diagnoses, reducing hospital backlogs and increasingly predicting illnesses before symptoms emerge, but human clinicians will remain central to medical decision-making.
Clinicians can rest easy in the knowledge that technology will not replace specialist roles in the healthcare industry, however. Instead, the role of AI will range from improving operational efficiency to assisting in diagnoses and, ultimately, predicting diseases before they can develop and spread through a patient's body.
Dutch health technology firm Philips has embedded AI across a range of diagnostic imaging equipment, including MRI, CT and ultrasound systems, where the technology is primarily focused on improving workflow.
One example is the company's Philips SmartSpeed technology for MRI scanners, which can significantly reduce scan times.
Faster imaging not only makes procedures more comfortable for patients, particularly those experiencing pain, but also allows hospitals to scan more people each day, addressing growing diagnostic backlogs in many markets, according to Atul Gupta, Global Chief Medical Officer, Diagnosis and Treatment at Philips.
"If you can get more people in and out of that scanner, we can significantly decrease that waiting time," Gupta said.
"It's particularly important for things like cancer. For instance, in the UK, not making a diagnosis with a CT or MRI can turn a stage one cancer to an untreatable higher stage cancer.
"So you really want to get that early."
Beyond efficiency, Philips is increasingly deploying AI as a clinical decision support tool.
In stroke care, AI embedded within CT scanners can automatically identify tiny blood clots responsible for strokes, alert specialist teams and prioritise the scan for immediate review by a radiologist.
Similarly, AI-assisted prostate MRI software can automatically highlight suspicious regions that may require biopsy while also identifying patients suitable for ongoing monitoring instead of immediate intervention.
Prevention will always beat treatment
The next frontier for Philips will be predictive medicine.
Recent advances in cloud computing and AI are enabling healthcare providers to anticipate disease before conventional diagnostic methods detect any abnormalities.
Among the technologies demonstrating that shift is the Philips ePatch, a lightweight wearable ECG monitor that patients wear continuously for up to two weeks.
Right now, the device is primarily used to detect intermittent heart rhythm disorders, including atrial fibrillation, that may only occur briefly, therefore bypassing traditional monitoring.
But clinical studies are beginning to show the technology could go considerably further.
The implications extend beyond detecting irregular heart rhythms. Because atrial fibrillation increases stroke risk significantly, earlier identification allows clinicians to prescribe preventative medication before a potentially devastating event occurs.
Gupta cited the case of a patient in the Netherlands, who suffered a stroke, despite initial hospital monitoring showing no evidence of atrial fibrillation.
After being discharged, he was instructed to wear an ePatch. Clinicians detected a brief episode of the condition, allowing them to identify the underlying cause and begin treatment to reduce the risk of another stroke.
The wearable also reflects broader advances in cloud-based healthcare, replacing older Holter monitors that relied on bulky and cumbersome recording equipment, and could typically only monitor patients for one or two days.
Could AI fix systemic healthcare understaffing?
Beyond improving patient outcomes, AI can play a significant role in addressing mounting workforce shortages in healthcare systems globally.
Many hospitals are struggling to recruit nurses, sonographers, and other specialists, as ageing populations increase demand for care.
AI can help bridge part of that gap by enabling less experienced clinicians to perform complex diagnostic procedures with greater confidence.
One example is cardiac strain imaging, an advanced ultrasound examination used to detect the earliest signs of heart failure in cancer patients undergoing chemotherapy.
This examination previously required highly experienced sonographers because of its complexity.
However, by embedding AI into the imaging workflow, Philips has largely automated the process, allowing junior staff to perform the examination while producing consistent, quantifiable results.
The benefits extend beyond productivity - new technology will also help retain and attract staff. For specialists, surgeons and clinicians, using outdated equipment and observing reduced efficiency can be frustrating.
"Staff want to work in a 'techie' place," Gupta explained. "They don't want to just use the same piece of equipment, or do the same old thing over and over again. They like to be challenged."
Despite rapid advances, Gupta cautioned that healthcare innovation inherently moves more slowly than consumer technology.
While Silicon Valley companies often embrace the idea of "move fast and break things", that philosophy is not quite as practical in a health setting.
"It takes us seven years, on average, to get some of our innovations out," he said.
"That's why we do some interesting things. We don't just test our technologies with senior physicians. We also want to test our technologies out on physicians of tomorrow.
"At our innovation centres in the United States, China, and the Netherlands, we actually bring teenagers into our operating rooms. We have them play with our interfaces. We get some very interesting insights."
These sessions provide data on the ways digital-native generations interact with emerging technologies such as augmented reality and voice-controlled systems.
Looking ahead to the short and medium-term, healthcare innovation will be defined less by any single breakthrough than by the convergence of multiple technologies.
Artificial intelligence, robotics, augmented reality, voice interfaces and image-guided minimally invasive procedures will increasingly combine inside operating theatres, creating the 'surgery suite of the future.'
Shadow AI becoming industry-wide issue
There's no doubt that AI can bring myriad benefits to the health industry.
But Philips' latest Future Health Index survey uncovered some concerning statistics around shadow AI.
Sixty-four per cent of 2000 physicians surveyed now use consumer AI chatbots to assist with patient diagnosis, rising to a whopping 79 per cent in China.
While the figures reflect mounting workload pressures on clinicians, they also expose significant risks.
Consumer AI systems are not clinically validated, remain susceptible to hallucinations, and are prone to confidently generating incorrect information.
"AI has its share of problems," Gupta said. "AI hallucinates. It does so with authority.
"It acts like it really knows. It has sycophantic behaviour. It tells you what it thinks you want to hear. There are huge dangers still with shadow AI."