Few industries are more in need of Artificial Intelligence (AI) than healthcare. As the technology can drive efficiency and other benefits, the health sector is pushing towards adoption to ease the workload of overburdened staff and services and deliver more value where it matters most.
From nurses and receptionist staff to doctors and surgeons, healthcare organizations across the globe are struggling to make sure regional demands are being met and illnesses are effectively treated.
In the UK, in particular, the NHS has suffered successive cuts to funding and staff and is also operating largely on severely outdated tech. This means staff and patients potentially suffer as tasks take longer and wait times skyrocket.
AI is changing that, at least partly. By using the technology to automate certain processes, scientists could, for example, spot diseases in patients more quickly, or patients could book appointments more efficiently.
“It could lead to faster, more accurate diagnoses, predict the development of disease, support doctors with treatment decisions, and help manage the demand for hospital beds,” the National Institute for Health and Care Research (NIHR) has said.
The effects could be revolutionary in an industry that is so heavily relied upon, and the real-world impact could include the elimination of diseases and the general improvement of national health conditions.
Healthcare’s improvements via AI can be summed up in three broad categories:
- Diagnosis
- Patient experience
- Staff and administration
The patient experience
The average general practitioner (GP) in the UK is responsible for 2,293 patients, according to the British Medical Association (BMA). This is an increase of 18% since 2015. With nearly 40,000 GPs in the country (37,677), that’s a huge number of individual appointments and follow-ups.
Ensuring that all of those patients - in addition to all the other patients globally in a similar position - have a seamless experience interacting with their healthcare provider is difficult.
There’s also the issue of engagement and non-compliance, in which patients who are prescribed, say, courses of treatment do not see them through to the end, or do not return for repeat appointments.
“Messaging alerts and relevant, targeted content that provoke actions at moments that matter” is a promising area in which AI could lead to increased patient engagement, according to a paper from the US National Library of Medicine (NLM).
AI will also increasingly allow patients to access “AI-generated opinions with increasing sophistication and accuracy,” according to another paper from the NLM, and it “seems likely that patients will arrive to clinical encounters with specific expectations for next steps in their care.”
While this can be helpful and promote a positive discussion between doctor and patient, it could also place strain on the dynamic if “the physician feels threatened or if families do not accept the current limitations of these tools.”
The employee factor
Staff experience can be just as important as patient experience, largely because the two are directly connected. The better tools staff have to do their job, the smoother a process patients will experience when interacting with healthcare organizations.
Many staff are craving these tools. In the NHS, a survey from Virgin Media O2 Business found that three in five IT Decision Makers (ITDMs) believe new digital tools could help the UK health service reach between three percent and five percent more patients.
On the AI front, the NHS has delivered a proof-of-concept (PoC) on an AI tool to predict long-term hospital stays. Staying in a hospital longer than needed generally correlates with worse outcomes for patients and higher mortality rates, and the longer a patient stays, the more resources are needed.
“The ability to identify and intervene early can make a real difference to these patients,” the NHS said.
Similar moves can be made in the UK’s accident and emergency (A&E) center, alleviating some of the burden that is placed on paramedic staff who need to decide which patients are taken to A&E.
Ambulances in England take around 350,000 people to A&E each month, according to the NIHR, and AI tools could help front-line medical staff work out which of those people don’t actually need to go.
Researchers developed an AI computer model and, in 80% of cases, it correctly predicted which patients did not need to attend A&E based on factors such as a patient's mobility or observations about pulse or oxygen.
Tools like this can save thousands of ambulance journeys and thousands of hours for medical staff, meaning these resources can be better used in the places that need them across the healthcare system.
Diagnosis and treatment
Perhaps the most exciting improvement AI is bringing to healthcare is through its ability to diagnose and even treat diseases, with large language models (LLMs) being trained to identify all manner of patient characteristics to make assessments about health.
Take Imagene, an Israeli AI firm that has just developed a foundation model in partnership with Oracle, designed to facilitate research and development in the pathology and oncology space.
The firm’s tool is dubbed CanvOI and has learned to capture “the complex features and patterns within biopsy images,” thus providing vital information to researchers and healthcare professionals looking to tackle cancer.
Wet age-related macular degeneration (wet AMD) - which leads to loss of central vision - can also be tackled using AI tools according to the NIHR, with the technology useful for shortening the amount of time needed to analyze scans.
A combination of the AI model and clinical experts were brought together to predict whether patients with wet AMD in one eye would develop it in their second eye (a common eventuality) within six months of the scan.
“AI correctly predicted the development of wet AMD in two in five (41%) patients. It out-performed 5 out of 6 experts,” the NIHR said.
Ultimately, AI can - and is - improving healthcare in almost every way imaginable, from the less flashy administrative tasks to the groundbreaking advancements in disease prediction and diagnosis.
Healthcare professionals must be careful with how they incorporate AI into their workflows, but they also mustn’t shy away from it - AI has revolutionary potential in the healthcare industry.
