Big data and artificial intelligence (AI) are already disrupting traditional models of healthcare with cognitive computing, machine learning and predictive analytics all important catalysts set to revolutionise healthcare delivery.
Big data incorporates medical records (electronic and hard copy) scans, reports and treatment data as well as information captured from wearable health trackers and sensors. AI can be used to ‘mine’ this data at a deeper level, combining logged clinical events with real-time monitoring to reveal comprehensive non-biased data on the patient’s journey through the healthcare system.
Health consumers are already leaving a massive digital footprint across multiple platforms. This data can be analysed in detail by AI to provide tailored preventive health information. On a larger scale, AI-analysed population health data can inform the design of future healthcare systems. Using sophisticated AI-generated algorithms, big data mining has the power to analyse clinical management systems to provide better and more cost-efficient health services.
But what else can AI and big data do?
Advances in trauma care
Professor Vasa is excited about the future of AI in health having seen its possibilities first hand. Under the leadership of Prof Mouzakis, the team collaborated with clinicians at the Alfred Hospital’s Trauma Centre to develop the Trauma Reception and Resuscitation System (TR&R). The TR&R is a clinical decision support system and algorithm builder that aims to reduce errors of omission in the first 30 minutes of a patient arriving at the Trauma Centre.
“You need to protect your most precious resource in this situation, the health professionals,” says Vasa. The high-stress environment of trauma care increases the likelihood of staff deviating from standard hospital procedure. Following deployment of the TR&R, results of an independent study revealed that the system contributed to a 21 per cent reduction in errors of omission, a 30 per cent reduction in blood transfusions and significantly reduced time spent in intensive care.
“This is robust or ‘hard decision systems’ at work,” explains Vasa, “If you can analyse the clinical decision chain and work out more efficient treatment pathways, you can design a more cost-efficient system resulting in better outcomes for trauma patients, their treating health professionals and the entire hospital system.”
The medical Internet of Things
When it comes to the smaller clinic environment, Vasa sees disruption coming directly from the medical Internet of Things (mIoT) in the form of increasingly sophisticated health apps and sensor technology. “Monitoring systems that can be attached to a patient’s smart device are able to assess whether or not a visit to the GP is necessary, cutting down on travel and waiting times, and reducing costs.”
Targeted health solutions will increasingly be based on biometric data collected and analysed from a patient’s mobile device or smart watch enabling clinics to monitor each step of the patient journey, measuring outcomes, identifying gaps in treatment plans and improving access to appropriate care and services.
“Patients will be able to access tailored treatment plans and take more of an active interest in their personal health,” says Vasa. “AI systems can be designed to adapt to patient behaviour over time. They help us to understand the patient experience, improving adherence to treatment plans and resulting in improved health outcomes across diverse health populations.”
Emerging health careers
Telehealth apps already exist as part of the mIoT landscape and have enjoyed reasonably high uptake especially in rural areas, says Vasa. But he admits there are barriers interpreting vast volumes of sensor and tracking data.
“There are definitely knowledge gaps with very few technicians qualified to analyse telehealth data. The GP may or may not have the confidence or training in this type of data analysis, rendering the data useless.”
Predictive data analysis is going to be a job of the future with every new technology demanding specialised training. Health professionals will have to keep up, says Vasa, and ultimately big data, mIoT and AI will influence how medical education curricula maintains pace with the changing architecture of the health system.
According to Vasa, the jobs needed to support the health tech revolution don’t even exist yet. “Digital health coaches and telehealth data technicians are cyber careers that are just biding their time,” he says. As the sensitivity and performance of sensor technology and mIoT expands over the next few years, specialist big data analysts will emerge, powering up to progress AI across all areas, including healthcare.
Preparing for big data
When it comes to preparing your clinic environment for the big data future, it pays to learn as much as you can about AI and health data, says Vasa. What are his top tips for big data?
1. Read up on the emerging trends in health tech so you can recommend user-friendly apps that encourage collaborative development of patient-centric treatment plans.
2. Remember that consumers are more engaged than ever in the management of their health. Include questions about the use of health apps and sensor technology when you take a patient’s medical history.
3. Attend digital health conferences and enrol in big data training when it becomes available.
Rajesh spoke at Australian Health Week in March 2017; his presentations explored how EMR and AI systems can give clinicians quality information that can reduce treatment costs, eliminate unnecessary hospital procedures and provide patient-centric treatment plans.