We look at the challenges and potential benefits that artificial intelligence (AI) may bring to the healthcare sector.
Estimated growth rates for Artificial Intelligence (AI) vary, but all highlight an expectation that there will be significant annual growth in AI over the next 10 years, with a substantial portion of that growth within the healthcare sector.
It’s easy to imagine how AI might positively impact healthcare planning and operations, but what challenges are preventing rapid uptake and what potential benefits might AI really bring?
Frequently, AI is noted as having the potential to complete tasks typically handled by people, but faster and cheaper. However, in healthcare, tasks are complex and rely on multiple streams of information and a range of skillsets.
For any AI system to either replicate or improve on existing operations and service provision, several challenges need to be overcome; we consider some of them, along with the potential benefits AI may bring to the healthcare sector.
Current challenges to the adoption of AI in healthcare
AI systems have been noted in recent years as providing insight into, and the capability of, diagnosing and providing solutions towards healthcare issues. However, a range of challenges still exist that prevent adoption of AI across the healthcare industry. These include:
Fragmented uptake of technology: Healthcare systems are understandably complex, and like other industries have had to adopt new technologies as and when they demonstrate a clear benefit to operations and/or patient care. Unfortunately, the rate of adoption frequently differs across geographical areas, even within the same country. This leads to a fragmented system of willingness and adaptability to innovate.
Offering a system that will reduce inequality across areas and provide a benefit to patients and workers will be a planning, technical and administrative challenge. To date, AI systems developed by IBM and Google (Watson and DeepMind respectively) have demonstrated a clear capability to improve existing systems in targeted areas, including oncology and neuroscience. Translating these benefits across a wider range of specialisms will require an ecosystem approach, which might take a considerable amount of time to implement.
Making sense of the data: Whilst healthcare systems may collect vast quantities of data, the quality of which is uneven, the way it is collected may follow more traditional methods. In many cases the data is gathered via paper forms, for example, while information that is kept in silos will prevent an AI system from making best use of it. However, fixing this problem will take a substantial amount of time and investment – and that’s before investing in an AI system to work with the data after it has been digitised and sorted.
Using computational practices for holistic treatment: Applying computational systems to quantified information, that can be sorted and analysed with a clear target, may be an easy initial first use for AI in healthcare. However, healthcare is more complex than reaching a diagnosis and ‘treatment’ isn’t always about curing a problem: applying AI to mental healthcare and emotional wellbeing is a harder challenge – and removing or reducing the ability for individuals to access meaningful support in place of an automated tool may have a negative impact. Applying a ‘success’ condition in mental healthcare and emotional wellbeing is subjective to the individual; how an AI system could be trained to understand and respond to this appropriately will require additional consideration.
Though the complexity of planning and implementing AI across a healthcare system may be daunting, the potential benefits cited by those developing such systems are many.
As noted above, healthcare systems produce and rely on vast quantities of data, but only currently utilise what is collected from within the healthcare system: if an individual visits a GP or hospital, basic personal information and symptom specific information is collected. While this helps treat the condition immediately requiring intervention, it does not necessarily mean information is used effectively to support long-term health and overall wellbeing.
AI may help healthcare systems make better use of existing data, considering patterns and patient history when diagnosing a current problem or predicting a possible future one.
However, AI may also enable healthcare services to utilise data generated by new technologies such as wearables, offering insight into patients’ daily lives with consistent, quantified information on patient activities. By adding a wider range of information, including data that is consistently generated each day, AI may be able to identify patterns or anomalies, leading to quicker diagnosis and treatment of some conditions.
Alongside this, automating part of the data analysis for diagnosis may support quicker processes of much larger quantities of information. The benefits of this could range from clinical healthcare improvements to shortening time spent on admin-based tasks. In the long-term, this type of analysis may also support more precise approaches for treatment plans, offering a tailored plan that meets the unique needs of each patient.
There are several challenges, and many assumed benefits, of implementing AI at scale in healthcare systems. Transitioning towards this approach will likely take time and investment.
With risks so potentially great, and the consequences of something going wrong having an impact on human lives, slow progression will offer technology developers and healthcare sector workers the opportunity to better understand areas for immediate uptake and long-term development.
To support the development of AI in relation to healthcare systems, grant funding is available for businesses looking to get involved. Alongside this, organisations such as Health Innovation Manchester provide insight into how businesses may begin to work with healthcare organisations when a product is at the right stage.
If you’re interested in hearing more about AI and healthcare, or if you have a product you’re looking to develop, reach out to GC Business Growth Hub’s Innovation team today.
This blog is part of a series of articles focussing on different areas in healthcare that are embracing –or could embrace – new innovations. Next time, we will consider telemedicine, the growth of remote healthcare services, and how adapting to the restrictions of COVID-19 is shaping how we’ll access healthcare in future.
Clare Cornes, Innovation Development Manager (University of Salford)
Clare joined the Business Growth Hub as the Innovation Development Manager for the University of Salford in July 2019. Within this position, Clare uses her passion for new technologies and innovation to support SMEs in working with the University.
Prior to this role, Clare has led an autonomous vehicle development and trials programme for a British automotive manufacturer; managed multiple UK and European funded projects that utilised new technologies to improve local challenges; written national and international position papers analysing new innovations in relation to health and sustainable transport initiatives; and inputted into regional transport strategies to ensure new technologies are considered when designing schemes to solve city region challenges.
Alongside professional roles, Clare is also undertaking a PhD in her spare time, researching the barriers and challenges associated with implementing a sustainable Mobility as a Service (MaaS) system in Greater Manchester, including the policy and regulatory considerations. The research includes understanding what MaaS means in practical terms for transport planners, policy makers, related businesses and users. Through this experience, Clare has developed a skill for translating technical developments into socio-economic impacts and is keen to support SMEs developing innovative products and services as part of their business growth.