By Jemima Lawrence
Introduction
Artificial intelligence (AI) is defined simply as ‘the capacity of computers or other machines to exhibit or simulate intelligent behaviour’. AI is already widely and successfully used in healthcare in prevention, diagnosis and treatment. However, given the innovations in AI and with increased adoption there is potential for a huge improvement to healthcare services and outcomes.
AI’s current role in healthcare
We are already utilising AI to prevent, diagnose and treat illnesses within multiple global healthcare systems. For example in the UK, the NHS has used AI as a way of diagnosing patients through automated analysis of X-ray images, e.g. mammograms. This is extremely beneficial to radiologists as it provides them with support in making assessments. This in turn makes the whole process more efficient, allowing for radiologists to spend more time with patients and increasing the rate at which screens are carried out. AI is also being used to help clinicians read brain scans more quickly and this leads to quicker treatment and better quality of care for the patient.
As a way of treatment, the NHS is utilising AI to support people in virtual wards. These patients, who would otherwise be in hospital, are able to receive care and treatment from their own place of residence. In addition, remote monitoring technology, such as apps, allow for the monitoring of a patient’s health condition while at home. Healthcare-associated infections (HAIs) are a huge problem, so being able to be treated from home could be extremely beneficial to those who are more vulnerable to infections. In the United States the Centre of Disease Control and Prevention has estimated that roughly 99,000 deaths each year are related to HAIs. This perfectly illustrates the benefits of AI’s role not just in treatment but also in preventing disease.
But what could AI look like in the future of healthcare? How could it be further developed and what impact will further developments have?
The future of AI in prevention
AI has been invaluable in the past to track infectious diseases. During the COVID-19 pandemic, in 2020, we saw evidence of AI’s benefits in contact-tracing apps. AI algorithms could determine the risk of cross infection, through a smartphone user’s location and alert users of such risks. This allowed for those without the infection to be alerted if they had a risk of being in contact with someone who had contracted COVID-19 symptoms.
In the future, AI could be used once again to help predict outbreaks and high-risk areas, whilst also monitoring disease progression to improve prevention. AI could have a huge impact on disease suppression, as it can be used to assist vaccine development by predicting how viruses or other causative agents will evolve over time. Delivering vaccines faster means preventing the effects of the spreading disease at an earlier stage, which could be extremely beneficial to those who are vulnerable.
The future of AI in diagnosis
AI could be utilised in the future of diagnosis through image recognition and predictive modelling. According to Havard’s School of Public Health, using AI to make diagnoses may improve health outcomes by 40%.
Although there is much concern about the inaccuracy of some models, a study found that AI could actually recognise skin cancer better than experienced doctors. The researchers used deep learning (a type of AI) on over 100,000 images to identify skin cancer. They compared the results of AI against 58 international dermatologists, and found that AI did better. Due to AI being able to be trained on thousands of images, it can recognise patterns or abnormalities that may indicate disease or injury much faster than a radiologist, therefore offering faster treatment.
Predictive modelling is ‘a mathematical process used to predict future events or outcomes by analysing patterns in a given set of input data’. AI algorithms are able to analyse large amounts of data from electronic health records, genetic testing results, lifestyle factors, environmental factors and more, to help identify if an individual is at risk of a particular disease.
The future of AI in treatment and research
AI could be used to treat disease in many different ways including assisting the development of drugs and developing treatment plans. According to the California Biomedical Research Association, for a drug to travel from the research lab to the patient it takes an average of 12 years and shockingly only 1 out of 5,000 of the drugs that begin preclinical testing are ever approved for human usage. A drug will cost on average US $359 million to develop from the research lab to the patient. AI could play a significant role in the future of drug development by streamlining the drug discovery and drug repurposing process which could in turn save companies a huge amount of time and money.
Generative AI (AI used to create new content) could be used to assist the development of treatment plans for patients by analysing diverse patient records including medical records, treatment history and lifestyle factors. Machine Learning (a type of AI) can be used to identify patterns and correlations from this data to develop personalised treatment strategies and assist healthcare professionals in making data-driven decisions. This ensures that the most effective and appropriate treatment options can be advised to and received by patients.
General disadvantages of AI in healthcare
Drawbacks of AI in healthcare include the threat it poses to certain jobs within the healthcare industry. A lot of administrative roles could be made redundant. In addition, those using the AI models will need extensive training to learn how to use them properly, which may take them away from other important duties. Also the tools themselves need to be trained with carefully curated data so as to avoid any unintentional bias.
Furthermore, it could be argued that although it removes the majority of human-based error, when AI models are dealing with large amounts of data, inaccuracies can occur and within healthcare any serious inaccuracies could be fatal.
Conclusion
In summary, AI will inevitably have a huge positive impact on healthcare in the future as its benefits will certainly outweigh the negatives that come alongside it. Two clear benefits will be improved diagnosis of disease and accelerating vaccine development. It will rapidly improve the efficiency of healthcare in many different settings.
However, it will remain necessary to run any AI test or data analysis results by professionals, in order to limit any inaccuracies. In addition, AI lacks emotional intelligence so it is vital to have doctors, nurses and surgeons to advise patients and make them aware of the risks when making life changing decisions.
Bibliography
NHS. (2023, February 10). Artificial Intelligence - NHS Transformation Directorate. NHS England. (Accessed 12 Oct. 2023, 29 Oct. 2023) Available at: https://transform.england.nhs.uk/information-governance/guidance/artificial-intelligence/
NHS. (2023, August 9). Virtual wards - NHS Transformation Directorate. NHS England. (Accessed 12 Oct. 2023) Available at: https://transform.england.nhs.uk/information-governance/guidance/virtual-wards/
IBM Education. (2023, July 11). The Benefits of AI in Healthcare. IBM. (Accessed 12 Oct. 2023) Available at: https://www.ibm.com/blog/the-benefits-of-ai-in-healthcare/
GlobalData Healthcare. (2023, August 10). AI is Essential For The Future of Drug Discovery and Development. Pharmaceutical Technology. (Accessed 12 Oct. 2023) Available at: https://www.pharmaceutical-technology.com/comment/ai-drug-discovery-development/
FutureLearn. (2023, May 2). What is AI in Healthcare?. FutureLearn. (Accessed 12 Oct. 2023) Available at: https://www.futurelearn.com/info/blog/what-is-ai-in-healthcare
PwC. No Longer Science Fiction, AI and Robotics are Transforming Healthcare. PwC. (Accessed 12 Oct. 2023) Available at: https://www.pwc.com/gx/en/industries/healthcare/publications/ai-robotics-new-health/transforming-healthcare.html
Morgan, D.J., Lomotan, L.L., Agnes, K., McGrail, L., Roghmann, M.-C. (2010, August 31). Characteristics of Healthcare-Associated Infections Contributing to Unexpected In-Hospital Deaths. National Library of Medicine. (Accessed: 29 Oct. 2023) Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3528178/
Emerging India Analytics. (2023, September 27). AI-Powered Diagnostics: The Future Of Early Disease Detection and Prevention. LinkedIn. (Accessed: 30 Oct. 2023) Available at: https://www.linkedin.com/pulse/ai-powered-diagnostics-future-early
Sharma, V. (2023, August 29). AI For Healthcare Apps. BusinessofApps. (Accessed: 30 Oct. 2023) Available at: https://www.businessofapps.com/insights/ai-for-healthcare-apps/
Comments