Artificial Intelligence and Medicine

  • Simon Paul Attard Montalto

Abstract

Currently, a ‘hot’ debate circulating throughout the more serious international media platforms focuses on the downside – or potential downside – of artificial intelligence (AI). The doomsday view warns that AI will become smarter than mankind and, essentially, become uncontrollable and take over the planet! Indeed, one AI expert has suggested that, since AI-driven processes will be far more accessible and will be put to nefarious use with relative ease by erstwhile crooks, it will become more dangerous than nuclear weapons where access and safeguards against their injudicious use is so great. In stark contrast, the use of AI-based methodology to identify Abaucin, a potentially highly effective antibiotic against acinobacter bauinaemii, was recently heralded on BBC radio as a groundbreaking advance in the fight against superbugs.1 Certainly, by using rapid gene sequencing, AI speeded up the selection process of this antibiotic from amongst hundreds of other potential contenders and curtailed the pre-clinical trial research period that would otherwise have taken several years.1 Clinical trials have yet to start and will still be required to verify the efficacy of this promising drug. Similar areas where AI-based technology has been applied include vaccine research (e.g. for RSV, Ebola),2 management and prediction in type 1 diabetes,3,4 diagnosing retinal disorders5 and increasing early detection of sepsis by up to 20%.6

 

Artificial Intelligence describes the use of computers and/or technology to simulate intelligent and critical thinking/processing comparable to human beings, a notion pioneered by Alan Turing in 1950 and first coined by John McCarthy in 1955.7 In reality, AI has since been applied in medicine for several decades and generally uses computer-based ‘virtual’ data processing essentially centred around smart algorithms/databases/pattern recognition, and ‘physical’ machine-based interventions and robotics.7 Both have already been widely applied in radiological imaging and tumour detection, diverse robotic interventions, routine workflows and equipment maintenance, screening and risk assessment programmes (e.g. in mammography), including early alert systems based on symptomatology and early detection of metastases.7-9 This exponential increase in the application of AI in medicine has resulted in a substantial increase in submissions in AI-related domains, such that dedicated Journals on AI in Medicine and AI in Healthcare already exist and the New England Journal of Medicine (NEJM) plans to launch a sub-journal aptly named NEJM-AI in 2024.8

AI programmes have significantly reduced the burden of day-to-day health administration and patient care including early clinical diagnoses and identification of non-compliance in treatment.10 AI is excellent in performing arduous and repetitive tasks that humans find tedious and rapidly lose interest, concentration and performance. Nonetheless, development of AI needs to follow clearly accepted guidelines, be aware of commercial interest and bias, avoid secrecy whilst safeguarding confidentiality, and ensure transparency. Issues relating to insurance, liability and responsibilities still need to be defined.

Undoubtedly, AI has already proven to be extremely helpful in numerous medical applications. But is it equal or better than doctors? Sometimes, the answer to this question is “yes”: one study reported that AI outperformed 17 from 18 physicians in detecting cancer on chest X-rays, but this is certainly not always the case.11,12 Will AI make doctors redundant?11Unlikely. Can it free up quality time for doctors to perform hands-on clinical work? Definitely.11 Hence, the medical fraternity should view this expanding modality as an integral and extremely useful adjunct. AI should be introduced at undergraduate level where, in all probability, already-computer-savvy students will embrace it seamlessly.

  1. J Gallagher. New superbug-killing antibiotic discovered using AI. BBC News May 2023. https://www.bbc.com/news/health-65709834.
  2. A Papi, MG Ison, JM Langley, DG Lee, I Leroux-Roels, F Martinon-Torres, TF Schwarz, RN van Zyl-Smit, L Campora, N Dezutter, N de Schrevel, L Fissette, et al., for the AReSVi-006 Study Group. Respiratory Syncytial Virus Prefusion F Protein Vaccine in Older Adults. NEJM. 2023; 388: 595-608.
  3. S Montani, R Bellazzi, A Riva, C Larizza, L Portinale, M Stefanelli. Artificial intelligence techniques for diasbetes management: the T-IDDM project. https://www.researchgate.net/publication/220836594_Artificial_Intelligence_Techniques_for_Diabetes_Management_the_T-IDDM_Project
  4. ON Ergun, HO Ilhan. Early Stage Diabetes Prediction Using Machine Learning Methods. Eur J Science Technol. 2021; 29: 52-57.
  5. Y Ji, Y Ji, Y Zhao, L Zhang. Research progress on diagnosing retinal vascular diseases based on artificial intelligence and fundus images. Front Cell Dev Biol. 2023; 11.  https://doi.org/10.3389/fcell.2023.1168327
  6. R Adams, KE Henry, A Sridharan, H Soleimani, A Zhan, N Rawat, L Johnson, DN Hager, SE Cosgrove, A Markowski, EY Klein, ES Chen, MO Saheed, M Henley, S Miranda, K Houston, RC Linton, AR Ahluwalia, AW Wu, S Saria. Prospective, multi-site study of patient outcomes after implementation of the TREWS machine learning-based early warning system for sepsis. Nature Medicine. 2022; 22: 1455-60.
  7. Amisha, P Malik, M Pathania, VK Rathaur. Overview of artificial intelligence in medicine. J Family Med Prim Care. 2019; 8:(7) 2328-31.
  8. AL Beam, JM Drazen, IS Kohane, TY Leong, AK Manral, EJ Rubin. Artificial intelligence in medicine. NEJM 2023; 388: 1220-21.
  9. D Greenfield. Artificial Intelligence in Medicine. Science News, Harvard. 2023 https://sitn.hms.harvard.edu/flash/2019/- Applications, implications and limitations.
  10. DL Labovitz, L Shafner, M Reyes Gil, D Virmani, A Hanina. Using artificial intelligence to reduce the risk of nonadherence in patients on anticoagulation therapy. Stroke. 2017; 48: 1416-19.
  11. C Sinsky, L Colligan, L Li, M Prgomet, S Reynolds, L Goeders L, et al. Allocation of physician time in ambulatory practice: A time and motion study in 4 specialities. Ann Inter Med. 2016; 165: 753-60.
  12. The Digital mammography DREAM challenge. https://www.synapse.org/#!Synapse:syn4224222/wiki/401744.

Figure

Test image

Issue
Vol. 35 No. 2 (2023)
Section
Editorial
Published
08-11-2023

License

The Editorial Board retains the copyright of all material published in the Malta Medical Journal. Any reprint in any form of any part will require permission from the Editorial Board. Material submitted to the Editorial Board will not be returned, unless specifically requested.