(MENAFN- Gulf Times) Large language AI models such as ChatGPT are transforming professions and sectors of the economy. Some of the most significant impacts are in medicine
Large language artificial intelligence (AI) models such as ChatGPT have become markedly more advanced in recent years, with huge potential benefits – yet also risks – across probably the full range of human activity. Some of the most important developments are in healthcare.
Some forms of AI in healthcare are well established – for example tools that can detect from scans the early symptoms of a range of conditions, including cancers, diabetes and dementia. But large language models create more intimate and powerful applications. As an experiment, ChatGPT itself was used to provide a summary of medical applications for this article. This produced a list of five significant examples:
Medical imaging – accurate and efficient analysis of scans.
Diagnosis and treatment – ability to analyse large amounts of patient data helps doctors identify patterns and provide personalised treatment plans.
Drug discovery – identifying potential and predicting likely efficacy of new drugs, cutting development time and costs.
Medical records – healthcare providers can analyse trends and developments.
Patient monitoring –real-time monitoring provides early warning signals.
Another application is the use of AI as a training tool for medical students – by being able to draw on a wide range of data and patients, an AI computer can challenge the trainee with questions, advice and help with decision-making. Physical robots are also likely to have a role, even acting as an assistant in surgery by stitching wounds.
Prognoses through AI can increase the accuracy in projecting the course and outcome of a patient's condition, advising the medical team on a date for discharge and the ongoing treatment plan.
The US Food & Drug Administration (FDA) has approved over 500 medical AI models, and around 100 in 2022 alone.
Public health management can be transformed for the better. Virtual connections and automated diagnostic tools can mean that populations living in rural areas without a nearby hospital or clinic can have improved access to specialist diagnosis and other forms of care.
Healthcare providers are beginning to identify both the benefits and the hazards of AI, and to appreciate that the advantages can be substantial. One of the biggest benefits can be in cost savings; moreover the type of savings achieved can have a multiplier effect because, by freeing doctors and nurses from analysing scans and data, and from administrative tasks, medical staff can have more time with patient care. They will also have more time to update themselves on medical research – for which AI tools can help by selecting a reading list, appropriate for their specialism.
Hospitals can therefore cut costs and improve the standard of care simultaneously with the right application of AI. This is an especially important consideration given high healthcare cost inflation in recent years owing to supply issues, increased longevity and cost of treatments.
Costs can be substantially reduced in development and testing of new drugs, with AI able to identify patients for testing and analyse trial results more thoroughly and rapidly. With pre-AI approach, the development of a new drug could take up to 15 years and $2bn. There are further cost savings in insurance-based healthcare systems, because AI tools can be used to detect financial fraud, including insurance fraud, as noted in my earlier article on AI in finance.
Qatar has taken a noteworthy step in advancing healthcare services and promoting the development of health-related artificial intelligence (AI) applications through the establishment of the Qatar Biobank and the Qatar Genome Program. These initiatives have generated substantial amounts of big data, including whole genome sequencing and other omics data, which are of great value for research and innovation in the field of AI health applications.
Researchers and innovators can leverage these large datasets to utilise AI algorithms for data analysis and the generation of new insights into disease diagnosis, treatments, and patient outcomes. A prime example of interdisciplinary research findings on the use of AI for disease diagnosis is described in a research paper published by the College of Science and Engineering at Hamad Bin Khalifa University. The study employs deep learning algorithms to analyse retinal images and bone health data collected by the Qatar Biobank. The AI system was trained to identify markers associated with cardiovascular diseases and demonstrated high diagnostic accuracy.
The development of this infrastructure is paving the way for innovative new applications that have the potential to revolutionize healthcare and enhance outcomes for patients.
The usual cautionary notes for applications of AI must always be added: the most technologically sophisticated systems are useless or even dangerous if the data is inaccurate. There are also privacy concerns when it comes to the use of patients' medical records, and cyber-security has to be a priority. Transparency is the best safeguard against misapplication, so attention needs to be paid to governance arrangements and upholding an ethical code.
AI machines are sophisticated, but they are still machines, and are best deployed by those who are expert in the domain to which it is being applied – in the same way that powerful machinery is safer and more effective in the hands of a trained specialist operator. But with responsible governance and specialist oversight, AI can help ensure substantial improvements for public health and well-being, and cost savings for health departments.
The author is a Qatari banker, with many years of experience in the banking sector in senior positions.
MENAFN21052023000067011011ID1106283336
Legal Disclaimer:
MENAFN provides the information “as is” without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the provider above.