Table of Contents |
|---|
| Abstract |
| Introduction |
| Main Text |
| References |
Emerging AI advancements have had a large impact on society. Services such as ChatGPT, and facial detection and recognition for mobile phones have transformed the way we live. In particular, the healthcare industry has begun to introduce AI into the field. This article presents the influence of artificial intelligence in medicine for medical diagnostic imaging, patient diagnosis and robotic surgery. Medical diagnostic imaging analysis focuses on using AI tools to examine and identify patterns, such as irregularities, in medical scans. There is potential for artificial intelligence to be used for patient diagnosis. A study has shown that by combining AI with voice recording analysis, it is possible to detect type 2 diabetes in individuals. Robotic surgery refers to procedures that are performed with the assistance of robotic systems. The use of robotic surgery allows for more precision and stability when compared to manual techniques.
Artificial intelligence has had a transformative impact in the field of medicine. In this article, the applications of AI in medical diagnostic imaging analysis, patient diagnosis, and robotic surgery, are explored through the use of studies.
Artificial intelligence has taken many roles in medicine. Currently, the most common role for AI is medical imaging analysis. This is where artificial intelligence tools are being used to analyse and identify abnormalities in x-rays, CT scans and other scans that might be missed by professional healthcare staff. It is very likely that AI will become an integral part of the healthcare system. Research has shown that AI systems are as effective as human radiologists at finding abnormalities and detecting signs of cancer as well as other conditions (Lang et al, 2023).
A growing area of AI research is on developing AI algorithms for predicting the risk of disease in patients, as opposed to simply detecting patterns in images (Sharma et al, 2020).
The most accurate method to diagnose type 2 diabetes is through a blood test. However, researchers from Klick Applied Sciences have developed a tool that may become the next technological advanced test to use at home to diagnose type 2 diabetes. There have been many studies that show that there is an influence of type 2 diabetes on the voice. A systematic review in 2019 concluded there was a “higher prevalence (12.5%) of voice problems among individuals with diabetes […] as compared to the general population”. (Ravi et al, 2019). This includes hoarseness, increased strained voice and excessive throat clearing.
The study took the voice recording data of 267 individuals, who were diagnosed as nondiabetic or type 2 diabetic on the basis of American Diabetes Association guidelines, and analysed them for voice features. The data was collected using a mobile application. Participants recorded an exact phrase up to 6 times a day for 2 weeks which resulted in 18,465 recordings in total. The recordings were analysed using an AI model for the voice features separately for men and women, and then in an age-matched and BMI-matched model, and were used to predict whether the participant had been diagnosed with type 2 diabetes. Figure 1 (Kaufman et al, 2023) shows the procedure for the initial model through a flowchart of the analysis and the machine learning process.
Figure 1: Flowchart of analysis and machine learning process. T2DM refers to type 2 diabetes mellitus.
The most accurate model for women was created by averaging the female voice recording results with the BMI frequency of type 2 diabetes. The most accurate model for men was found by averaging the male recording voice results and cross-referencing this with the age and BMI predominance of type 2 diabetes. The variation in the acoustics of voice features found that women with type 2 diabetes reported a “slightly lower pitch with less variation”, while men “reported slightly weaker voices with more variation.” (Kaufman et al, 2023).
Ultimately, the study concludes that there is a promising future for voice analysis using AI to predict the diagnosis of type 2 diabetes. The results with the study were encouraging, given that the most accurate prediction method found a maximum test accuracy of 89% for women and 86% for men.
There are currently an estimated 850,000 people living with diabetes in the UK who have yet to be diagnosed (Diabetes UK, 2023). The approval of this AI tool may lead to a quicker and more efficient method to diagnose diabetes at home.
The American company ‘Intuitive Surgical’ created a surgical system called ‘Da Vinci’ in honour of Leonardo DaVinci who studied human anatomy. The ‘Da Vinci’ system can perform complex, minimally invasive surgery via a console that is controlled by surgeon (Hamet et al, 2017). As of December 2022, over 7500 da Vinci systems were in operation in hospitals worldwide, with more than 11 million robotic surgeries carried out with Da Vinci robots (Guthart, 2023).
Robotic surgery offers a spectrum of benefits and consequences for patients. Robotic surgery offers patients a shorter hospital stay, less risk of infection and faster recovery (The Leeds Teaching Hospitals, 2023). However, there are also concerns about the reliability of robotic surgery. In a study by the Severance Hospital, Yonsei University Health System in Seoul, Korea, 10267 Da Vinci robotic surgeries were carried out in 7 departments by 47 surgeons. This study allowed an extensive analysis based on multiple surgeries across multiple departments. Among the 10267 procedures performed at Severance Hospital, system malfunctions were reported in 185 cases (1.8%) as seen in Table 1 (Koh et al, 2018) which shows the failures and malfunctions using the system. 32 cases had a mechanical failure/malfunction error, 23 had a system error, while 130 had an instrument error. Most of the malfunctions were related to an instrument error, but this was easily solved by simply replacing the malfunctioning instrument. Mortality was linked to 12 Da Vinci surgery cases (0.12%) (Koh et al, 2018).
Table 1: Mechanical Failure and Malfunction of Robotic Surgery using the Da Vinci Surgical System
Ultimately, the study concluded that robotic surgery was a safe alternative to traditional surgery. This study could be limited as it only assessed the safety of robotic surgeries during a specific period of time at a single hospital which would not be reflective of the practice of robotic surgery as a whole. Given that the mortality rates from using the Da Vinci system are only recorded at this institution, it does not represent global trends.
In conclusion, AI has made remarkable strides in medicine. The integration of AI in medical diagnostic imaging, patient diagnosis, and robotic surgery is the just the beginning of what AI will be able to achieve.
Lång, K., Josefsson V., Larsson, A.-M., Larsson, S., Högberg, C., Sartor, H., Solveig Hofvind, Andersson, I. and Rosso, A. (2023). Artificial intelligence-supported screen reading versus standard double reading in the Mammography Screening with Artificial Intelligence trial (MASAI): a clinical safety analysis of a randomised, controlled, non-inferiority, single-blinded, screening accuracy study. Lancet Oncology, 24(8), pp.936–944. Available at: https://doi.org/10.1016/s1470-2045(23)00298-x.
Sharma, P., Suehling, M., Flohr, T. and Comaniciu, D. (2020). Artificial Intelligence in Diagnostic Imaging. Journal of Thoracic Imaging, 35, pp.S11–S16. Available at: https://doi.org/10.1097/rti.0000000000000499.
Ravi, R. and Gunjawate, D.R. (2019). Effect of diabetes mellitus on voice: a systematic review. Practical Diabetes, 36(5), pp.177–180. Available at: https://doi.org/10.1002/pdi.2240.
Kaufman, J.M., Thommandram, A. and Fossat, Y. (2023). Acoustic Analysis and Prediction of Type 2 Diabetes Mellitus Using Smartphone-Recorded Voice Segments. Mayo Clinic Proceedings: Digital Health, [online] 1(4), pp.534–544. Available at: https://doi.org/10.1016/j.mcpdig.2023.08.005.
Diabetes UK (2023). Number of People Living with Diabetes in the UK Tops 5 Million for the First Time. [online] Diabetes UK. Available at: https://www.diabetes.org.uk/about_us/news/number-people-living-diabetes-uk-tops-5-million-first-time.
Hamet, P. and Tremblay, J. (2017). Artificial intelligence in medicine. Metabolism, [online] 69(69), pp.S36–S40. Available at: https://doi.org/10.1016/j.metabol.2017.01.011.
Guthart, G. (2023). Intuitive Surgical J.P. Morgan Healthcare Conference 2023. [online] Available at: https://isrg.intuitive.com/static-files/6683d2bb-75e2-4fa0-b0cd-463ead7c30a4.
The Leeds Teaching Hospital (2023). Robotic Surgery. [online] www.leedsth.nhs.uk. Available at: https://www.leedsth.nhs.uk/a-z-of-services/urology/procedures/robotic-surgery/.
Koh, D.H., Jang, W.S., Park, J.W., Ham, W.S., Han, W.K., Rha, K.H. and Choi, Y.D. (2018). Efficacy and Safety of Robotic Procedures Performed Using the da Vinci Robotic Surgical System at a Single Institute in Korea: Experience with 10000 Cases. Yonsei Medical Journal, 59(8), p.975. Available at: https://doi.org/10.3349/ymj.2018.59.8.975.
