Artificial Intelligence and Urology

Artificial intelligence (AI) is now impacting our lives in so many ways, it is increasingly shaping what we see, believe and do. The magnitude of development in AI and other related technologies in the last ten years is mind-boggling. OpenAI was founded only ten years ago and generative AI/agentic AI was not yet recognised then. TikTok was not launched, 5G connectivity was experimental and Google DeepMind had not discovered AlphaFold, for which Demis Hassabis received the 2024 Nobel Prize in Chemistry. [1]

Development of AI has not been smooth across all aspects of life. There has been gradual progress in speech, handwriting and image recognition over the years. However significant progress has been made in the last ten years in reading comprehension, language understanding and predictive reasoning primarily through wider use of Large Language Models.

A timeline to show the rapid change in computer technology [2]

“Artificial intelligence is transforming our world; it is on all of us to make sure that it goes well” [3]

Why does it matter to us?

There are currently a small number of people at a few tech firms working on artificial intelligence (AI) who understand how powerful this technology is now becoming. If we do not become engaged, then it will be this small elite who will decide how this technology will change our lives.

Most significant advancement of AI in healthcare have so far been in the diagnostic specialities such as radiology, pathology and dermatology, where there are most acute workforce issues. Of over 1000 AI based medical programs approved so far by FDA, majority are focused on these three areas. [4]

Medical and surgical specialities are lagging behind in applications powered by AI. This is however going to change with recent advances through wider use of generative AI in predictive analysis of surgical patients by processing large multimodal datasets, including data from registries, imaging data, electronic health records and wearables.

These developments will ultimately help develop digital surgical pathways by capturing data across patients’ surgical journey. This will ultimately lead to development of ‘digital surgical twins’ which would have real-time impact on patients’ journey.

Intraoperative Assistance: Guidance and Execution of Simple Tasks: In robotic surgery, AI will analyse operations as they’re being performed and potentially provide decision support to surgeons as they are operating, AI will also be able to perform simple tasks through the robot, including closing a port site and tying a suture or a knot. It is not expected that we will have completely autonomous surgical robots anytime soon.

AI’s Role in Medical Education and Training: A recent study found that version 4 of the ChatGPT bot can answer Membership of the Royal College of Physicians (MRCP) written examination questions, without additional prompts, to a level that would equate with a comfortable pass for a human candidate. [5]

American College of Surgeons have developed an online educational course ‘Artificial Intelligence and Machine Learning: Transforming Surgical Practice and Education’, on how AI technology can inform clinical decision-making and help surgeons more accurately assess risk, predict disease progression, and manage patients. The program includes eight modules. [6]

Real-world Data: Developing highly predictive algorithms in the future will depend on improving the granular quality and diversity of the data that is being used to train AI models. This would mean using the health records of a large number of patients with tens of thousands of variables. [7] The power and accuracy of AI prediction models will depend on access to data from a diverse pool, and representative of the population where such AI programmes will be deployed.

Despite these potential opportunities, there are challenges of how to regulate rapid technological advances brought about by pace and complexity of innovation, balancing speed with safety in the resources-constrained environment. Important issues include; ethical considerations, cybersecurity, liability and public trust in wider rollout of AI based medical technologies

Surgeons should look at AI as “an opportunity to augment the great work we do more than as a threat to what we do”.

Potential AI applications in surgery could include

Admin:

• Improving triage and referrals to surgical services
• Perform secretarial and administrative roles
• Generating operation notes
• Writing referral letters and other correspondence
• Managing on-call rotas and workforce requirements
• Automating follow-up investigations and requests
• Managing waiting lists

Surgical patient workflow:

• Making diagnoses of diseases and conditions treated with surgery
• Preoperative risk assessment and stratification (e.g. risk scores, body scanning, reviewing electronic records and investigations)
• Clinical documentation (e.g. patient interactions, outpatients, discharge summaries)
• Delivering preoperative information to support informed consent
• Improving internal handover processes between medical colleagues
• Collating data and information for multidisciplinary team discussions
• Postoperative coaching to support personalised approaches to recovery
• Providing an ‘expert’ second opinion
• Applying evidence-based surgical practice (e.g. reviewing research and providing rapid summaries)
• Prescribing and making suggestions for medication optimisation

Surgical planning:

• Interpretation of medical imaging (e.g. X-rays, CT scans)
• Rapidly creating 3D models for operative planning

Surgical skills:

• Intraoperative surgical skill and competency assessment
• Training of surgeons outside of the operating room (e.g. knowledge, anatomy, procedural steps, simulation, augmented reality)
• Improving the precision of robotic and minimally invasive surgery
• Training semi-autonomous robots to perform common operations and procedures

Written by:

Prof. Naeem Soomro

Council member Royal College of surgeons England.

Co Director NorthFutures Digital Health Hub

[1] Nobel Prize in Chemistry (2024), https://www.nobelprize.org/prizes/chemistry/2024/summary/

[2] Roser, M. (2022) The brief history of artificial intelligence: The world has changed fast – what might be next? Our World in Data. Available at: https://ourworldindata.org/brief-history-of-ai.

[3] Roser (2022), Artificial Intelligence, Our World in Data https://ourworldindata.org/artificial-intelligence

https://ourworldindata.org/brief-history-of-ai

[4] FDA (2024), AI/ML‑Enabled Medical Devices https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-software-medical-device

[5] O’Connor et al. (2024), ChatGPT‑4 MRCP Performance, BMJ https://doi.org/10.1136/bmj.q675

[6] American College of Surgeons (2024), AI & Machine Learning Course https://www.facs.org/for-medical-professionals/education/programs/artificial-intelligence-and-machine-learning-transforming-surgical-practice-and-education/

[7] Rajpurkar et al. (2022), AI in Healthcare & Multimodal Data https://www.nature.com/articles/s41746-022-00656-7