The role of Artificial Intelligence in modern surgery

Posted on by admin

ACHAIKI IATRIKI | 2024; 43(2):63–66

Editorial

Apollon Zygomalas1,2, Dimitris Kalles2, George Skroubis2,3

1Department of General Surgery, Olympion General Clinic of Patras, Greece
2Hellenic Open University, Artificial Intelligence in Laparoscopy Research Program
3Department of General Surgery, University Hospital of Patras, Greece

Received: 14 Nov 2023; Accepted: 13 Feb 2024

Corresponding author: George Skroubis, Professor of Surgery, Department of Surgery, University of Patras, University Hospital of Patras, Greece, Tel.: +30 2610999460, E-mail: skroubis@med.upatras.gr

Key words: Artificial intelligence, surgery, image analysis, computer vision, predictive analytics, decision support

 

INTRODUCTION

Modern medicine is undergoing a profound transformation, with the advent of Artificial Intelligence (AI) ushering in an era of unprecedented possibilities. More than any other domain, this transformation is evident in surgery. Artificial Intelligence, with its capabilities in data analysis, pattern recognition, and machine learning, is rapidly becoming a dynamic partner for surgeons, redefining the boundaries of what is achievable within the operating theater.

Historically, surgery has always demanded a delicate blend of art and science, with the human surgeon at its epicenter. The surgeon’s expertise, intuition, and dexterity have been the driving forces behind surgical advancements. However, the limitations of the human eye, hands, and mind have naturally imposed constraints on the precision, efficiency, and outcomes of surgical procedures. This is where AI steps in as a transformative force, augmenting the capabilities of surgeons and pushing the boundaries of what was once deemed impossible.

This editorial explores the multifaceted role of AI in modern surgery, investigating the theoretical underpinnings, the practical applications across clinical settings, and a discussion of the future perspectives that hold the promise of revolutionizing surgical practices for years to come.

Theoretical Foundations

AI refers to the simulation of human-like intelligence in machines and software. It encompasses a wide range of technologies and techniques that enable computers to perform tasks typically requiring human intelligence, such as learning from data, recognizing patterns, making decisions, and solving complex problems [1]. Today, the most successful AI systems are based on artificial neural networks. A neural network is a computational model inspired by the structure and functioning of the human brain. It consists of layers of interconnected nodes, each simulating a neuron. The number of neurons in an AI system can vary widely. In some simple AI models, there may be just a few hundred or thousand neurons, while in more advanced AI systems, intense learning models used in tasks like natural language processing or computer vision can have millions or even billions of neurons [1,2].

The theoretical foundations of AI in surgery are built upon principles of machine learning, computer vision, and data analysis. Machine learning algorithms enable computers to learn from surgical data and medical records, aiding in decision support and predictive modeling. Computer vision allows AI systems to interpret and analyze surgical images and videos, while data analysis helps extract meaningful insights from vast medical datasets, ultimately enhancing surgical precision and patient outcomes.

Clinical Applications

Image Analysis

AI-driven image analysis is crucial in surgery, offering precise insights and decision support tools mainly in the preoperative setting. The primary application of AI is the interpretation of medical images, such as magnetic resonance imaging (MRI), computed tomography (CT) scans, PET scans and X-rays [3,4]. AI algorithms can accurately identify and classify abnormalities like tumors, anatomical variants, or vascular anomalies. This accelerates the diagnosis process and assists surgeons in developing tailored treatment plans, ensuring more effective and personalized patient care.

Regarding intraoperative image analysis, AI can be used for Real-time Monitoring. AI systems can continuously analyze data from intraoperative X-rays, ultrasound devices or other imaging sources like head-mounted cameras, fluoroscopy, etc., enabling real-time feedback [5]. Special algorithms process the images in real time to enhance their quality, making it easier for surgeons to discern fine details and structures. This aids in accurate navigation during surgery. Systems can detect anomalies or unexpected changes like bleeding, tissue damage or the displacement of critical structures in real time, providing immediate alerts to the surgical team. In oncological surgeries, AI can help surgeons identify tumor margins in real time, ensuring that all cancerous tissue is removed while preserving healthy tissue. Augmented reality (AR) systems, backed by AI, can merge radiological images with the surgeon’s view, allowing for real-time superimposition of critical information directly onto the surgical field.

An exciting development of AI in image analysis is the field of radiomics. Radiomics involves extracting many quantitative features from radiological images often imperceptible to the human eye. Radiomics relies on advanced image processing techniques and machine learning algorithms [6]. It has the potential to revolutionize how medical imaging data is used to make diagnostic and therapeutic decisions, offering a more detailed and personalized approach to patient care [7].

Computer vision

Computer vision is an application field of AI that enables computers to understand and interpret visual information from images and videos. AI facilitates real-time and post-procedure video analysis in laparoscopic, robotic surgeries and other endoscopic procedures [8]. Computer vision algorithms interpret visual data from endoscopic cameras, providing surgeons with enhanced depth perception and augmented reality overlays [9]. Real-time tracking systems can monitor the position of laparoscopic instruments, effectively minimizing the risk of unintended injuries. Computer vision can identify various anatomical structures or anomalies during surgery, improving surgical navigation resulting in more accurate minimally invasive procedures.

Surgical videos are a perfect tool for training and skill assessment. AI video analysis provides objective and data-driven evaluation of surgical steps and instrument handling. It can also evaluate factors such as hand-eye coordination, instrument path accuracy, and tissue manipulation, allowing trainees to improve their skills progressively [10]. Computer systems generate detailed metrics and analytics of a surgeon’s performance. These metrics can be used to identify areas for improvement and develop personalized training plans.

Predictive Analytics and Decision Support

Decision support systems work in tandem with predictive analytics to assist surgeons in making complex decisions. Machine learning models leverage preoperative and intraoperative data to forecast patient outcomes and potential complications or longer hospital stays [11]. AI algorithms predict and model surgical risk, determining the most effective interventions for high-risk patients. This approach involves the analysis of large volumes of data, enabling the identification of potential risks before symptoms manifest. Utilizing electronic health records, patient history, and real-time patient monitoring enhances the accuracy and reliability of predictive models, ensuring that surgical teams are better prepared and that patients receive timely, personalized care. This predictive capacity is valuable for tailoring treatment plans to individual patient needs, optimizing resource allocation, and enhancing surgical efficiency and safety.

In trauma and emergency surgery, AI-based tools are being used to support complex analysis, aiding surgeons in making informed decisions under time-sensitive conditions [12]. Predictive analytics also plays a crucial role in anticipating and reducing risk based on current patient data. For instance, it can help determine the likelihood of a cancer patient suffering complications from surgery or being readmitted to the intensive care unit (ICU) within 48 hours of discharge. Predictive algorithms can be particularly valuable in ICUs, where timely intervention is critical to patient survival.

Moreover, predictive analytics can assist in managing surgical schedules, reducing patient wait times, optimizing operating room utilization, and improving hospital throughput and cost-effectiveness.

AI supports postoperative care by monitoring patients’ progress. AI-powered systems track vital signs, detect early warning signs of complications, and trigger alerts to healthcare providers. This constant vigilance ensures timely interventions and contributes to faster recoveries.

Future Perspectives

AI may significantly improve surgical training by providing personalized, data-driven, immersive experiences. AI-driven simulations and virtual reality environments provide surgeons with realistic training scenarios, allowing them to refine their skills in a risk-free setting [13]. AI analyzes trainees’ performances and offers real-time feedback to correct errors. It also enables remote training and collaboration among surgeons, making expertise accessible worldwide. AI-based 3D imaging and augmented reality aid in visualizing complex anatomical structures, hence raising the possibility of a better understanding of surgical field and operation steps. Overall, AI enhances surgical training by ensuring competence, reducing errors, and providing continuous access to the latest surgical advancements.

AI may play a pivotal role in telesurgery. AI could enable rapid patient triage and diagnosis by analyzing medical data and thus facilitating remote consultations and preoperative planning. It enhances decision support through predictive analytics, aiding healthcare providers in making more accurate treatment recommendations. AI could aid surgeons in remote environments by providing real-time guidance and enhancing precision. In the near future, robots driven by AI may execute surgical procedures under remote surgeon supervision, overcoming geographical barriers. AI could make surgical treatments accessible and effective, irrespective of physical location.

Machine learning algorithms can analyze vast datasets to allow robots to make real-time decisions during surgery. These robots can assist or perform complex procedures under the guidance of a surgeon, ensuring steady hands and minimizing the risk of human error. AI can enable predictive modeling to anticipate surgical complications and assist in preoperative planning. Autonomous AI robotic surgery systems can combine advanced technology with human expertise to deliver safer, more precise, and efficient surgical procedures.

Limitations of AI

Apart from those proven or anticipated advantages, AI-aided surgery also has some limitations. AI systems may struggle to adapt to unforeseen situations during surgery as they rely on pre-trained models. The effectiveness of AI can vary with the surgeon’s skill, and inexperienced surgeons may not fully leverage AI [14]. The training of AI systems requires high-quality data, and consequently, inaccurate or incomplete data can lead to errors. Surgeons may become overly reliant on AI, affecting their skills.

As AI becomes more integrated into surgical practice, ethical and regulatory considerations will become increasingly important. Ensuring patient privacy, transparency, and accountability will be crucial. Determining legal responsibility in cases of AI-related surgical errors is complex. Furthermore, compliance with healthcare regulations can be challenging. Implementing AI in surgery can be costly, limiting accessibility for some healthcare facilities.

CONCLUSION

Modern surgery is rapidly evolving with the advancements in technology. While AI offers powerful tools to improve surgical outcomes, the surgeon remains a critical part of the process. It is the surgeon’s experience, expertise, and judgment that guide the surgical process, ultimately ensuring the best possible outcomes for patients. This collaborative approach harnesses the strengths of both human skill and artificial Intelligence.

Conflict of interest disclosure: None to declare.

Declaration of funding sources: None to declare.

Author contributions: AZ and GS conceived the study, collected the literature, and wrote the manuscript; DK provided specialized knowledge and data on AI. DK and GS reviewed and corrected the manuscript.

REFERENCES
  1. Russell SJ, Norvig P, Davis E. Artificial Intelligence: A Modern Approach. Prentice Hall; 2010. 1153 p.
  2. Domingos P. A few useful things to know about machine learning. Commun ACM. 2012;55(10):78–87.
  3. Zheng Q, Yang L, Zeng B, Li J, Guo K, Liang Y, et al. Artificial intelligence performance in detecting tumor metastasis from medical radiology imaging: A systematic review and meta-analysis. EClinicalMedicine. 2021;31:100669.
  4. Vorontsov E, Tang A, Roy D, Pal CJ, Kadoury S. Metastatic liver tumour segmentation with a neural network-guided 3D deformable model. Med Biol Eng Comput. 2017;55(1):127–39.
  5. Zhou XY, Guo Y, Shen M, Yang GZ. Application of artificial Intelligence in surgery. Front Med. 2020;14(4):417–30.
  6. Shur JD, Doran SJ, Kumar S, ap Dafydd D, Downey K, O’Connor JPB, et al. Radiomics in Oncology: A Practical Guide. RadioGraphics. 2021;41(6):1717–32.
  7. Xia TY, Zhou ZH, Meng XP, Zha JH, Yu Q, Wang WL, et al. Predicting Microvascular Invasion in Hepatocellular Carcinoma Using CT-based Radiomics Model. Radiology. 2023;307(4):e222729.
  8. Anteby R, Horesh N, Soffer S, Zager Y, Barash Y, Amiel I, et al. Deep learning visual analysis in laparoscopic surgery: a systematic review and diagnostic test accuracy meta-analysis. Surg Endosc. 2021;35(4):1521–33.
  9. Kitaguchi D, Takeshita N, Hasegawa H, Ito M. Artificial intelligence‐based computer vision in surgery: Recent advances and future perspectives. Ann Gastroenterol Surg. 2021;6(1):29–36.
  10. Lam K, Chen J, Wang Z, Iqbal FM, Darzi A, Lo B, et al. Machine learning for technical skill assessment in surgery: a systematic review. NPJ Digit Med. 2022;5(1):1–16.
  11. Colborn K, Brat G, Callcut R. Predictive Analytics and Artificial Intelligence in Surgery—Opportunities and Risks. JAMA Surgery. 2023;158(4):337–8.
  12. Cobianchi L, Piccolo D, Dal Mas F, Agnoletti V, Ansaloni L, Balch J, et al. Surgeons’ perspectives on artificial Intelligence to support clinical decision-making in trauma and emergency contexts: results from an international survey. World Journal of Emergency Surgery. 2023;18(1):1.
  13. Cacciamani GE, Anvar A, Chen A, Gill I, Hung AJ. How the use of the artificial Intelligence could improve surgical skills in urology: state of the art and future perspectives. Curr Opin Urol. 2021;31(4):378–84.
  14. Hashimoto DA, Rosman G, Rus D, Meireles OR. Artificial Intelligence in Surgery: Promises and Perils. Ann Surg. 2018;268(1):70–6.