Data-driven Human-centric Surgery

By Julio Mayol, Professor of Surgery, Chief Medical and Innovation Officer, Instituto de Investigación Sanitaria San Carlos, Universidad Complutense de Madrid, Hospital Clínico San Carlos, Madrid, Spain.

Luis Sánchez-Guillén, Colorectal Surgeon, Department of Surgery, Hospital General Universitario de Elche. Universidad Miguel Hernández. Elche, Alicante, Spain.

Technology has had a tremendous impact on both surgeons and surgical patients over the last 30 years. Minimally invasive surgery, a disruptive technological innovation, emerged in the last decade of the 20th century and changed the old surgical paradigms, forcing surgeons to anarchically acquire new skills that were unnecessary for conventional procedures. Patients benefited from this disruption by getting faster postoperative recovery, better cosmesis, and fewer approach-related complications at the price of a transient increase in intraoperative complications.

This is true for most patients in developed countries. In contrast, many people all over the world do not have access to high quality surgery because many surgeons, not only those in developing countries, cannot get adequate surgical training (1).

Recent technological breakthroughs (3D imaging for endoscopic surgery, robotic assisted-surgery, powered staplers, tissue-sealing devices, etc) are basically designed to increase operative precision and productivity. Unfortunately, other powerful computing innovations will not be so rapidly incorporated into surgeons’ practice. Big data analytics and cognitive technologies are still far from being implemented in our medical schools, clinics, wards and operating theatres. This is due to several factors, such as lack of clear understanding of the potential benefits, short-sighted data-storage strategies, incomplete understanding of surgeons’ cognitive processes, surgical complexity, unstructured scenarios, traditional business models at risk, variations in definition and assessment of relevant patients’ outcomes, and fear of change, among others.

It may not be obvious that healthcare is undergoing a profound transformation. However, surgeons, as other healthcare professionals, will be forced to evolve from service providers to value generators for their patients. Personalised and population value will be highly demanded by society in the near future (2). In order to achieve that objective, major changes in training will be required, not only with regard to technical dexterity and spatial orientation, but also to cognitive learning and decision-making. Unfortunately, there is limited knowledge about how surgeons learn, think. make decisions and execute actions to produce the best outcomes. In fact, little is known about what features define a good surgeon.

Surgical data science (3) is key to tackle these limitations. Generating big data repositories from structured (preoperative and intraoperative imaging, haptics, lab tests, electronic medical records, serious games, etc) and unstructured data (clinical notes) will be needed to train neural networks that replicate how surgeons function (Figure 1) and the kind of outcomes they get. These simulations will make possible to build robust models to relate cognitive processes with technical dexterity, performance and outcomes. As a result, the validated models could be used to train surgeons and assist them in clinical practice through the use of computerized cognitive training platforms. Eventually, autonomous surgical robots will become a reality.

Computerized cognitive training platforms (4) will become essential to globally:

  • Train surgeons to analyze clinical scenarios and critically think.
  • Capture data to understand how surgeons make decisions and operate.
  • Enhance surgeons’ performance.
  • Simulate clinical scenarios, procedures and outcomes.

However, there are still some barriers and limitations for the incorporation of surgical data science into practice. Although the IDEAL framework (5)could be useful to guide de incorporation of these surgical innovations, some questions remain: Is simulation enough to validate these technologies? When must the results of these solutions be evaluated? How should patients be engaged in the assessment?

References

  • Bunogerane GJ, Taylor K, Lin Y, Costas-Chavarri A. Using Touch Surgery to Improve Surgical Education in Low- and Middle-Income Settings: A Randomized Control Trial. J Surg Educ. 2018 75:231–237. https://doi.org/10.1016/j.jsurg.2017.06.016.
  • Gray M. Value based healthcare BMJ 2017; 356 :j437 https://doi.org/10.1136/bmj.j437
  • Maier-Hein L et al. Surgical data-science: enabling next-generation surgery https://arxiv.org/pdf/1701.06482.pdf
  • Kahol K, Vankipuram M, Smith ML. Cognitive simulators for medical education and training. J Biomed Inform, 2009,43: 593–604, https://doi.org/10.1016/j.jbi.2009.02.008.
  • McCulloch P, Cook JA, Altman DG, Heneghan C, Diener MK; IDEAL Group. IDEAL framework for surgical innovation 1: the idea and development stages. BMJ. 2013

Jun 18;346:f3012. https://doi.org/10.1136/bmj.f3012

Figure 1. Functional analogy between surgeons and computers

2018-05-23T13:29:27+00:00