The goal of this research initiative will be to explore the development of artificial intelligence (AI) tools which can be used to objectively analyze general surgical operations, extract key insights, and provide specific learning moments and recommendations to surgeons.
With the provided data, we will use a global team of expert annotators and AI algorithms to identify the critical moments of a procedure. The data will be processed by both humans and AI algorithms to produce analytics on a variety of procedures. Advances in Surgery (AIS) Channel will develop the guidelines for how these critical portions of the procedure are identified, and the technical analysis will be performed with the assistance of a third-party company. AIS Channel may use the surgical videos to create de-identified data which may be used and disclosed on a royalty free basis as AIS Channel deems appropriate.
To develop AI systems for surgery, we need numerous examples to teach the AI system to recognize key steps in each operation. The surgical video data from laparoscopic and/or robotic surgical procedures will be de-identified so that the patient’s identification remains anonymous. The data set will then be analyzed in aggregate. By providing data to this effort, surgeons and patients will be helping bring the objective analysis of surgical technique into continuing medical education. It is essential to have a large volume and variety of surgical procedures present in the data set to produce a robust analysis.
This initiative will help to develop a surgical video dataset capturing specific examples on how surgeons approach operations and solve problems. The ultimate goal is to extract learnings and insights that can benefit surgeons and their patients in surgical procedures worldwide.
To assess the role of the AI tools which can be used to objectively analyze general surgical operations, extract key insights, and provide specific learning moments to surgeons. Data will be used to develop machine learning algorithms and produce analytics based on defined targets of annotation.
This is an international multicenter descriptive observational study that will analyze historical (retrospective) and prospectively collected operational data with a particular focus on identifying the critical moments of the procedure.
Population and sample size
De-identified surgical videos (regardless of the year of performance) that include complete scheduled surgeries performed by minimally invasive approach (both laparoscopic and robotic, understanding robotic surgery as minimally invasive surgery carried out, totally or partially, through any of the robotic platforms available on the market, e.g., Da Vinci surgical system). A total of 4,000 surgical videos are planned. Surgical Procedures that will be targeted include (1000 of each):
Primary:
For the creation of an automatic model of recognition of the steps of each surgery as well as to recognize operational activities based on their impact on task-based and surgeon efficiency metrics, a 3-phase approach will be used:
The study protocol was reviewed and accepted by the Ethical Committee of the coordinator center, the Hospital Clinic of Barcelona. In each collaborating center, the decision of the Competent Ethics Committee (CEC) and the competent authority concerning the conduct of the study will be made in writing to the Sponsor and Principal Investigator before commencement of this study. The clinical study can only begin once approval from all required authorities has been received. Any additional requirements imposed by the authorities shall be implemented. The responsible investigator at the investigation site ensures that approval from an appropriately constituted CEC is sought for the clinical study.
It is expected that the results obtained will be of great interest to the scientific and surgical community. The AIS research team will plan to publish the results in scientific journals with the maximum possible Impact Factor, being, in any case, a priority that is included in the D1 of each subspecialty.
In any case, the external communication of results will be initially and preferentially through publication in a scientific journal. Once the work is accepted, it will be disseminated in general media: scientific community, and scientific symposiums and medical journals.
The research team from AIS will be in charge of the weight of the medical, scientific part, carrying out the interpretation of the data obtained after the visualization, labeling, and analysis of all the videos sent by the collaborators.
For scientific publications extracted from this study, surgeons will be credited based on the number of videos submitted as indicated below. In the first place, the first authors of each participating center will appear ordered by the number of videos sent and then in alphabetical order. Next, the second, third and fourth authors will appear ordered according to the same criteria.