The AIDE/OBS Study: an international multicenter protocol for AI-ready clinical and intraoperative visual data collection in non-elective minimally invasive surgery for complicated diverticulitis

In the high-stakes world of emergency surgery, doctors often face a split-second dilemma that can define a patient's future. When a section of the large intestine becomes dangerously inflamed or infected, a condition known as complicated diverticulitis, surgeons must decide how to repair the damage. They can either reconnect the two ends of the bowel immediately, restoring the body's natural path, or they can remove the damaged section and bring the healthy end out through the abdominal wall to create a temporary opening, known as a stoma. This second option, historically called a Hartmann's procedure, is a safer bet for unstable patients but leaves them with a bag to manage and often requires a second, difficult surgery later to reverse it. The choice between these two paths is rarely clear-cut. It depends on a complex mix of the patient's overall health, the severity of the infection, and, crucially, what the surgeon sees inside the body the moment they begin the operation. For decades, this decision has relied heavily on a surgeon's experience and intuition, a process that is difficult to study because the reasoning happens in real time, often under pressure, and is rarely recorded in a way that others can analyze.

A new international research effort, led by a team of surgeons and scientists, is attempting to change how we understand this critical moment. The study, known as AIDE/OBS, does not test a new drug or a new surgical technique. Instead, it is building a massive, organized library of information to see if we can teach computers to help surgeons make these life-altering choices. The researchers are gathering data from hospitals around the world, focusing specifically on patients who are undergoing urgent, minimally invasive surgery—using small cameras and instruments rather than large open cuts—to treat severe diverticulitis. Their goal is to collect not just the standard medical notes, but also the actual video footage and still images from inside the operating room. By pairing these visual records with detailed information about the patient's condition and the surgeon's final decision, the team hopes to create a structured dataset that captures the invisible logic of surgical judgment.

The project is designed as a feasibility study, meaning its primary purpose is to prove that this kind of data collection is possible across different countries and hospitals. The researchers are asking participating centers to record specific moments during surgery: the initial look inside the abdomen, the removal of scar tissue, the assessment of how well the bowel is healing, and the final decision to either reconnect the intestine or create a stoma. They are also asking surgeons to write down exactly why they made their choice, noting factors like how much infection was present, how healthy the tissue looked, and whether the blood supply seemed strong enough to support a connection. This information is being gathered from at least 150 complete cases, a number chosen to provide a solid foundation for future analysis without being so large that it becomes unmanageable in this initial phase.

What makes this approach unique is its focus on the visual evidence that usually disappears once the surgery is over. In traditional medical studies, researchers might know that a surgeon chose to reconnect the bowel, but they often lack the context of what the surgeon actually saw to make that call. Did the tissue look too swollen? Was there too much pus? Was the blood flow poor? The AIDE/OBS study aims to capture these visual details by organizing video clips into specific phases of the operation. This allows the researchers to link the surgeon's real-time observations directly to the outcome. If a surgeon decides to create a stoma, the study records the visual cues that led to that decision, such as the presence of widespread contamination or poor tissue quality. If they choose to reconnect the bowel, the study captures the visual signs of health and stability that gave them the confidence to do so.

The team is not yet building an artificial intelligence tool that can make decisions for surgeons. Instead, they are laying the groundwork for such a tool by ensuring the data is high-quality, consistent, and ready for computer analysis. They are testing whether it is possible to collect standardized video and clinical data from dozens of different hospitals, each with its own equipment and recording habits. The study acknowledges that emergency surgery is messy and unpredictable, with conditions that can vary wildly from one patient to the next. By focusing on a specific group of patients—those stable enough to undergo minimally invasive surgery but sick enough to require urgent care—the researchers are creating a controlled environment to study these complex decisions. They are also paying close attention to the use of special cameras that can show blood flow in real time, a technology that is becoming more common but is not yet fully understood in emergency situations.

The findings of this initial phase will not tell us which surgical strategy is better for every patient. The study is not designed to prove that one method is superior to the other. Instead, it will reveal whether we can successfully build a database that connects the visual reality of the operating room with the clinical decisions surgeons make. If the data collection is successful, it will provide the raw material needed for future researchers to develop computer programs that can recognize patterns in surgical videos. These future tools could one day offer surgeons a second opinion, highlighting visual signs of risk or safety that a human eye might miss in the heat of the moment. The ultimate hope is that by understanding the visual and clinical factors that drive these decisions, medicine can move toward a future where the choice between reconnecting the bowel and creating a stoma is based on clear, reproducible evidence rather than just individual experience.

The study has already attracted interest from more than 300 surgical centers worldwide, suggesting that the need for better data is a shared global challenge. Surgeons from different countries, working in different healthcare systems, are joining forces to record their cases. This collaboration is essential because the way surgeons handle these emergencies can vary significantly depending on their training and the resources available to them. By bringing these diverse perspectives together, the study aims to identify which factors are universal and which are specific to certain environments. The researchers are careful to note that this is just the first step. They are not claiming to have solved the problem of emergency surgery, but they are demonstrating that it is possible to capture the complexity of these decisions in a way that computers can eventually learn from.

In the end, the AIDE/OBS study is about making the invisible visible. It seeks to turn the fleeting, subjective judgments of a surgeon into a permanent, analyzable record. By doing so, it opens the door to a new era of surgical research where artificial intelligence can assist in the most critical moments of patient care. The work is methodical and deliberate, focusing on the quality of the data rather than the speed of the results. If successful, this project will not replace the surgeon's skill but will provide a powerful new tool to support it, potentially reducing the number of patients who undergo unnecessary stoma surgeries and helping more people return to a normal life after a life-threatening illness. The journey from a messy, unpredictable emergency room to a structured, data-driven future begins with this simple act of recording what happens, one case at a time.