Total Small Bowel Length Measurement Using Computed Tomography and Magnetic Resonance Imaging in Obese Patients

The total length of the small intestine (TSBL) represents a crucial parameter for obtaining a safe and successful minimally invasive surgery in metabolic/bariatric bypass surgery.

Nowadays, the standard of small intestine measurement is the intraoperative measurement. Laparoscopy represents the standard approach for baratric/metabolic, making the TSBL measurement time-consuming and risky in case of intestinal lesions. An accurate and effective non-invasive preoperative measurement of the TSBL will allow to evaluate the variability of the TSBL, which affects the surgical strategy. Cross-sectional imaging could play an important role in this setting thanks to the possibility of measuring in a non-invasive way the TSBL. Some studies performed with both Computed Tomography (CT) and Magnetic Resonance (MR) report promising results. However, they are limited by the small size of the sample, the lack of standardized technique and the lack of an automatic method based on Artificial Intelligence (AI).

The evaluation of a reliable preoperative method to measure TSBL using cross-sectional imaging will potentially reduce intraoperative complications and insufficient long-term weight loss or nutritional deficiencies. In this scenario a possible solution could be the implementation of analysis method through the development of an AI algorithm capable of automatically segmenting the small intestine.

The PRIMARY END POINT of this study is to set up and validate a reliable and reproductible automatic method to measure the TSBL in patients candidates for laparoscopic bariatric/metabolic surgery, based on preoperative radiological imaging

The main phases of the project will be:

1. evaluate the feasibility of preoperative CT and MRI-base measurement of the TSBL in a large cohort of obese patients and compare radiological measurement with intraoperative laparoscopic measurement (method of elongation) as a reference standard (1).

2. Evaluate the more accurate cross-sectional imaging between CT and MRI to measure the length of the small intestine.

3. Build an AI tool that can automatically measure TSBL on transversal slice imaging.

Three high-volume Italian centers will enroll 195 obese patients who are candidates for metabolic surgery for obesity. Part of them will be established training cohort (total = 105 patients), used to set up the AI-based method of TSBL measurement. The other 90 patients (30 for each center) will represent the validation cohort.