This study investigated abdominal wall distension due to pneumoperitoneum using RGB-D imaging. Abdominal wall distension varied considerably between individuals, with global distension ranging from 1.3 to 5.8 cm (mean 3.6 ± 0.9 cm). Distension occurred primarily in the anterior direction, with a greater magnitude in the central abdominal regions compared to the lateral regions. Univariate and multivariate analyses showed that global distension is independently associated with patient height, age, medio-lateral region, and cranio-caudal region, suggesting that both individual patient characteristics and location on the abdominal surface influence the abdominal wall distension due to pneumoperitoneum. The results of this study underscore the importance of patient-specific simulation of pneumoperitoneum to provide realistic, interpretable trocar placement guidance to surgeons.

The presented study shows that global wall distension is independently associated with patient height, age, medio-lateral region position, and cranio-caudal region position. This is in line with the findings of Malbrain et al., who described that abdominal compliance is higher in patients with increased height and age [10]. They also suggested that BMI and sex influence abdominal wall compliance, which contradicts the findings of the present study. Yildirim et al. also found a positive correlation between age and compliance, and no correlation between BMI and compliance, thus confirming our findings [11]. Furthermore, our observations that abdominal wall distension has greater magnitude in the medial regions are consistent with the previous reports [8, 12]. Those reports indicate that pneumoperitoneum mainly causes expansion in the central abdomen, likely due to higher compliance of the midline and uniform pressure distribution during pneumoperitoneum.

While port placement is performed based on a fully insufflated abdomen, our study findings offer insights for pre-operative insufflation simulation. This simulation can be used for education or personalized pre-operative surgical planning. The findings of this study offer two crucial implications. First, the measured variation in global wall distension between patients highlights the individual variations in abdominal wall distension due to pneumoperitoneum. These variations suggest that standardized pneumoperitoneum simulation may not be optimal for all patients, thus emphasizing the need for patient-specific pneumoperitoneum simulation methods. Second, we found that abdominal wall distension was independently associated with patient height and age. These relationships suggest that it may be possible to simulate abdominal wall distension preoperatively, using those patient characteristics. Additionally, we observed that lateral abdominal regions show smaller wall distension, compared to the medial part of the abdomen. This spatial variation in shape change can inform more precise surgical planning. Together, these insights contribute to a more personalized understanding of the operative scene and support the development of patient-specific simulations to guide trocar placement in minimally invasive surgery.

For patient-specific trocar placement, multiple factors must be considered, such as the procedure type, the modality (robotic vs. laparoscopic), the target organ, the patient’s abdominal shape, and how this shape changes due to abdominal wall distension caused by pneumoperitoneum. This study focused specifically on this abdominal wall distension due to pneumoperitoneum, and its relationship with patient characteristics. To our knowledge, this is the first study that used RGB-D cameras to generate realistic data to quantify the abdominal wall distension automatically and to assess its relationship with patient characteristics. Unlike previously used methods utilizing, for example, external motion tracking systems, our acquisition method enabled a non-invasive method of acquiring data without influencing patient treatment, surgery duration, or abdominal deformation [8].

As a next step, future research should aim to develop a patient-specific trocar placement planning approach that incorporates all the mentioned patient-specific factors. A first step in this planning would be the prediction of the abdominal shape, including its changes due to pneumoperitoneum, enabling the creation of an intuitive, realistic simulation of the intra-operative situation. Initially, this representation could assist surgeons during trocar placement. In the future, this knowledge could be used for automatic trocar placement planning. While recent studies have aimed to provide such a realistic simulation of the patient, including a simulation of pneumoperitoneum, most methods are based on mathematical models and/or validated on porcine data, instead of realistic human data [13,14,15,16,17,18]. The dataset collected in this study provides the opportunity to build and validate more accurate, human-based models.

This study has several limitations that warrant consideration. First, the quantification of abdominal shape change was based on a relatively small dataset, although this dataset is bigger than previous studies that investigated abdominal wall distension due to pneumoperitoneum [8, 12]. In our study population, BMI was not associated with abdominal wall distension. The current population had a relatively narrow BMI distribution (27.6 ± 4.5), which limits the assessments of the correlation. Assessment based on a more diverse group could prove a correlation between BMI and abdominal wall distension. Second, while patient characteristics extracted from the EHR were included in our analysis, prior research suggests that abdominal wall distension may also be influenced by factors observable in Computed Tomography (CT) imaging [12]. Future research should focus on incorporating a larger dataset with comprehensive CT imaging, to investigate the relationship between CT-derived features and abdominal wall distension due to pneumoperitoneum. Third, the visibility of the abdomen during application of pneumoperitoneum was limited. As the patients underwent upper abdominal procedures, the sterile drapes partially covered abdominal areas, specifically below the umbilicus and the lateral abdomen. The cranial edge of the sterile drapes was placed near the inferior border of the sternum. Mark placement and the subsequent analysis of regional wall distension were restricted to the visible surface, which led to the exclusion of a substantial portion of the abdomen. The limited field of view may also explain the lack of displacement in the lateral direction. In future studies, ensuring wider abdominal exposure or the inclusion of lower abdominal surgery could improve the accuracy and completeness of regional analysis. Finally, chest wall motion from mechanical ventilation may have contributed to minor additional movements of the abdominal wall. Although these movements were small compared to pneumoperitoneum-induced distension, we did not compensate for this effect.