Improved safety for minimally invasive surgery by incalculation of pathology-specific acute and chronic effects of tissue manipulation.

The lack of intraoperative haptic feedback limits the benefits that are associated with robotic minimally invasive surgery. The development of mathematical models to implement force information in surgical robots requires biological data describing the relationship between mechanical loading and consequential tissue damage. In this dissertation we aimed to quantify this relationship for aortic clamping, a procedure that is often crucial during surgery. We have shown that in vivo clamping of mouse thoracic aorta results in damage on both the functional and morphological level, which partly persists up to one month after the initial clamping injury.

To investigate if implementation of patient-specific information is necessary in the developed models, we examined whether or not age and pathology influence the relation between clamping force and tissue injury.

We observed a trend towards impaired endothelium-dependent relaxation in aging mice, Probably this difference in response would become significant when mice older than 40 weeks would be included in the study.

Given the high prevalence of atherosclerosis in patients who need surgery, we also studied the effect of aortic clamping in a mouse model displaying atherosclerotic lesions (LDLR knockout mice fed western diet). We could not find a consistent genotype-based difference in response to arterial clamping at intima level. However, investigation of effects located in the tunica media showed that atherosclerotic arteries have a higher osteopontin score (synthetic smooth muscle cells) compared to wildtype vessels. Quantification of curvature showed more flattening of the innermost elastic membrane in response to clamping in atherosclerotic samples compared to wildtype segments.

While flattening did occur in both genotypes, no recovery was observed in LDLR-/- arteries. These conclusions suggest that atherosclerotic arteries are more prone to medial damage, especially in the lighter clamping load range. Additionally, we observed a consistent rise in inflammatory cells (CD45) after clamping. Interestingly, wildtype inflammation scores were higher than LDLR-/- levels in the intima, while media analysis showed an opposite pattern.

We can conclude that in vivo clamping of arteries results in damage to the vessel wall, even when utilizing low clamp loads. While the comorbid factor aging did not affect the results significantly, atherosclerotic arteries respond differently to mechanical injury compared to healthy vessels. These results suggest that mathematical models to be implemented in surgical robots should account for possible pathology at the tissue of interest. This would improve safety during and after minimally invasive surgery.