Intraoperative Perfusion Assessment in Enhanced Reality Using Quantitative Optical Imaging: An Experimental Study in a Pancreatic Partial Ischemia Model.

Taiga Wakabayashi,Raoul Pop,Anne-Laure Charles,Eric Felli,Takeshi Urade,Emilie Seyller,Manuel Barberio,Bernard Geny,Margherita Pizzicannella,Michele Diana,Pietro Mascagni,Andrea Baiocchini,Yuta Abe,Yuko Kitagawa,Jacques Marescaux

doi:10.3390/diagnostics11010093

Abstract

To reduce the risk of pancreatic fistula after pancreatectomy, a satisfactory blood flow at the pancreatic stump is considered crucial. Our group has developed and validated a real-time computational imaging analysis of tissue perfusion, using fluorescence imaging, the fluorescence-based enhanced reality (FLER). Hyperspectral imaging (HSI) is another emerging technology, which provides tissue-specific spectral signatures, allowing for perfusion quantification. Both imaging modalities were employed to estimate perfusion in a porcine model of partial pancreatic ischemia. Perfusion quantification was assessed using the metrics of both imaging modalities (slope of the time to reach maximum fluorescence intensity and tissue oxygen saturation (StO2), for FLER and HSI, respectively). We found that the HSI-StO2 and the FLER slope were statistically correlated using the Spearman analysis (R = 0.697; p = 0.013). Local capillary lactate values were statistically correlated to the HSI-StO2 and to the FLER slope (R = −0.88; p < 0.001 and R = −0.608; p = 0.0074). HSI-based and FLER-based lactate prediction models had statistically similar predictive abilities (p = 0.112). Both modalities are promising to assess real-time pancreatic perfusion. Clinical translation in human pancreatic surgery is currently underway.

Highlights

Given the technical complexity of pancreatic surgery, the complication rate following pancreatectomy is high [1] and this may affect the longterm oncological prognosis negatively [2]
Augmented reality (AR) of the overlaid Hyperspectral imaging (HSI) (Figure 2c) and fluorescencebased enhanced reality (FLER) (Figure 2d) images wasthree possible cases, and the the virtual demarcation between vital perform, and ischemic the ROIsininallreal-time onto pancreas and toline chronologically a confocal zones was visually consistent in both imaging modalities
augmented reality (AR) allowed for accurate confocal laser endoscopy (CLE) scanning, local lactate (LCL) samcorrespondence of each ROI

Summary

Introduction

Given the technical complexity of pancreatic surgery, the complication rate following pancreatectomy is high (approximately 50% of patients) [1] and this may affect the longterm oncological prognosis negatively [2]. As already pointed out by Strasberg et al, a satisfactory blood flow at the pancreatic stump is necessary to promote proper healing and reduce the risk of PF [6]. Clinical parameters, such as tissue color or the presence of a local pulse, are unreliable to assess the viability of visceral organs intraoperatively [7]. Fluorescence angiography (FA), an intraoperative imaging modality based on the use of near-infrared cameras and on the injection of an exogenous fluorophore (e.g., indocyanine green (ICG)) has shown promising results as a tool to assess the viability of the gastrointestinal tract. The perfusion evaluation is based on the principle that ICG propagates in vital regions but not in ischemic ones

Objectives

Methods

Results

Discussion

Conclusion