Quantitative serosal and mucosal optical imaging perfusion assessment in gastric conduits for esophageal surgery: an experimental study in enhanced reality

Abstract

Introduction/objective Gastric conduit (GC) is used for reconstruction after esophagectomy. Anastomotic leakage (AL) incidence remains high, given the extensive disruption of the gastric circulation. Currently, there is no reliable method to intraoperatively quantify gastric perfusion. Hyperspectral imaging (HSI) has shown its potential to quantify serosal StO2. Confocal laser endomicroscopy (CLE) allows for automatic mucosal microcirculation quantification as functional capillary density area (FCD-A). The aim of this study was to quantify serosal and mucosal GC’s microperfusion using HSI and CLE. Local capillary lactate (LCL) served as biomarker. Methods GC was formed in 5 pigs and serosal StO2% was quantified at 3 regions of interest (ROI) using HSI: fundus (ROI-F), greater curvature (ROI-C), and pylorus (ROI-P). After intravenous injection of sodium-fluorescein (0.5 g), CLE-based mucosal microperfusion was assessed at the corresponding ROIs, and LCLs were quantified via a lactate analyzer. Results StO2 and FCD-A at ROI-F (41 ± 10.6%, 3.3 ± 3.8, respectively) were significantly lower than ROI-C (68.2 ± 6.7%, p value: 0.005; 18.4 ± 7, p value: 0.01, respectively) and ROI-P (72 ± 10.4%, p value: 0.005; 15.7 ± 3.2 p value: 0.001). LCL value at ROI-F (9.6 ± 4.7 mmol/L) was significantly higher than at ROI-C (2.6 ± 1.2 mmol/L, p value: 0.04) and ROI-P (2.6 ± 1.3 mmol/L, p value: 0.04). No statistically significant difference was found in all metrics between ROI-C and ROI-P. StO2 correlated with FCD-A (Pearson’s r = 0.67). The LCL correlated negatively with both FCD-A (Spearman’s r = − 0.74) and StO2 (Spearman’s r = − 0.54). Conclusions GC formation causes a drop in serosal and mucosal fundic perfusion. HSI and CLE correlate well and might become useful intraoperative tools.

Optimization of anastomotic technique and gastric conduit perfusion with hyperspectral imaging in an experimental model for minimally invasive esophagectomy

“…In values between 60% and 70% as described in other studies 46 to below 45% during magnet-induced ischemia resulted in a pre-necrotic demarcation of gastric tissue of 35.7±9.7% of histological surface area. This tendency also applied incrementally for the stages in between, so that with lower StO2 values between 45% and 60%, there were correspondingly greater surface areas between 10% and 35% with pre-necrotic changes.…”

Deep Learning Analysis of In Vivo Hyperspectral Images for Automated Intraoperative Nerve Detection

“…Likewise, HSI could be easily integrated into current surgical optical navigation tools, such as laparoscopes [12], endoscopes [13] or even robotic systems; hence, it has captured the attention of the surgical community as a potential innovative surgical guidance tool [14]. In particular, HSI technology has been previously applied in digestive surgery to quantify intestinal perfusion before anastomosis during several procedures [15][16][17], as well as in the case of mesenteric ischemia [18,19], or to quantify liver perfusion [20]. A number of previous works focused successfully on the ability of HSI to discriminate between normal and tumoral tissue, particularly in prostate cancer [21], colorectal cancer [22][23][24], gastric cancer [25,26], glioblastoma [27] and head and neck cancers [28][29][30][31].…”

Single Snapshot Imaging of Optical Properties (SSOP) for Perfusion Assessment during Gastric Conduit Creation for Esophagectomy: An Experimental Study on Pigs

“…Our group has previously used HSI and a system called HYPER (HYperspectral-based Enhanced Reality) to assess StO 2 % and intraoperatively localize preselected ROIs during esophagectomy [ 43 ], small bowel ischemia [ 24 ], and hepatectomy [ 44 ]. HYPER allowed to compare HSI to LCLs and mucosal scan with confocal laser endomicroscopy (CLE) on the same ROI [ 43 ] with high accuracy. However, the time currently necessary for HSI acquisition is around 10 s and still not yet fully “real-time”.…”