The study was approved by the Ethics Commission for Animal and Human Experimentation of the Autonomous University of Barcelona, to which our hospital is affiliated (Ref. CEEAH 10; 26-2016)
Experimental, controlled, prospective study in animals. After laparoscopic surgery and ICG administration, the vascularization of the colon was quantitatively determined using instantaneous images with the SERGREEN program. All methods were carried out in accordance with relevant guidelines and regulations. This study followed the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines.
Eight farm pigs (Large-White breed: four males and four females, weight 30–35 kg and age 2–4 months) were used. They were imported from SPECIFIC PIGS Ltd. (CIF-B65488017. Autovia de Castelldefels Km. 31, El prat de Llobregat, 08820, (Barcelona), SPAIN) with all the required health and transport documentation. The use of all pigs were allowed by informed consent from SPECIFIC PIGS Ltd.
Animals were operated upon at the Experimental Surgery Unit of the Parc Taulí University Hospital by the hospital’s colorectal surgery team, with support from the veterinary team. The surgery was performed under general anesthesia in accordance with the center’s animal anesthesia protocol.
Sedation was performed by intramuscular injection of zolethyl 100 (2 mg/kg tiletamine + 2 mg/kg zolazepam), xylazine (2 mg/kg) and atropine (0.01 mg/kg). The atrial vein was channeled, and propofol was injected in a bolus (1.5 mg/kg). Once the animals were sedated and relaxed, orotracheal intubation was performed. Sedation and analgesia were maintained by a continuous infusion pump using propofol and fentanyl, and fluid therapy was maintained with Ringer’s lactate solution.
Animals’ vital signs were recorded at baseline, after anesthesia, analgesia and relaxation, and prior to surgery. They were maintained stable throughout the procedure, with heart rates ranging between 70–95 beats per minute and mean blood pressure between 60 and 80.
The laparoscopic equipment used was the IMAGE 1 H3-Z FI and IMAGE 1 HUB HD Camera Control Unit SCB (Karl-Storz®). Pictures were taken in STANDARD, SPECTRA A and ICG mode. Images were recorded from the time just prior to the infusion until the end of the procedure.
Two groups were created for the study of the anastomoses. The control group underwent two phases of surgery, one open and one laparoscopic; in the second group, which also underwent open and laparoscopic surgery, an ischemic segment was created at the level of the left colon, while the anastomosis of the right colon was left undisturbed.
All the experimental procedures were carried out by the same members of the research team at the same ICG dose, with the same observation time and the same light source and laparoscopic optic exposure. In view of the drug’s technical specifications and elimination half-life, a 30-minute interval was observed between each determination and each new administration to ensure adequate clearance of the blood flow.
A midline laparotomy was performed, and the right colon and left colon segments of interest were located. The second loop of the right colon (spiral) and the segment of the left colon located in the center of the arcade of the inferior mesenteric artery and vein were selected. Both colon segments were sectioned and anastomosed again using a running end-to-end 3/0 silk suture.
After creation of the anastomoses, the laparotomy was closed with a running, hand-sewn 0 silk suture.
Ten-millimeter laparoscopic trocars were inserted at the level of the mesogastrium to the right of the midline laparotomy incision to allow introduction of the laparoscopic camera. A 10-mm trocar in the right quadrant and a 5-mm trocar in the left quadrant were used to introduce the laparoscopic equipment. Intra-abdominal pressure was kept constant with a 12 mmHg self-regulating insufflator.
The ICG was injected at a dose of 0.2 mg/kg through the atrial vein, and video and image recordings began. Starting two minutes after contrast infusion, standard and Spectra An images were taken of both the right and left colon anastomoses. (Fig. 1)
Thirty minutes after the administration of the first dose of ICG, the abdominal cavity was accessed again through midline laparotomy performed earlier. At the level of the anastomosis of the left colon, the arcade of the mesenteric artery and vein was divided less than 5 cm proximal and 5 cm distal from the anastomosis, creating an area of approximately 10 cm without vascularization.
After dividing the artery, the midline laparotomy was closed again using a running suture 0-silk. We waited 15 minutes to ensure complete ischemia of the segment of interest.
A 0.2 mg/kg dose of ICG was again administered through the atrial vein. Starting two minutes after contrast infusion, Standard and Spectra An images were taken of the two anastomoses (right colon and ischemic left colon).
This procedure was repeated every 30 minutes under the same conditions for up to a maximum of 120 minutes (135 minutes after the vascular section) (Fig. 2-A and 2-B).
Experimental outcomes: The SERGREEN software
The results were analyzed using the SERGREEN program, measuring the segments of interest in areas of uniform vascularization two minutes after ICG injection. The program was created using the MATLAB-2014b software tool (R2014b, 8.4.0. 150421. 64-bit -maci64- September 15, 2014).
The images obtained during the experimental procedure were recorded in * jpg format with laparoscopy equipment. This format is processed by the computer program and broken down into three matrices according to the RGB (red, green, blue) color model. Each cell in each matrix corresponds to a pixel with its respective intensity in these three colors. All pixels obtain 24 bits (eight in each matrix) with the information from the three colors.
On the decimal scale, eight bits are represented from 0 to 255, and this is how these values are
represented. By overlapping the three channels from the RGB image, we obtain the color seen in the original image. SERGREEN processes the matrix corresponding to green and compares it
with the other two to determine the predominant color.
Pixels whose predominant color is not green are discarded. In this way, false readings caused by
reflections are avoided, since white is the composition of the three colors at their maximum value (R: 255, G: 255 and B: 255).
The program’s main tool allows users to mark a rectangle wherever they like in the image and obtain statistical information on the intensity of the green in that region. All the pixels corresponding to the green channel within the specified rectangle can be exported into an Excel file. Another document is generated to check the image selected, as well as a boxplot chart, a histogram and a summary of the central tendency values.
Experimental outcomes: Image processing and analysis of the results obtained with SERGREEN
This measurement was carried out in 10 different areas with 10x10 pixel squares. A sampling of different areas of interest was generated within the same intestinal segment to ensure its correct representation. A matrix was obtained that included the eight pigs, all their intestinal segments and their 10 measurements according to area of interest and for each time of observation, including all the pixels contained in the squares.
At the level of the anastomosis, between 1–2 cm proximal and distal to the suture line, 10 areas of interest were recorded in all the images obtained of both the right and left colon, avoiding the areas that correspond to visible vessels (Fig. 3).
Statistical and calculation methods
The ICG saturation obtained by applying the SERGREEN program to the images of interest was taken as the main variable. The Kolmogorov-Smirnov test was applied to check the normal distributions.
The results are expressed as the means for each segment (the sample mean of the pigs for each segment and observation time).
Student’s t test was used for comparisons between two means, while ANOVA was used to compare different means since the distributions were comparable to normal. The decreases in the saturation of the ICG after generating the ischemic area are expressed in absolute and relative terms.
The absolute mean ICG saturation decrease (AMISD) was calculated as the mean ICG saturation of the preischemic left colon (MISPreI = Mean ICG Saturation PreIschemia) minus the mean ICG saturation of all postischemic left colon saturation determinations (MISPostI = Mean ICG Saturation PostIschemia).
We describe the relative mean ICG saturation decrease (RMISD), which is calculated as the unit minus the division of the nonvascularized area (MISPostI) with respect to the vascularized area (MISPreI) of the left colon. This can be expressed as a percentage of the relative decrease by multiplication by 100.