PURPOSE: The aim of this study was to introduce thermal imaging in the intr
aoperative detection of bowel ischemia by comparing thermal imaging with co
nventional techniques in detecting acutely ischemic bowel, using histologic
evidence for intestinal necrosis as the standard. METHODS: A prospective s
tudy was performed using a porcine model. Laparotomy was performed on four
pigs under general anesthesia. A 25-cm segment of mid jejunum was tagged wi
th proximal and distal sutures, and its mesentery was ligated and divided.
Thermal imaging, visual inspection, Doppler ultrasound, and fluorescence wi
th Wood's lamp after fluorescein were used to estimate the extent of bowel
ischemia five minutes after ligation of the mesentery. Measurements were ta
ken in reference to both the proximal and distal tags to obtain two data po
ints per animal for each method. After two hours of warm ischemia, the jeju
num was harvested and sectioned longitudinally. Comparisons were made betwe
en the estimated region of necrosis for each method and microscopic evidenc
e of necrosis. RESULTS: Visual inspection was the only method unable to det
ect a difference between vascularized and devascularized bowel for each of
the eight data points. Fluorescein dye missed 3 cm of ischemic bowel. Doppl
er ultrasound and thermal imaging were 100 percent sensitive for necrotic b
owel, with thermal imaging overestimating necrosis to a greater extent than
Doppler ultrasound. The positive predictive value of fluorescein dye, Dopp
ler ultrasound, and thermal imaging for determining nonviable bowel was 91.
8, 80.8, and 69.5 percent, respectively. CONCLUSIONS: Thermal imaging has t
he potential to be a useful adjunct in the intraoperative determination of
bowel ischemia. Further studies are indicated to study this technique.