RATIONALE AND OBJECTIVES: The authors experimentally assess and compare the detection limit of gastrointestinal bleeding in digital subtraction angiography (DSA) and conventional screen-film angiography. MATERIALS AND METHODS: Arterial blood flow was simulated using a tube model in which bleeding was imitated by exudation of liquid containing contrast material. Gut peristalsis was imitated using silicone tubes filled with air and liquid. Images were acquired by DSA and conventional screen-film angiography. The iodine concentration was increased in increments from 1 mg I/mL, with and without simulated peristalsis, and with both free and circumscribed extravasation of contrast material. RESULTS: The detection limit for free extravasation in DSA was 1 mg I/mL without peristalsis and 60 mg I/mL with peristalsis. The corresponding figures for circumscribed extravasation were 1 mg I/mL and 34 mg I/mL. The detection limit for free extravasation in screen-film angiography was 172 mg I/mL both with and without simulated peristalsis. Circumscribed extravasation was detected at 9 mg I/mL without peristalsis and 7 mg I/mL with peristalsis. CONCLUSION: Digital subtraction angiography is the more sensitive angiographic technique for detection of gastrointestinal bleeding and is superior to conventional screen-film angiography, provided that it is performed with adequate parasym-pathicolysis and suspended respiration. Only when these requirements cannot be achieved is screen-film angiography advantageous.
RATIONALE AND OBJECTIVES: The authors experimentally assess and compare the detection limit of gastrointestinal bleeding in digital subtraction angiography (DSA) and conventional screen-film angiography. MATERIALS AND METHODS: Arterial blood flow was simulated using a tube model in which bleeding was imitated by exudation of liquid containing contrast material. Gut peristalsis was imitated using silicone tubes filled with air and liquid. Images were acquired by DSA and conventional screen-film angiography. The iodine concentration was increased in increments from 1 mg I/mL, with and without simulated peristalsis, and with both free and circumscribed extravasation of contrast material. RESULTS: The detection limit for free extravasation in DSA was 1 mg I/mL without peristalsis and 60 mg I/mL with peristalsis. The corresponding figures for circumscribed extravasation were 1 mg I/mL and 34 mg I/mL. The detection limit for free extravasation in screen-film angiography was 172 mg I/mL both with and without simulated peristalsis. Circumscribed extravasation was detected at 9 mg I/mL without peristalsis and 7 mg I/mL with peristalsis. CONCLUSION: Digital subtraction angiography is the more sensitive angiographic technique for detection of gastrointestinal bleeding and is superior to conventional screen-film angiography, provided that it is performed with adequate parasym-pathicolysis and suspended respiration. Only when these requirements cannot be achieved is screen-film angiography advantageous.