Investigation on the optical scan condition for imaging of multi-slice spiral CT liver perfusion in rats

Multi-slice CT liver perfusion has been widely used in experimental studies of hemodynamic changes in liver lesions, and is usually performed as an adjunct to a conventional CT examination because of its high temporal and spatial resolution, simple protocol, good reproducibility, and ability to measure hemodynamic changes of liver tissues at the capillary level. Experimental rat models, especially those of induced liver cancer, are often used in studies of hemodynamic changes in liver cancer. Carcinogenesis in rats has a similar pathological progression and characteristics resembling those in human liver cancer; as a result, rat models are often used as ideal animal models in the study of human liver cancer. However, liver perfusion imaging in rats is difficult to perform, because rats' livers are so small that different concentrations, flow rates, and dose of contrast agents during the CT perfusion scanning can influence the quality of liver perfusion images in rats. The purpose of this study, therefore, was to investigate the optimal scan protocol for the imaging of hepatic perfusion using a deconvolution mathematical method in rats by comparing the results of rats in different injection conditions of the contrast agent, including concentration, rate and time.

Plain CT scan conditions in eighty 2-month-old male Wistar rats were 5.0 mm slice thickness, 5.0 mm interval, 1.0 pitch, 120 kV tube voltage, 60 mA tube current, 512 × 512 matrix, and FOV 9.6 cm. Perfusion scanning was carried out with different concentrations of diatrizoate (19%, 38%, 57%, and 76%), different injection rates (0.3 and 0.5 ml/s), and different injection times (1, 2-3, 4-5, and 6 seconds). The above conditions were randomly matched and adjusted to determine the best perfusion scan protocol. Three-phase contrast-enhanced scanning was performed after CT perfusion. Histological examination of the liver tissues with hematoxylin and eosin stains was done after CT scanning.

When the concentration of the contrast agent was 19% or 38%, no pseudo-color map was created. The viscosity increased when the concentration of the contrast agent was 76%; so it is difficult to inject the contrast agent at such a high concentration. Also no pseudo-color map was generated when the injection time was short (1, 2-3, and 4-5 seconds) or the injection rate was low (0.3 ml/s). The best perfusion images and perfusion parameters were obtained during 50 seconds scanning. Each rat was given an injection of 57% diatrizoate at 0.5 ml/s via the tail vein using a high-pressure syringe for 6 seconds. The perfusion parameters included hepatic blood flow (HBF), hepatic blood volume (HBV), mean transit time (MTT) of the contrast agent, capillary permeability-surface area product (PS), hepatic arterial index (HAI), hepatic artery perfusion (HAP), and hepatic portal perfusion (HPP). All these parameters reflected the perfusion status of liver parenchyma in normal rats. Three phases of enhancement were modified according to the time-density curves (TDCs) of the perfusion imaging: hepatic arterial phase (7 seconds), hepatic portal venous phase (15 seconds), and a delayed phase (23-31 seconds). On examination by microscopy, the liver tissues were pathologically normal.

The appropriate protocol with multi-slice spiral CT liver perfusion reflected normal liver hemodynamics in rats. This study laid a solid foundation for further investigation of the physiological characteristics of liver cancer in a rat model, and was an important supplement to and reference for conventional contrast-enhanced CT scans.