Methods used to develop CT scan protocols

Methods used to develop the CT scan contrast agent injection protocols

The development of these CT scanning protocols took place gradually over a 4 year period from 1995 to 1999. During this time, the scanning protocols evolved through numerous changes. The current protocols are the "I series", and they were preceded by series A,B,C,D,E,F,G and H.

The A protocols were the starting point. They divided patients into subjective small, medium and large groups.
The B protocols used specific weight ranges for the small, medium and large groups.
The C protocols divided the weight categories into 5 groups, and introduced different injection rates for each group. The transition from A to B to C protocols occurred quickly, during the first two months.

The D protocols added minor improvements. These protocols were studied in a "qualitative" study, comparing subjective scan quality of the D series CT images compared to CT images from fixed doses of 100ml and 120ml. The results showed overall better quality images from the D series compared to fixed doses.

The E, F, G and H protocols were studied "quantitatively". The types of data collected were:

Liver "enhancement" is the post-contrast CT density, in Hounsfield units, measured on a mid-liver slice, by tracing a Region-Of-Interest (ROI) on the liver parenchyma. (The computer software automatically calculated the mean density of all pixels within the ROI). The ROI avoided major intra-hepatic blood vessels and biliary ducts. The studies did not include pre-contrast scans, so it was not possible to subtract pre-contrast from post-contrast images.
Intra-hepatic portal venous enhancement was scored on the following scale: 0= hypodense portal veins, 1=portal veins isodense with liver parenchyma, 2= portal veins mildly brighter than liver parenchyma. 3=portal veins moderately brighter than liver parenchyma, 4=portal veins were so bright as to appear completely white.
Hepatic venous enhancement was scored on a similar scale: 0=hypodense hepatic veins, 1=hepatic veins isodense with liver parenchyma, 2= hepatic veins mildly brighter than liver parenchyma, 3=hepatic veins moderately brighter than liver parenchyma, 4=hepatic veins were so bright as to appear competely white.
Kidney parenchymal enhancement was scored on the following scale: 0=thin cortex enhancement only, 1=mix of cortical and partial medulla enhancement, 2=equal cortex and medulla enhancement, 3= medulla has hyperdense streaks from excreted contrast in the pyramids, 4=opacified urine is visible in the renal pelvis.

The E and F protocols used a 60ml dose for the weight group 38-47 kilograms. This was changed to 70ml in the G and H protocols. ( The 60ml dose in this weight group resulted in a mean liver enhancement of 106.0 Hounsfield units, and the 70ml dose in this weight group resulted in a mean liver enhancement of 117.4 Hounsfield units, which is closer to the means of the other weight groups.)

The G protocol added a new weight category for patients weighing > 117 kilograms and changed their dose to 160ml from 140ml. (However, there was hardly any difference: the mean liver enhancement changed from 101.6 to 103.8 Hounsfield units.)

The G protocols shifted the scan Delays earlier, in order to accumulate data from earlier delay times, especially in lighter patients.

The H protocols were developed when scan slice thickness changed from 10mm to 8mm thick. All the delays were adjusted.

The Current I protocols were designed after analysing all data from previous series. The Delays were adjusted slightly to their optimum values.

Below, is a record of the parameters used in the E,F,G & H protocols. The Rates haven't changed, and are not shown.

	E Delays Abd	E Delays Chest	E Chest Abd		F Delays Abd	F Delays Chest	F Chest Abd		G Delays Abd	G Delays Chest	G Chest Abd		H Delays Abd	H Delays Chest	H Chest Abd
								160ml	66	53	41	160ml	66	57	38
140ml	74	57	29	140ml	69	57	29	140ml	66	53	41	140ml	66	56	38
120ml	71	56	27	120ml	68	56	27	120ml	65	52	40	120ml	65	55	37
100ml	69	55	26	100ml	67	55	26	100ml	64	51	39	100ml	62	53	35
80ml	66	54	24	80ml	66	54	24	80ml	63	51	38	80ml	60	51	34
60ml	62	54	22	60ml	64	54	22	70ml	63	51	38	70ml	59	51	33
50ml	58	53	20	50ml	60	53	20	60ml	62	50	38	60ml	58	50	32

There are some pit-falls in the methods, but they were not bad.

back to CT scan weight based protocol index