Patients: Development of the study:
Between August 7th and September 3rd 2013,
237 patients (132 males and 105 females),
mean aged of 61.3 years (SD 1.4) and mean weight of 74.1 kg (SD 15.27) were refered for follow up of their oncology disease. Fig. 4
The amount of contrast medium was calculated by P3T system according to weight (group B).
Each abdomen CT was compared to another in the same patient and the same standard protocol performed with fixed contrast medium dose of 120 ml (group A). Fig. 5
Studies carried out using another protocol or patients who had contraindication for the use of intravenous contrast (as a high creatinine level),
were excluded.
MDCT protocol
All examinations were performed with a 64 slices CT scan (Somatom Definition and Somatom sensation,
Siemens,
Forcheim,
Germany) in the portal phase at 120 kilovoltage and with an automatic tube current modulation program Caredose 4D-Siemens.
This program selects an effective miliampere second (mAs) for a reference adult or pediatric patient.
In our adult protocol,
this value was set at 210 mAs and it was applied depending on patient-specific parameters calculated according the scout view.
Other parameters of data acquisition were: detector collimation 0.6 mm,
section acquisition 5 mm,
gantry rotation time 0.5 ms,
and pitch 1.4. Fig. 6
In both studies we used Iohexol 300 mg iodine/ml.
Examinations in group A were performed with fixed contrast dose of 120 mg at 2 ml/s,
while examinations in group B (with P3T system) used a contrast dose according to patient's weight with at mean rate of 2.2 ml/s.
This rate was calculated using a fixed time duration of the injection of 58 seconds. Fig. 7
Quantitative analysis: [3-7]
For quantitative analysis,
we measure the signal intensity,
given in Hounsfield Units (HU),
in the liver and spleen parenchyma,
infrarenal abdominal aorta and in the common portal vein,
with three circular regions of interest (ROIs) drawn in these structures. Fig. 8 Fig. 9
The mean background signal was calculated by measuring the signal intensity with three ROIs drawn in the erector spinal muscles. Fig. 10
In the same way,
the image noise was calculated by measuring the standard deviations of HU with three ROIs drawn in three different regions outside the patient body.
In these cases,
we used lung window to avoid the noise produced by external devices. Fig. 11
The signal-to-noise ratio (SNR) was defined as SI structures/ Noise.
The contrast-to-noise ratio (CNR) was defined as (SI structures - SI background)/Noise.
Qualitative analysis: [3-7]
For qualitative evaluation,
two radiologists,
with a double-blind evaluation,
with the same window and width values assessed the enhancement and the diagnostic quality of three structures (portal vein,
pancreas and kidneys).
They scored them from 1 (poor enhancement or poor quality) to 3 (excellent enhancement or excellent quality). Fig. 12 Fig. 13
The final score of every studies was categorised in three different groups: non diagnostic (3-4 points),
diagnostic with suboptimal quality (5-7 points); and diagnostic with optimal quality (8-9 points).
Statistical analysis:
Student's T test for related samples was employed to compare quantitative results.
We analyzed global results and also stratifying by weight.
A two-sided p value < 0.05 was considered to indicate a statistically significant difference.
Statistical analyses were performed with the SPSS software package (SPSS for Windows,
version 20).