Contrast medium injection optimisation in spiral CT for the diagnosis of pulmonary embolism.

Purpose: Spiral CT, normally a highly accurate diagnostic method to diagnose pulmonary embolism, has its weak point in the synchronisation of contrast medium (CM) injection and the start of the acquisition, essential to obtain optimal vascular enhancement. The aim of this paper is to introduce a method to control the CM injection based on the enhancement of blood vessels in the diagnosis of pulmonary embolism.
Materials and Methods: The CARE bolus software pilots an electronic trigger that first monitors the CM passage, then starts the acquisition procedure when the intensity of enhancement reaches a pre-set value. Our spiral CT has a 6-second scan delay between the trigger's "go-ahead" and the start of the acquisition. During this interval, the CM reaches the pulmonary venous system, enhancing it and making the diagnosis of pulmonary embolism more difficult. This problem was overcome by injecting a slow bolus (30 ml; 1.5 ml/s flow rate) before the CM that triggers the start of the scan when the CM is only present in the pulmonary arteries. We examined 80 patients (36 men, 44 women, mean age 66.9, age range 18 to 89 years). All patients were examined for clinically, radiographically or scintigraphically suspected pulmonary embolism. We evaluated the enhancement of pulmonary arteries on a scale from 0 (poor) to 10 (excellent), image quality (excellent, fair, poor), the examination time and patient tolerance. The results were compared with those obtained in a group of 80 patients studied with CARE bolus without a timing bolus.
Results: Monitor scans were performed with the ROI that triggers the sequence centred on the right heart (trigger value set at 30/35 HU). There were no diagnostic artefacts caused by the enhancement of pulmonary veins due the timing bolus. The average time per procedure was less than 30 min and the time needed to reach the trigger value was 15 sec (range: 10-24 sec). The average volume of CM injected was 130 ml (timing bolus: 30 ml, scan bolus: 100 ml). There were no adverse events to CM injection. The arterial enhancement scored 7 to 10 for 45 patients (56%), 4 to 6 for 23 (28.5%) and 1 to 3 for 12 (15.0%). Image quality was excellent in 52 patients (65.0%), fair in 18 (22.5%) and poor in 10 (12.5%). Comparing arterial enhancement and image quality, we observed that in the cases where enhancement had scored 7 to 10, image quality was excellent, and in the 7 cases where image quality was poor, so was enhancement (1 to 3). We also made a comparison between procedures carried out with and without the timing bolus. We observed that the use of the timing bolus increases the number of exams with high scores for arterial enhancement, and therefore increases the overall number of examinations with optimal enhancement.
Discussion and Conclusions: In order to be diagnostically useful, spiral CT requires good vascular enhancement and synchronisation of the start of acquisitions with the highest concentration of CM, as an incorrect scan delay will lead to artefacts and interpretation errors. The proposed method allows correct timing of the CM injection. The diagnostic bolus is preceded by a slow-flow timing bolus that is intercepted by the electronic trigger, which starts the scan when the CM passes into the right heart and pulmonary arteries. The slow-flow bolus volume was 30 ml injected at 1.5 ml/s, whereas the volume of the real bolus was 100 ml, injected at 4.5 ml/s. Monitor scans were performed with the trigger ROI centred on the right heart (trigger value set at 30/35 HU). The time needed for the complete spiral CT exam did not exceed 30 min. The first low-flow bolus injection takes approximately 10 min, but this time becomes shorter as the operator's experience grows. The correct positioning of the ROI on the right heart is the most time-consuming step in the procedure. The procedure was well accepted by all patients with no complaints due to the CM or to the duration of the procedure. There is a high level of concordance between arterial enhancement and image quality. In conclusion, the proposed method is simple, easy to reproduce, and does not give rise to interpretation problems. It is well accepted by patients and suffers few limitations, mainly represented by patients with severe cardiac arrhythmia.