Session Item

In the field of pelvic MRI-only treatment the use of coil bridges for the anterior MRI receiver coil is recommended to avoid deformation of the external contour. The increased distance between patient and coil will decrease the SNR. Development in coil technology has enabled lightweight, blanket-like coils to be commercially available (air coils, GE healthcare). This work aimed to evaluate the benefits and effects of air coil technology in a pelvic MRI-only workflow, with and without coil bridges. Anatomical and dosimetric quality measures were assessed.

Six patients referred to HYPO (42.7 Gy in 7 fractions) prostate MRI-only radiotherapy were included in the analysis. One large field of view T2 weighted sequence (T2 _cli), for synthetic CT (sCT) generation and consequently treatment planning, was acquired for each patient using coil bridges according to clinical routine. The coil bridge was then removed, and the air coil was placed directly on the patient. A second identical T2 was added to the examination (T2_ncb). A 3T GE scanner, large anterior air coil and a posterior built-in spine coil were used for all acquisitions.

Anatomical difference between the T2_cli and T2_ncb was assessed to evaluate potential deformation from the on-patient air coil placement by measuring the absolute maximum anterior-posterior thickness (AMT) of the patient in the same left-right position on a slice in the middle of the prostate.

The T2_cli and T2_ncb were used to generate sCTs (sCT_cli and sCT_ncb). The clinically approved treatment plan, created on sCT_cli, was transferred and recalculated on the 6D rigidly registered sCT_ncb, using the same number of monitor units. The clinical dose metrics for the mean PTV dose, rectum V_96%, V_89% and V_77%, and mean bladder dose were analysed using the clinical structure set. A paired Wilcoxon signed-rank test was used for dose metric comparisons, p < 0.05 was considered statistically significant.

The SNR using coil bridges was measured in a homogenous phantom where the air coil and bridge was placed at a 5, 10, 15 and 20 cm distance from the surface of the phantom. 

Median AMT difference between T2_cli and T2_ncb was 1.9 mm, range [0.5 to 8.9 mm]. sCTs for the worst case (8.9 mm) is presented in Figure 1.

For the PTV no significant difference was seen (p = 0.16) and the median mean dose difference was -0.03 Gy (range -0.20 to 0.03 Gy). No significant differences were found for either rectal or bladder dose metrics (Table 1).

In phantom measurements the SNR decreased with 18, 30, 40 and 44% for increased distances of 5, 10, 15 and 20 cm, respectively. 

These preliminary results show small anatomical and no significant dosimetrical differences, when air coils are placed directly on the patient in a pelvic MRI-only workflow. This supports the removal of coil bridges from the clinical workflow. Additional subjects must be included to cover a wider range of anatomical variations.