ESTRO 2024 - Abstract Book

S3875

Physics - Image acquisition and processing

ESTRO 2024

comparing pre-treatment and fraction 5 R2* maps (Figure 2). When assessing changes between each imaging session, only weak evidence of inter-fraction differences (Table 1) was obtained (p<0.124 using a single-factor analysis of variance (ANOVA) test). Using volunteer-measured repeatability, four out of five patients experienced median R2* value changes above the %RC. However, the median R2* value changes were only above the patient-measured %RC of 13.0% in one out of five patients. R2* distributions were positively skewed, indicating that the median rather than the mean was the appropriate summary statistic to report.

Figure 2: Median R2* values of the whole prostate before (pre-treatment imaging session) and after treatment (immediately following fraction 5).

Table 1: Median R2* values of the whole prostate throughout five patients’ treatments.

Median R2* (s -1 )

Patient

Pre-treatment

Fraction 1

Fraction 3

Fraction 5

1

20.2

19.7

18.2

17.7

2

20.2

17.9

14.0

16.1

3

15.0

15.7

12.2

13.1

4

19.3

20.5

18.3

18.1

5

18.6

18.1

16.2

16.3

Conclusion:

A BOLD-MRI sequence was successfully developed for use on the MR Linac with its suitability robustly assessed both quantitatively and qualitatively. The sequence is capable of measuring R2* changes in the prostate but further work is required to improve the repeatability in patient groups and to focus the analysis on localised regions of the prostate