In Vivo Dosimetry using Plastic Scintillation Detectors for External Beam Radiation Therapy

 
5.5 Conclusion 
Because the PSD has been demonstrated to be a good in vivo detector in prior studies and 
because we have shown here that PSDs can be used for entrance dosimetry for proton 
fields without undue difficulty, we believe that PSDs are a good candidate for in vivo 
entrance dosimetry in patients undergoing proton therapy. PSDs are small and water 
equivalent, so the beam should not be perturbed in a clinically significant manner, and 
skin dose results would be available immediately following treatment. This may prove 
helpful in assessing the risk of radiation dermatitis for patients undergoing treatment. 
PSDs for in vivo entrance dosimetry would also be capable of detecting gross errors 
during dose delivery. Unfortunately, other treatment parameters such as beam range 
could not be verified with such a system. Nonetheless, we believe that PSDs could serve 
as a useful tool to perform in vivo entrance dosimetry. 
 
 
100 

---

CHAPTER 6 
 
 
 CONCLUSION 
101 

---

6.1 Summary and Conclusions 
The work presented in this dissertation represents a significant step towards 
implementing plastic scintillation detectors for general use as in vivo dosimetry devices in 
external beam radiation therapy. Three specific projects were undertaken as steps towards 
this goal: characterization and correction of the temperature dependence of common 
scintillating fibers, a clinical protocol employing the PSD as an in vivo dosimeter for 
patients undergoing prostate radiation therapy, and a characterization of the performance 
of PSDs in the context of proton skin dosimetry. 
The first study demonstrated that, in contrast with prior knowledge, two 
scintillating fibers commonly used in PSDs exhibit temperature dependence. The 
temperature dependence was found to be approximately linear with temperature. The 
first, BCF-60, was found to lose 0.5% of its signal with each °C increase in temperature, 
relative to 22°C. BCF-12 was found to lose 0.09% with each °C increase. This 
corresponds to a 7.5% and 1.4% under-response at body temperature, respectively. The 
shape of the scintillation spectrum was found to change slightly for each scintillating 
fiber. This effect was small however, and it was suggested that a simple correction factor 
consisting of the ratio of dose measured at a given temperature to that at a reference 
temperature was sufficient to account for this effect. 
The second study utilized PSDs to perform in vivo dosimetry for patients 
undergoing IMRT for prostate cancer. Pairs of PSDs were attached to endorectal balloons 
which were then inserted into patient’s rectums for the duration of each fraction of 
treatment. This positioned the PSD pair in close proximity with the rectal wall where they 
were used to measure the dose delivered during treatment. Temperature correction factors 
102 

---

derived from the previous study were used to ensure accurate results. Prior to each 
treatment, a CT image set was acquired for the purpose of locating and determining the 
expected dose to each detector. This procedure was repeated twice weekly for five 
patients, generating a total of 142 in vivo measurements. 
The average difference between the expected dose and the measured dose ranged 
from -3.3% to 3.3% over the five patient population. The standard deviation fell between 
5.6% and 7.1% for four of the five patients, and was 13.9% for the fifth patient for 
reasons explained in detail in chapter 4. The average difference over all five patients was 
-0.4% with a standard deviation of 2.8%. The implementation of an in vivo dosimetry 
system did not interrupt or alter the clinical workflow, and the patients reported that the 
detectors attached to endorectal balloons were as tolerable as the endorectal balloon 
alone. 
During the course of this study, a method of localizing the detector using three 
ceramic fiducials attached in a rigid geometry was implemented. This was necessary 
because PSDs are radiographically indistinguishable from tissue as a result of being water 
equivalent. When experimentally validated in an anthropomorphic phantom, this method 
localized detectors to within 1 mm in the lateral and anterior-posterior directions, 
exhibiting an average deviation of just 0.1 mm from the true location. 
The final study investigated the use of PSDs for entrance dosimetry in proton 
beams. Of particular interest was the problem of ionization quenching, an under-response 
when measuring dose delivered by high LET radiation. Comparisons between ion 
chamber measurements and PSD measurements revealed that PSDs under-respond by 7% 
to 10% at the entrance of passively scattered proton beams of energies between 140 MeV 
103 

---

and 250 MeV, with lower energy beams producing a greater under-response. The width 
of the spread out Bragg peak was found to have a negligible effect on the magnitude of 
the under-response. In spite of the reduced signal due to ionization quenching, the PSD 
was found to exhibit excellent relative accuracy and a high SNR. On the basis of this 
work it is expected that the PSD can be used effectively as an in vivo skin dosimeter in 
proton therapy with the use of empirically determined ionization quenching correction 
factors or direct calibration in the proton beam of interest. 
Overall, this work has demonstrated that two non-negligible response-altering 
effects can be accurately corrected for, permitting high accuracy in vivo dosimetry. 
Furthermore, it has been demonstrated that PSDs are effective and practical when used 
for in vivo dosimetry, producing accurate results even when placed in a high dose-
gradient region such as the rectal wall in prostate intensity modulated radiation therapy.