Research Articles (Medical Physics)

Permanent URI for this collection

http://hdl.handle.net/11660/290

Browse

Browsing Research Articles (Medical Physics) by Issue Date

Filter results by year or month
Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Accuracy of patient-specific I-131 dosimetry using hybrid whole-body planar-SPECT/CT I-123 and I-131 imaging
    (SpringerOpen, 2024) Morphis, Michaella; van Staden, Johan A.; du Raan, Hanlie; Ljungberg, Michael; Sjögreen Gleisner, Katarina
    𝗣𝘂𝗿𝗽𝗼𝘀𝗲: This study aimed to assess the accuracy of patient-specific absorbed dose calculations for tumours and organs at risk in radiopharmaceutical therapy planning, utilizing hybrid planar-SPECT/CT imaging. 𝗠𝗲𝘁𝗵𝗼𝗱𝘀: Three Monte Carlo (MC) simulated digital patient phantoms were created, with time-activity data for mIBG labelled to I-123 (LEHR and ME collimators) and I-131 (HE collimator). The study assessed the accuracy of the mean absorbed doses for I-131-mIBG therapy treatment planning. Multiple planar whole-body (WB) images were simulated (between 1 to 72 h post-injection (p.i)). The geometric-mean image of the anterior and posterior WB images was calculated, with scatter and attenuation corrections applied. Time-activity curves were created for regions of interest over the liver and two tumours (diameters: 3.0 cm and 5.0 cm) in the WB images. A corresponding SPECT study was simulated at 24 h p.i and reconstructed using the OS-EM algorithm, incorporating scatter, attenuation, collimator-detector response, septal scatter and penetration corrections. MC voxel-based absorbed dose rate calculations used two image sets, (i) the activity distribution represented by the SPECT images and (ii) the activity distribution from the SPECT images distributed uniformly within the volume of interest. Mean absorbed doses were calculated considering photon and charged particle emissions, and beta emissions only. True absorbed doses were calculated by MC voxel-based dosimetry of the known activity distributions for reference. 𝗥𝗲𝘀𝘂𝗹𝘁𝘀: Considering photon and charged particle emissions, mean absorbed dose accuracies across all three radionuclide-collimator combinations of 3.8 ± 5.5% and 0.1 ± 0.9% (liver), 5.2 ± 10.0% and 4.3 ± 1.7% (3.0 cm tumour) and 15.0 ± 5.8% and 2.6 ± 0.6% (5.0 cm tumour) were obtained for image set (i) and (ii) respectively. Considering charged particle emissions, accuracies of 2.7 ± 4.1% and 5.7 ± 0.7% (liver), 3.2 ± 10.2% and 9.1 ± 1.7% (3.0 cm tumour) and 13.6 ± 5.7% and 7.0 ± 0.6% (5.0 cm tumour) were obtained for image set (i) and (ii) respectively. 𝗖𝗼𝗻𝗰𝗹𝘂𝘀𝗶𝗼𝗻: The hybrid WB planar-SPECT/CT method proved accurate for I-131-mIBG dosimetry, suggesting its potential for personalized treatment planning
  • Thumbnail Image
    ItemOpen Access
    Evaluating the accuracy of planar gated blood pool processing software using simulated patient studies
    (Elsevier, 2024) Pieters, H.; van Staden, J. A.; du Raan, H.; Nel, M. G.; Engelbrecht, G. H. J.
    Planar gated blood pool (GBP-P) radionuclide imaging is a valuable non-invasive technique for assessing left ventricular ejection fraction (LVEF). Serial cardiac imaging can be performed to monitor the potential decline in LVEF among patients undergoing cardiotoxic chemotherapy. Consequently, accurate LVEF determination becomes paramount. While commercial software programs have enhanced the LVEF values' reproducibility, concerns remain regarding their accuracy. This study aimed to generate a database of GBP-P studies with known LVEF values using Monte Carlo simulations and to assess LVEF values' accuracy using four commercial software programs. We utilised anthropomorphic 4D-XCAT models to generate 64 clinically realistic GBP-P studies with Monte Carlo simulations. Four commercial software programs (Alfanuclear, Siemens, General Electric Xeleris, and Mediso Tera-Tomo) were used to process these simulated studies. The accuracy and reproducibility of the LVEF values determined with these software programs and the intra- and inter-observer reproducibility of the LVEF values were assessed. Our study revealed a strong correlation between LVEF values calculated by the software programs and the true LVEF values derived from the 4D-XCAT models. However, all the software programs slightly underestimated LVEF at lower LVEF values. Intra- and inter-observer reliability for LVEF measurements was excellent. Accurate LVEF assessment is crucial for determining the patient's cardiac function before initiating and during chemotherapy treatment. The observed underestimation, particularly at lower LVEF values, emphasises the need for the accurate and reproducible determination of these values to avoid excluding suitable candidates for chemotherapy. The software programs' excellent intra- and inter-observer reliability highlights their potential to reduce subjectivity when using the semi-automatic processing option. This study confirms the accuracy and reliability of these commercial software programs in determining LVEF values from simulated GBP-P studies. Future research should investigate strategies to mitigate the underestimation biases and extend findings to diverse patient populations.