医学
前列腺
核医学
放射治疗
前列腺癌
超声波
边距(机器学习)
放射科
癌症
计算机科学
机器学习
内科学
作者
Nur Syazana Mohd Zahir,Marniza Saad,Adlinda Alip,Munira Rejab,Zulaikha Jamalludin,Nur Diyana Afrina Hizam,Yih Miin Liew,Ngie Min Ung
标识
DOI:10.1007/s13246-023-01230-x
摘要
Transperineal ultrasound (TPUS) is an image-guided radiotherapy system used for tracking intrafraction prostate displacements in real time. The objectives of this study are to evaluate intrafraction prostate displacements and derive planning target volume (PTV) margins for prostate radiotherapy at our institution. The ultrasound (US) data of nine prostate cancer patients referred for VMAT radiotherapy was retrieved. Prior to beam on, patient position was set up with the US probe positioned transperineally with the aid of reference images (fused US and computed tomography images). In each fraction, prostate displacements in three directions [superior/inferior (SI), left/right (LR) and anterior/posterior (AP)] were recorded. PTV margins were determined using Van Herk’s formula. To assess the prostate displacement time trend, continuous displacement data were plotted in 30-s intervals for eight minutes. The intrafraction prostate monitoring found a population mean setup error (Mp) of 0.8, 0.1, − 1.7 mm, a systematic error of (∑p) 0.7, 0.4, 0.9 mm and random error (σp) of 0.2, 0.1, 0.3 mm in SI, LR and AP directions, respectively. The PTV margin was found to be the largest in the AP direction at 2.5 mm compared with 1.9 mm and 1.1 mm for SI and LR directions, respectively. The PTV margin allowed for prostate radiotherapy at our institution was 2.5 mm in all directions. The prostate displacement time trend showed an increase in intrafraction displacements, with most patients were observed to have strong positive correlation between time and intrafraction prostate displacements in SI direction. TPUS is feasible for monitoring intrafraction displacement of the prostate and may facilitate PTV margin generation to account for such displacements during radiotherapy.
科研通智能强力驱动
Strongly Powered by AbleSci AI