Clinical implementation of artificial intelligence-driven cone-beam computed tomography-guided online adaptive radiotherapy in the pelvic region

Abstract

Background and purpose Studies have demonstrated the potential of online adaptive radiotherapy (oART). However, routine use has been limited due to resource demanding solutions. This study reports on experiences with oART in the pelvic region using a novel cone-beam computed tomography (CBCT)-based, artificial intelligence (AI)-driven solution. Material and methods Automated pre-treatment planning for thirty-nine pelvic cases (bladder, rectum, anal, and prostate), and one hundred oART simulations were conducted in a pre-clinical release of Ethos (Varian Medical Systems, Palo Alto, CA). Plan quality, AI-segmentation accuracy, oART feasibility and an integrated calculation-based quality assurance solution were evaluated. Experiences from the first five clinical oART patients (three bladder, one rectum and one sarcoma) are reported. Results Auto-generated pre-treatment plans demonstrated similar planning target volume (PTV) coverage and organs at risk doses, compared to institution reference. More than 75% of AI-segmentations during simulated oART required none or minor editing and the adapted plan was superior in 88% of cases. Limitations in AI-segmentation correlated to cases where AI model training was lacking. The five first treated patients complied well with the median adaptive procedure duration of 17.6 min (from CBCT acceptance to treatment delivery start). The treated bladder patients demonstrated a 42% median primary PTV reduction, indicating a 24%-30% reduction in V 45Gy to the bowel cavity, compared to non-ART. Conclusions A novel commercial oART solution was demonstrated feasible for various pelvic sites. Clinically acceptable AI-segmentation and auto-planning enabled adaptation within reasonable timeslots. Possibilities for reduced PTVs observed for bladder cancer indicated potential for toxicity reductions.

Development of in-house fully residual deep convolutional neural network-based segmentation software for the male pelvic CT

“…Consequently, deliverables with artificial intelligence will be needed for accuracy in "real world [18]". Currently, auto-segmentation including a CNN method was implemented in some commercial treatment planning support systems [19][20][21][22][23][24]. DLCExpert ™ (Mirada Medical Ltd., UK), Ethos therapy system (Varian Medical Systems, Palo Alto, CA), and Limbus Contour (Limbus AI Inc., Regina, Canada) have the function of auto-segmentation using modified U-nets such as semantic segmentation [25] and BibNet [27].…”

“…Furthermore, auto-segmentation has great real-world clinical potential with the possibility of reducing time consumption [18]. Despite the number of published studies in this area, it is difficult to generalize these outcomes because it can only be used with dedicated treatment planning support systems [19][20][21][22][23][24][25]. Therefore, in this study, we develop and evaluate the accuracy of a software that can be used on the commercial radiation treatment planning system (RTPS).…”

An in-silico assessment of the dosimetric benefits of MR-guided radiotherapy for esophageal cancer patients

“…(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) radiotherapy for esophageal cancer patients at dedicated X-ray based systems [25,26]. Previous studies on radiation-induced toxicity after chemoradiation for esophageal cancer showed multiple correlations between an increase in lung and heart dose and pulmonary and cardiac complications and mortality.…”

“…Recently, multiple online adaptive treatment platforms have been released by various vendors which allow for online replanning and thereby potentially by-passing all interfractional anatomy geometry variations. These platforms could either be based on X-ray imaging [25,26] or MR imaging [27][28][29] and could have onboard imaging capabilities that allow for online intrafraction monitioring to allow gating [30,31] and tracking [32] to mitigate effects from intrafraction motion.…”

Future mainstream platform for online adaptive radiotherapy will be using on‐board MR rather than on‐board (CB) CT images

“…Reasonable adaptive treatment time has been reported for the Varian Ethos platform, which allows potential high patient throughput with ART. 16 One concern over CTgART is the lowimaging quality for online delineation and the lack of soft-tissue contrast for target visualization. Both FBCT and CBCT on the current ART platforms provide comparable image quality to the planning CT, which is already sufficient for many disease sites (head and neck, thorax, pelvis, spine, et al).…”