Automatic post-implant needle reconstruction algorithm to characterize and improve implant robustness analyses

A postimplant dosimetry simulation framework for robustness evaluation in permanent breast seed implant brachytherapy

BACKGROUND

Permanent breast seed implant (PBSI) brachytherapy is a promising treatment that has the potential to be widely utilized with increased standardization, optimization, and robustness. Excellent early efficacy and very high patient acceptance were reported, however, to further evaluate and improve planning strategies, a framework to quantify plan robustness to implant uncertainties is necessary.

PURPOSE

In this study, we aim to quantify clinical seed displacement using an automated algorithm and develop and validate a PBSI post-implant dosimetry simulation framework to evaluate PBSI plan robustness to implant uncertainties.

METHODS AND MATERIALS

Clinical PBSI seed displacements were quantified for 63 consecutive patients. A PBSI simulator was developed in Matlab (2020) by resampling clinical seed displacements and computing a range of possible post-implant dosimetry outcomes under various seed displacement scenarios. Simulations were performed retrospectively on 63 previous clinical plans to evaluate plan robustness to seed displacement.

RESULTS

Mean seed displacement for the whole cohort was 10 ± 6 mm. A clinical seed displacement database was established and a user interface was developed for the simulation framework. For all clinical plans, the median (range) value of simulated median ETV V90 in various seed displacement scenarios was 97.8% (87.5-100%).

CONCLUSIONS

A PBSI postimplant dosimetry simulation framework was developed and validated. Simulation results showed that the current PTV planning margin is sufficient to provide adequate postimplant dose coverage of ETV. This simulator can be used to evaluate plan robustness to seed displacement and will facilitate future research in improving PBSI planning methods.

Establishing a simulation-based education program for radiation oncology learners in permanent seed implant brachytherapy: Building validation evidence

PURPOSE

The purpose of this study was to establish a simulation-based education program for radiation oncology learners in permanent seed implant brachytherapy. The first step in formalizing any education program is a validation process that builds evidence-based verification that the learning environment is appropriate.

METHODS AND MATERIALS

The primary education task allowed practitioners to use an anthropomorphic breast phantom to simulate a permanent seed implant brachytherapy delivery. Validation evidence is built by generating data to assess learner and expert cohorts according to their proficiency. Each practitioner's performance during the simulation was evaluated by seed placement accuracy, procedural time-to-complete, and two qualitative evaluation tools-a global rating scale and procedural checklist.

RESULTS

The average seed placement accuracy (±SD) was 8.1 ± 3.5 mm compared to 6.1 ± 2.6 mm for the learner and expert cohort, respectively. The median (range) procedural time-to-complete was 64 (60-77) minutes and 43 (41-50) minutes for the learner and expert cohort, respectively. Seed placement accuracy (student t-test, p < 0.05) and procedural time-to-complete (Mann-Whitney U-test, p < 0.05) were statistically different between the cohorts. In both the global rating scale and procedural checklist, the expert cohort demonstrated improved proficiency compared to the learner cohort.

CONCLUSIONS

This validation evidence supports the utilization of this simulation environment toward appropriately capturing the delivery experience of practitioners. The results demonstrate that, in all areas of evaluation, expert cohort proficiency was superior to learner cohort proficiency. This methodology will be used to establish a simulation-based education program for radiation oncology learners in permanent seed implant brachytherapy.

Post-implant analysis in permanent breast seed implant: automated plan reconstruction using simulated annealing

Purpose

Post-implant analysis in permanent breast seed implant (PBSI) brachytherapy is an important component of the quality assurance process that indicates dosimetric quality relevant to patient outcome, indicating salvage therapy if inadequate, as well as providing feedback to the brachytherapy team to improve future treatments. To measure geometric indices on implant quality, plan reconstruction must be performed to correlate each planned and post-implant seed location. In this work, a simulated-annealing-based algorithm is developed to perform this plan reconstruction automatically.

Material and methods

The plan reconstruction algorithm was developed in MATLAB, taking the patient pre-treatment and post-implant (Day 0) plan and associated contours as inputs. For 19 treated patients, a reconstruction was obtained that defined the correspondence between each planned and post-implant seed. The simulated-annealing algorithm was used to reconstruct each patient 10 times to assess the variability in convergence. Manual reconstructions performed by at least two independent observers to obtain consensus were defined as the ground truth; these were compared to the automatic reconstructions obtained by the algorithm. Metrics on seed placement accuracy and needle strand angulation were calculated for the patients.

Results

The algorithm performed reconstructions on 19 patients (1235 seeds) with ground-truth reconstructions, obtaining 97 ± 8% correct matches. This strong performance indicates the ability to incorporate this algorithm into the clinical quality assurance workflow.

Conclusions

The plan reconstruction algorithm developed herein performed very well in a 19-patient cohort. This algorithm can be incorporated into the clinical process to assist in the assessment of center-specific seed placement accuracy and can be used to gather implant metrics in an automated, standardized fashion for future PBSI trials.

Use of needle track detection to quantify the displacement of stranded seeds following prostate brachytherapy

We aim to compute the movement of permanent stranded implant brachytherapy radioactive sources (seeds) in the prostate from the planned seed distribution to the intraoperative fluoroscopic distribution, and then to the postimplant computed tomography (CT) distribution. We present a novel approach to matching the seeds in these distributions to the plan by grouping the seeds into needle tracks. First, we identify the implantation axis using a sample consensus algorithm. Then, we use a network flow algorithm to group seeds into their needle tracks. Finally, we match the needles from the three stages using both their transverse plane location and the number of seeds per needle. We validated our approach on eight clinical prostate brachytherapy cases, having a total of 871 brachytherapy seeds distributed in 193 needles. For the intraoperative and postimplant data, 99.31% and 99.41% of the seeds were correctly assigned, respectively. For both the preplan to fluoroscopic and fluoroscopic to CT registrations, 100% of the needles were correctly matched. We show that there is an average intraoperative seed displacement of 4.94±2.42 mm and a further 2.97±1.81 mm of postimplant movement. This information reveals several directional trends and can be used for quality control, treatment planning, and intraoperative dosimetry that fuses ultrasound and fluoroscopy.

Matching and reconstruction of brachytherapy seeds using the Hungarian algorithm (MARSHAL)

Intraoperative dosimetric quality assurance in prostate brachytherapy critically depends on discerning the three-dimensional (3D) locations of implanted seeds. The ability to reconstruct the implanted seeds intraoperatively will allow us to make immediate provisions for dosimetric deviations from the optimal implant plan. A method for seed reconstruction from segmented C-arm fluoroscopy images is proposed. The 3D coordinates of the implanted seeds can be calculated upon resolving the correspondence of seeds in multiple x-ray images. We formalize seed-matching as a combinatorial optimization problem, which has salient features: (a) extensively studied solutions by the computer science community; (b) proof for the nonexistence of any polynomial time exact algorithm; and (c) a practical pseudo-polynomial algorithm that mostly runs in O(N3) time using any number of images. We prove that two images are insufficient to correctly match the seeds, while a third image renders the matching problem to be of nonpolynomial complexity. We utilize the special structure of the problem and propose a pseudopolynomial time algorithm. Using three presegmented images, matching and reconstruction of brachytherapy seeds using the Hungarian algorithm achieved complete matching in simulation experiments; and 98.5% in phantom experiments. 3D reconstruction error for correctly matched seeds has a mean of 0.63 mm, and 0.9 mm for incorrectly matched seeds. The maximum seed reconstruction error in each implant was typically around 1.32 mm. Both on synthetic data and in phantom experiments, matching rate and reconstruction error achieved using presegmented images was found to be sufficient for prostate brachytherapy. The algorithm is extendable to deal with arbitrary number of images without any loss in speed or accuracy. The algorithm is sufficiently generic to provide a practical solution to any correspondence problem, across different imaging modalities and features.

The robustness of dose distributions to displacement and migration of 125I permanent seed implants over a wide range of seed number, activity, and designs

PURPOSE

To investigate the robustness of permanent prostate implant dosimetry for various (125)I seed activities and various seed models. The dosimetric impact of seed misplacement and seed migration (seed loss) is also taken into account using various standard dose indices.

METHODS AND MATERIALS

A dose-based inverse planning algorithm is used for automated dosimetric plan creation (45-60 s per plan) and provides an unbiased way to compare the robustness of various optimal dosimetric plans. Seed misplacement and seed migration are simulated by way of Monte Carlo, based on the measured displacement distributions from clinical postimplant cases. Plans were generated for seed activities between 0.2 and 1.4 mCi (0.25 to 1.78 U) and for 11 different seed models.

RESULTS

The numbers of seeds and needles are shown to decrease rapidly for a seed activity between 0.3 mCi and 0.6 mCi (0.38 and 0.76 U). The loss in V100, from 100%, because of seed misplacement is below 10% for an apparent activity ranging from 0.2 to 0.9 mCi (0.25 to 1.14 U). A minimum degradation in V100 is observed around 0.6-0.7 mCi (0.76-0.89 U). D90 increases from 150 to 170 Gy between 0.3 and 0.7 mCi (0.38 and 0.89 U) and decreases afterward to fall below 140 Gy at higher activity. V200 and D10 to the target volume both show an increase in hot spots up to 0.7 mCi, and then decrease linearly at higher activities for all seed models. V200 and D10 to the urethra remain about constant for all seed activities up to 0.8 mCi (1.02 U), at which point they start to decrease. All seed models follow this general trend.

CONCLUSIONS

Plans were shown to be robust to misplacement and migration of seeds over a wide range of seed activity and for various seed models. With a properly tuned inverse planning algorithm able to ensure the dose coverage and protection for the organs at risk in the presence of placement errors (displacement and migration), the choice of a preferred seed activity, in a range up to about 0.7 mCi (0.89 U), is open. The upper part of this range offers the opportunity to significantly reduce the number of seeds and needles, thus reducing surgical trauma to the patient, saving time in an operating room planning setting, and reducing the cost of a permanent prostate implant procedure.