Feasibility evaluation of novel AI-based deep-learning contouring algorithm for radiotherapy.

PURPOSE: To evaluate the clinical feasibility of the Siemens Healthineers AI-Rad Companion Organs RT VA30A (Organs-RT) auto-contouring algorithm for organs at risk (OARs) of the pelvis, thorax, and head and neck (H&N). METHODS: Computed tomography (CT) datasets from 30 patients (10 pelvis, 10 thorax, and 10 H&N) were collected. Four sets of OARs were generated on each scan, one set by Organs-RT and the others by three experienced users independently. A physician (expert) then evaluated each contour by assigning a score from the following scale: 1-Must Redo, 2-Major Edits, 3-Minor Edits, 4-Clinically usable. Using the highest-scored OAR from the human users as a reference, the contours generated by Organs-RT were evaluated via Dice Similarity Coefficient (DSC), Hausdorff Distance (HDD), Mean Distance to Agreement (mDTA), Volume comparison, and visual inspection. Additionally, each human user recorded the time to delineate each structure set and time-saving efficiency was measured. RESULTS: The average DSC obtained for the pelvic OARs ranged between (0.81 ± 0.06)Rectum and (0.94 ± 0.03)Bladder . (0.75 ± 0.09)Esophagus to ( 0.96 ± 0.02 ) Rt . Lung ${( {0.96 \pm 0.02} )}_{{\mathrm{Rt}}.{\mathrm{\ Lung}}}$ for the thoracic OARs and (0.66 ± 0.07)Lips to (0.83 ± 0.04)Brainstem for the H&N. The average HDD in cm for the pelvis cohort ranged between (0.95 ± 0.35)Bladder to (3.62 ± 2.50)Rectum , (0.42 ± 0.06)SpinalCord to (2.09 ± 2.00)Esophagus for the thoracic set and ( 0.53 ± 0.22 ) Cerv _ SpinalCord ${( {0.53 \pm 0.22} )}_{{\mathrm{Cerv}}\_{\mathrm{SpinalCord}}}$ to (1.50 ± 0.50)Mandible for the H&N region. The time-saving efficiency was 67% for H&N, 83% for pelvis, and 84% for thorax. 72.5%, 82%, and 50% of the pelvis, thorax, and H&N OARs were scored as clinically usable by the expert, respectively. CONCLUSIONS: The highest agreement registered between OARs generated by Organs-RT and their respective references was for the bladder, heart, lungs, and femoral heads, with an overall DSC≥0.92. The poorest agreement was for the rectum, esophagus, and lips, with an overall DSC⩽0.81. Nonetheless, Organs-RT serves as a reliable auto-contouring tool by minimizing overall contouring time and increasing time-saving efficiency in radiotherapy treatment planning.

Dose perturbation due to the polysulfone cap surrounding a Fletcher-Williamson colpostat.

We conducted a metrological evaluation of the dosimetric impact due to the polysulfone cap used with the Fletcher-Williamson (FW) colpostat for 192Ir high-dose rate and pulsed-dose rate intracavitary brachytherapy using Monte Carlo simulations. Polysulfone caps with diameter of 30 mm, 25 mm, 20 mm, and 16 mm (mini-ovoid) were simulated and the absorbed dose rate in the surrounding water was calculated and compared to the dose rate for a bare 192Ir source in water. The dose perturbation depended on the cap diameter, distance away from the cap surface, and angular position around the cap. The largest dose rate reductions were found to be in the direction of the tumor bed where the cap is thickest. The range of perturbation over all depths and cap diameters was +2.8% (dose enhancement) to -6.8% (dose reduction). The FW colpostat cap's material composition should be modified to reduce this dosimetric effect or brachytherapy treatment planning dose algorithms should be improved to account for this perturbation.

Development of prototype shielded cervical intracavitary brachytherapy applicators compatible with CT and MR imaging.

PURPOSE: Intracavitary brachytherapy (ICBT) is an integral part of the treatment regimen for cervical cancer and, generally, outcome in terms of local disease control and complications is a function of dose to the disease bed and critical structures, respectively. Therefore, it is paramount to accurately determine the dose given via ICBT to the tumor bed as well as critical structures. This is greatly facilitated through the use of advanced three-dimensional imaging modalities, such as CT and MR, to delineate critical and target structures with an ICBT applicator inserted in vivo. These methods are not possible when using a shielded applicator due to the image artifacts generated by interovoid shielding. The authors present two prototype shielded ICBT applicators that can be utilized for artifact-free CT image acquisition. They also investigate the MR amenability and dosimetry of a novel tungsten-alloy shielding material to extend the functionality of these devices. METHODS: To accomplish artifact-free CT image acquisition, a "step-and-shoot" (S&S) methodology was utilized, which exploits the prototype applicators movable interovoid shielding. Both prototypes were placed in imaging phantoms that positioned the applicators in clinically applicable orientations. CT image sets were acquired of the prototype applicators as well as a shielded Fletcher-Williamson (sFW) ovoid. Artifacts present in each CT image set were qualitatively compared for each prototype applicator following the S&S methodology and the sFW. To test the novel tungsten-alloy shielding material's MR amenability, they constructed a phantom applicator that mimics the basic components of an ICBT ovoid. This phantom applicator positions the MR-compatible shields in orientations equivalent to the sFW bladder and rectal shields. MR images were acquired within a gadopentetate dimeglumine-doped water tank using standard pulse sequences and examined for artifacts. In addition, Monte Carlo simulations were performed to match the attenuation due to the thickness of this new shield type with current, clinically utilized ovoid shields and a 192Ir HDR/PDR source. RESULTS: Artifact-free CT images could be acquired of both generation applicators in a clinically applicable geometry using the S&S method. MR images were acquired of the phantom applicator containing shields, which contained minimal, clinically relevant artifacts. The thickness required to match the dosimetry of the MR-compatible and sFW rectal shields was determined using Monte Carlo simulations. CONCLUSIONS: Utilizing a S&S imaging method in conjunction with prototype applicators that feature movable interovoid shields, they were able to acquire artifact-free CT image sets in a clinically applicable geometry. MR images were acquired of a phantom applicator that contained shields composed of a novel tungsten alloy. Artifacts were largely limited to regions within the ovoid cap and are of no clinical interest. The second generation A3 utilizes this material for interovoid shielding.

Monte Carlo calculations of the absorbed dose and energy dependence of plastic scintillators.

Detector systems using plastic scintillators can provide instantaneous measurements with high spatial resolution in many applications including small field and high dose gradient field applications. Energy independence and water equivalence are important dosimetric properties that determine whether a detector will be useful in a clinical setting. Using Monte Carlo simulations, we calculated the energy dependence of plastic scintillators when exposed to photon beams in the radiotherapeutic range. These calculations were performed for a detector comprised of a BC-400 plastic scintillator surrounded by a polystyrene wall. Our results showed the plastic scintillation detector to be nearly energy independent over a range of energies from 0.5 to 20 MeV. The ratio of the dose absorbed by the scintillator to that absorbed by water was nearly a constant, approximately equal to 0.98 over the entire energy range of interest. These results confirm the water equivalence of the plastic scintillation detector and are in very good agreement with earlier results obtained using Burlin cavity theory.

Comparison of Monte Carlo calculations around a Fletcher Suit Delclos ovoid with radiochromic film and normoxic polymer gel dosimetry.

The Fletcher Suit Delclos (FSD) ovoids employed in intracavitary brachytherapy (ICB) for cervical cancer contain shields to reduce dose to the bladder and rectum. Many treatment planning systems (TPS) do not include the shields and other ovoid structures in the dose calculation. Instead, TPSs calculate dose by summing the dose contributions from the individual sources and ignoring ovoid structures such as the shields. The goal of this work was to calculate the dose distribution with Monte Carlo around a Selectron FSD ovoid and compare these calculations with radiochromic film (RCF) and normoxic polymer gel dosimetry. Monte Carlo calculations were performed with MCNPX 2.5.c for a single Selectron FSD ovoid with and without shields. RCF measurements were performed in a plane parallel to and displaced laterally 1.25 cm from the long axis of the ovoid. MAGIC gel measurements were performed in a polymethylmethacrylate phantom. RCF and MAGIC gel were irradiated with four 33 microGy m2 h(-1) Cs-137 pellets for a period of 24 h. Results indicated that MCNPX calculated dose to within +/- 2% or 2 mm for 98% of points compared with RCF measurements and to within +/- 3% or 3 mm for 98% of points compared with MAGIC gel measurements. It is concluded that MCNPX 2.5.c can calculate dose accurately in the presence of the ovoid shields, that RCF and MAGIC gel can demonstrate the effect of ovoid shields on the dose distribution and the ovoid shields reduce the dose by as much as 50%.

Dosimetric evaluation of the Fletcher-Williamson ovoid for pulsed-dose-rate brachytherapy: a Monte Carlo study.

We used radiochromic film dosimetry to validate a Monte Carlo (MC) model of a 192Ir pulsed-dose-rate (PDR) source inside a Fletcher-Williamson ovoid. MD-55-2 radiochromic film was placed in a high-impact polystyrene phantom in a plane parallel to and displaced 2.0 cm medially from the long axis of the ovoid. MC N-particle transport code (MCNPX) version 2.4 was used to model the ovoid and the 192Ir source. Energy deposition was calculated using a track-length estimator modified by an energy-dependent heating function, which is a good approximation of the collision kerma. To convert the estimates of the MC dose per simulated particle to clinically relevant absolute dosimetry, additional MC models of an actual and a virtual 192Ir source in dry air were simulated to determine air kerma strength for the penetrating part of the photon spectrum (>11.3 keV). The absolute dose distributions predicted by MCNPX agreed with the film results and were within +/-9.4% (k = 2) and within +/-2% or within a distance to agreement of 2 mm for 94% of the dose grid. Additional MC models characterized the uncertainty resulting from source positioning inside the ovoid. For a worst-case scenario of 1 mm off centre from the nominal source position in the 3 mm diameter ovoid shaft, the average dose deviation over the film plane was +/-5% (1sigma = +/-4%), with maximum deviation near the sharp dose-gradient provided by the shields of -20% to + 26%. A validated MC model is the first requirement to simulate common LDR clinical loadings (5-20 mgRaEq) and, thus, will aid in the transition from the current 137Cs Selectron LDR ICBT to PDR for treatment of gynecologic cancers.

Prostate implant dosimetric outcomes and migration patterns between bio-absorbable coated and uncoated brachytherapy seeds.

PURPOSE: To evaluate the lung and pelvic seed migration and intraprostatic dose variability for prostate seed implant (PSI) using bio-absorbable polymer "coated" seeds for intraoperative planning. METHODS AND MATERIALS: A total of 100 PSI patients were initially implanted with uncoated I-125 (STM 1251 or I125-SL, N = 85) or Pd-103 (mod 200, N = 15) seeds, and 105 PSI patients were implanted subsequently with coated seeds using inverse optimization with real-time planning. Implant technique, average number of needles, and dose objectives remained identical among the cohorts. RESULTS: Day 30 postimplant comparison of seed migration demonstrated a significant reduction in overall lung and pelvic seed migration from 25% (uncoated) to 4% (coated) (p < 0.0001). A measurable reduction in intraprostatic dose variability was observed in patients with the coated seeds when comparing 30 days dosimetry results for V100, V150, and D90 for prostate, and V110 for the rectum. A statistically significant reduction in the standard deviation from Day 0 to Day 30 for the above parameters for the prostate as well as for V110 of rectum was also observed. A significant improvement in implant quality at Day 30 was demonstrated using Radiation Therapy Oncology Group (RTOG) evaluation criteria range with the coated seeds cohort. CONCLUSIONS: PSI using coated seeds shows lower lung and pelvic seed migration compared with those using uncoated seeds and compares favorably to pelvic stranded seed migration reports. A higher concordance was observed with less dose variability in dosimetric parameters on Day 30 dosimetry compared with that on Day 0. Improvement in the implant quality was also observed using the RTOG criteria, suggesting reduced intraprostatic migration.