Clinical evaluation of a deep learning CBCT auto-segmentation software for prostate adaptive radiation therapy.

Purpose: Aim of the present study is to characterize a deep learning-based auto-segmentation software (DL) for prostate cone beam computed tomography (CBCT) images and to evaluate its applicability in clinical adaptive radiation therapy routine. Materials and methods: Ten patients, who received exclusive radiation therapy with definitive intent on the prostate gland and seminal vesicles, were selected. Femoral heads, bladder, rectum, prostate, and seminal vesicles were retrospectively contoured by four different expert radiation oncologists on patients CBCT, acquired during treatment. Consensus contours (CC) were generated starting from these data and compared with those created by DL with different algorithms, trained on CBCT (DL-CBCT) or computed tomography (DL-CT). Dice similarity coefficient (DSC), centre of mass (COM) shift and volume relative variation (VRV) were chosen as comparison metrics. Since no tolerance limit can be defined, results were also compared with the inter-operator variability (IOV), using the same metrics. Results: The best agreement between DL and CC was observed for femoral heads (DSC of 0.96 for both DL-CBCT and DL-CT). Performance worsened for low-contrast soft tissue organs: the worst results were found for seminal vesicles (DSC of 0.70 and 0.59 for DL-CBCT and DL-CT, respectively). The analysis shows that it is appropriate to use algorithms trained on the specific imaging modality. Furthermore, the statistical analysis showed that, for almost all considered structures, there is no significant difference between DL-CBCT and human operator in terms of IOV. Conclusions: The accuracy of DL-CBCT is in accordance with CC; its use in clinical practice is justified by the comparison with the inter-operator variability.

Impact of beam complexity on plan delivery accuracy verification of a transmission detector in volumetric modulated arc therapy.

OBJECTIVE: To study the effect of beam complexity on VMAT delivery accuracy evaluated by means of a transmission detector, together with the possibility of scoring plan complexity. METHODS: 43 clinical VMAT plans delivered by a TrueBeam linear accelerator to both Delta4 Discover and Delta4 Phantom+ for patient-specific quality assurance were evaluated. Global Dose-γ analysis, MLC-γ analysis, percentage of leaves with a deviation between planned and measured leaf tip position lower than 1 mm (LD) were computed. Modulation complexity score (MCSv), average leaf travel (LT), a multiplicative combination of LT and MCSv (LTMCS), percentage of leaves with speed lower than 5 mm/s (LS), from 5 to 20 mm/s (MS), higher than 20 mm/s (HS) and the average value of leaf speed (MLCSav) were evaluated by means of an home-made Matlab script. RESULTS: Dose-γ passing rate showed a moderate correlation with MCSv, LT, MLCSav, LS and HS, while a stronger positive correlation was found with LTMCS. A strong correlation was observed between LD and both LT and leaves speed, while a weak correlation was observed with MCSv. A correlation between MLC-γ pass rate and plan complexity parameters was found except for MCSv; a moderate correlation with LS was observed, while all other parameters showed weak correlations. CONCLUSIONS: The study confirmed the possibility to establish correlations between plan complexity indices versus dose distribution and MLC parameters measured by a transmissive detector. Further investigation is necessary to define specific values of the complexity indices to evaluate whether a VMAT plan is deliverable as intended.

Preliminary evaluation of a novel secondary check tool for intensity modulated radiotherapy treatment planning.

PURPOSE: To evaluate the dosimetric accuracy of the Delta4 Insight (DI) secondary-check dosimetry system. METHODS: Absolute dosimetry in reference conditions, output factors, percent depth doses normalized and off-axis dose profiles for different field sizes calculated by DI were compared with measurements. Dose calculations for 20 clinical IMRT/VMAT plans generated in the TPS using both AAA or AcurosXB algorithms were compared with measurements. The average difference between calculated and measured point dose in high-dose region was calculated for all cases. 3D dose measurements were performed in Delta4 Phantom+ and a comparison between calculated and measured dose distributions was performed by means of the gamma analysis with 3 %/2 mm criteria. The dose distributions calculated by DI for 20 IMRT/VMAT plans were compared with those calculated by the TPS. RESULTS: The absolute dosimetry computed by DI showed dose value in agreement with the measured one within 0.3 %. The average differences between measured and calculated output factors were less than 2.5 %. The average PDD differences were less than 0.6 %. An excellent agreement between calculations and off-axis measurements is found. The point doses calculated for the 20 recalculated plan showed good agreement with measurements with average differences less than 0.5 %. The average gamma pass rate values for the Delta4 Phantom + 3D dose analysis was greater than 97.%. The comparison of DI with theTPS showed good agreement for the used metrics. CONCLUSIONS: Delta4 Insight may provide a useful independent secondary dose verification system for IMRT/VMAT plans, complementing the traditional global QA protocols.

Implementation of a Commercial Deep Learning-Based Auto Segmentation Software in Radiotherapy: Evaluation of Effectiveness and Impact on Workflow.

Proper delineation of both target volumes and organs at risk is a crucial step in the radiation therapy workflow. This process is normally carried out manually by medical doctors, hence demanding timewise. To improve efficiency, auto-contouring methods have been proposed. We assessed a specific commercial software to investigate its impact on the radiotherapy workflow on four specific disease sites: head and neck, prostate, breast, and rectum. For the present study, we used a commercial deep learning-based auto-segmentation software, namely Limbus Contour (LC), Version 1.5.0 (Limbus AI Inc., Regina, SK, Canada). The software uses deep convolutional neural network models based on a U-net architecture, specific for each structure. Manual and automatic segmentation were compared on disease-specific organs at risk. Contouring time, geometrical performance (volume variation, Dice Similarity Coefficient-DSC, and center of mass shift), and dosimetric impact (DVH differences) were evaluated. With respect to time savings, the maximum advantage was seen in the setting of head and neck cancer with a 65%-time reduction. The average DSC was 0.72. The best agreement was found for lungs. Good results were highlighted for bladder, heart, and femoral heads. The most relevant dosimetric difference was in the rectal cancer case, where the mean volume covered by the 45 Gy isodose was 10.4 cm3 for manual contouring and 289.4 cm3 for automatic segmentation. Automatic contouring was able to significantly reduce the time required in the procedure, simplifying the workflow, and reducing interobserver variability. Its implementation was able to improve the radiation therapy workflow in our department.

Hypofractionation and Concomitant Boost in Ductal Carcinoma In Situ (DCIS): Analysis of a Prospective Case Series with Long-Term Follow-Up.

We previously reported on a cohort of breast cancer patients affected with ductal carcinoma in situ (DCIS) that were treated with breast conservative surgery and hypofractionated whole-breast radiotherapy with a concomitant boost to the lumpectomy cavity. We now report on the long-term results of the oncological and toxicity outcomes, at a median follow-up of 11.2 years. We also include an analysis of the predictive factors for local recurrence (LR). Eighty-two patients with long-term observation were considered for this report. All received hypofractionated post-operative radiotherapy with a concomitant boost (45 Gy/20 fractions to the whole breast and 50 Gy/20 fractions to the lumpectomy cavity). We report on LC rates at 5 and 10 years, overall survival (OS), and breast-cancer-specific survival (BCSS), employing the Kaplan-Meier method. Cox proportional regression analysis was used to determine the role of selected clinical parameters on the risk of local recurrence, by the univariate and multivariate models. After a median follow-up of 11.2 years (range 5-15 years), 9 pts (11%) developed LR. The LR rates at 5 years and 10 years were 2.4% and 8.2%, respectively. The 5- and 10-year overall survival rates were 98.8% and 91.6%, respectively. The 5- and 10-year breast-cancer-specific survival rates were 100.0% and 99.0%. Late skin and subcutaneous toxicities were generally mild, and cosmetic results were good-excellent for most patients. For the univariate regression analysis, ER positive status (HR; 95% CI, p = 0.021), PgR positive status (HR; 95% CI, p = 0.012), and the aggregate data of positive hormonal status (HR; 95% CI, p = 0.021) were inversely correlated to LR risk. Conversely, a high tumor grade (G3) was directly correlated with the risk of LR (HR; 95% CI, p = 0.048). For the multivariate regression analysis, a high tumor grade (G3) confirmed its negative impact on LR (HR 0.40; 95% CI 0.19-0.75, p = 0.047). Our long-term data demonstrate hypofractionated whole-breast radiotherapy with a concomitant boost to be feasable, effective, and tolerable. Our experience suggests positive hormonal status to be protective with respect to LR risk. A high tumor grade is a risk factor for LR.

Effect of delaying surgery by more than 10 weeks after neoadjuvant therapy in rectal cancer: a single institution experience.

The optimal timing of surgery after neoadjuvant chemoradiotherapy (CRT) in locally advanced rectal cancer is still controversial. Aim of this study was to evaluate the effect of increasing time interval between the end of CRT and surgery on pathological outcomes. This is a retrospective analysis on 114 patients treated with long-course neoadjuvant RT with or without chemotherapy between January 2005 and September 2020. 43 patients underwent surgery within 10 weeks from the end of CRT (1st group), whereas 71 patients underwent total mesorectal excision with a time interval equal or greater than 10 weeks (2nd group). Primary endpoint was pCR (pathological complete response). Secondary endpoints were near pCR (ypT0-1 N0), tumor downstaging (ypT less than cT), nodal downstaging (ypN less than cN), and overall response comparing clinical with pathological TN stage. Overall, the pCR rate was 8.8%, whereas we observed no significantly difference in primary endpoint between the two groups. Considering near pCR, a trend toward significant difference in favor of 2nd group was seen (p = 0.072). Tumor and nodal downstaging rates were 39.5%, 41.9%, 59.2%, and 56.3% in the 1st and 2nd group, respectively, with a statistically significant difference for T category (p = 0.042). Overall response rates (TN stage) showed a trend toward significant difference in favor of patients of the ≥ 10 week group (p = 0.059). Our study suggests that a prolonged time interval between the end of CRT and surgery (≥ 10 weeks) increases pathological response rates.

Delta4 Discover transmission detector: A comprehensive characterization for in-vivo VMAT monitoring.

OBJECTIVE: To investigate the dosimetric behaviour, influence on photon beam fluence and error detection capability of Delta4 Discover transmission detector. METHODS: The transmission detector (TRD) was characterized on a TrueBeam linear accelerator with 6 MV beams. Linearity, reproducibility and dose rate dependence were investigated. The effect on photon beam fluence was evaluated in terms of beam profiles, percentage depth dose, transmission factor and surface dose for different open field sizes. The transmission factor of the 10x10 cm2 field was entered in the TPS's configuration and its correct use in the dose calculation was verified recalculating 17 clinical IMRT/VMAT plans. Surface dose was measured for 20 IMRT fields. The capability to detect different delivery errors was investigated evaluating dose gamma index, MLC gamma index and leaf position of 15 manually modified VMAT plans. RESULTS: TRD showed a linear dependence on MU. No dose rate dependence was observed. Short-term and long-term reproducibility were within 0.1% and 0.5%. The presence of the TRD did not significantly affect PDDs and profiles. The transmission factor of the 10x10 cm2 field size was 0.985 and 0.983, for FF and FFF beams respectively. The 17 recalculated plans met our clinical gamma-index passing rate, confirming the correct use of the transmission factor by the TPS. The surface dose differences for the open fields increase for shorter SSDs and greater field size. Differences in surface dose for the IMRT beams were less than 2%. Output variation ≥2%, collimator angle variations within 0.3°, gantry angle errors of 1°, jaw tracking and leaf position errors were detected. CONCLUSIONS: Delta4 Discover shows good linearity and reproducibility, is not dependent on dose rate and does not affect beam quality and dose profiles. It is also capable to detect dosimetric and geometric errors and therefore it is suitable for monitoring VMAT delivery.

Computed Tomography to Cone Beam Computed Tomography Deformable Image Registration for Contour Propagation Using Head and Neck, Patient-Based Computational Phantoms: A Multicenter Study.

PURPOSE: To investigate the performance of various algorithms for deformable image registration (DIR) for propagating regions of interest (ROIs) using multiple commercial platforms, from computed tomography to cone beam computed tomography (CBCT) and megavoltage computed tomography. METHODS AND MATERIALS: Fourteen institutions participated in the study using 5 commercial platforms: RayStation (RaySearch Laboratories, Stockholm, Sweden), MIM (Cleveland, OH), VelocityAI and SmartAdapt (Varian Medical Systems, Palo Alto, CA), and ABAS (Elekta AB, Stockholm, Sweden). Algorithms were tested on synthetic images generated with the ImSimQA (Oncology Systems Limited, Shrewsbury, UK) package by applying 2 specific deformation vector fields (DVF) to real head and neck patient datasets. On-board images from 3 systems were used: megavoltage computed tomography from Tomotherapy and 2 kinds of CBCT from a clinical linear accelerator. Image quality of the system was evaluated. The algorithms' accuracy was assessed by comparing the DIR-mapped ROIs returned by each center with those of the reference, using the Dice similarity coefficient and mean distance to conformity metrics. Statistical inference on the validation results was carried out to identify the prognostic factors of DIR performance. RESULTS: Analyzing 840 DIR-mapped ROIs returned by the centers, it was demonstrated that DVF intensity and image quality were significant prognostic factors of DIR performance. The accuracy of the propagated contours was generally high, and acceptable DIR performance can be obtained with lower-dose CBCT image protocols. CONCLUSIONS: The performance of the systems proved to be image quality specific, depending on the DVF type and only partially on the platforms. All systems proved to be robust against image artifacts and noise, except the demon-based software.

Dosimetric characterization and behaviour in small X-ray fields of a microchamber and a plastic scintillator detector.

OBJECTIVE: The aim of this work was to investigate the main dosimetric characteristics and the performance of an A26 Exradin ionization microchamber (A26 IC) and a W1 Exradin plastic scintillation detector (W1 PSD) in small photon beam dosimetry for treatment planning system commissioning and quality assurance programme. METHODS: Detector characterization measurements (short-term stability, dose linearity, angular dependence and energy dependence) were performed in water for field sizes up to 10 × 10 cm2. Polarity effect (Ppol) was examined for the A26 IC. The behaviour of the detectors in small field relative dosimetry [percentage depth dose, dose profiles often called the off-axis ratio and output factors (OFs)] was investigated for field sizes ranging from 1 × 1 to 3 × 3 cm2. RESULTS: Results were compared with those obtained with other detectors we already use for small photon beam dosimetry. A26 IC and W1 PSD showed a linear dose response. While the A26 IC showed no energy dependence, the W1 PSD showed energy dependence within 2%; no angular dependence was registered. Ppol values for A26 IC were below 0.9% (0.5% for field size >2 × 2 cm2). A26 IC and W1 PSD depth-dose curves and lateral profiles agreed with those obtained with an EDGE diode. No differences were observed among the detectors in OF measurement for field sizes larger than 1 × 1 cm2, with average differences <1%. For field sizes <1 × 1 cm2, the effective volume of ionization chamber and non-water equivalence of EDGE diode become significant. A26 IC OF values were significantly lower than EDGE diode and W1 PSD values, with percentage differences of about -23 and -13% for the smallest field, respectively. W1 PSD OF values lay between ion chambers and diode values, with a maximum percentage difference of about -10% with respect to the EDGE diode, for a 6 × 6-mm2 field size. CONCLUSION: The results of our investigation confirm that A26 IC and W1 PSD could play an important role in small field relative dosimetry. Advances in knowledge: Dosimetric characteristics of Exradin A26 ionization microchamber and W1 plastic scintillation detector for small field dosimetry.