Physics and Imaging in Radiation Oncology最新文献

Background and Purpose

Treatment planning is a time-intensive task that could be automated. We aimed to develop a “single-click” workflow, fully deployed within a commercial treatment planning system (TPS), for autoplanning prostate radiotherapy treatment plans using predictions from a deep learning model (DLM).

Materials and Methods

Automatically generated treatment plans were created with a single script, executed from within a commercial TPS scripting environment, that performed two stages sequentially. Initially, a 3D dose distribution was predicted with a ResUNet DLM. The DLM was trained and validated using previously treated datasets (n = 120) which used 3D contours as inputs. Following this, dose predictions were converted into treatment plans by extracting dose-volume metrics from the predictions to use as objectives for the inverse optimizer within the TPS. An independent test dataset (n = 20) was used to evaluate the similarity between automated and clinical plans.

Results

For planning target volumes, the median percentage difference and interquartile range between the automatically generated plans and clinical plans were 0.4% [0.2-1.1%] for the V_100%, −0.5% [(−1.0)-(−0.2)%] for D_99% and −0.5% [(−1.0)-(−0.2)%] for D_95%. Bladder and rectum volume-at-dose objectives agreed within −6.1% [(−12.5)-0.9%]. The conversion of the DLM prediction into a treatment plan took 15 min [13-16 min].

Conclusions

An automatic plan generation workflow that uses a DL model with scripted optimization was fully deployed in a commercial TPS. Autoplans were compared to previously treated clinical plans and were found to be non-inferior.

Background and purpose

Image-based data mining (IBDM) requires spatial normalisation to reference anatomy, which is challenging in breast radiotherapy due to variations in the treatment position, breast shape and volume. We aim to optimise spatial normalisation for breast IBDM.

Materials and methods

Data from 996 patients treated with radiotherapy for early-stage breast cancer, recruited in the REQUITE study, were included. Patients were treated supine (n = 811), with either bilateral or ipsilateral arm(s) raised (551/260, respectively) or in prone position (n = 185). Four deformable image registration (DIR) configurations for extrathoracic spatial normalisation were tested. We selected the best-performing DIR configuration and further investigated two pathways: i) registering prone/supine cohorts independently and ii) registering all patients to a supine reference. The impact of arm positioning in the supine cohort was quantified. DIR accuracy was estimated using Normalised Cross Correlation (NCC), Dice Similarity Coefficient (DSC), mean Distance to Agreement (MDA), 95 % Hausdorff Distance (95 %HD), and inter-patient landmark registration uncertainty (ILRU).

Results

DIR using B-spline and normalised mutual information (NMI) performed the best across all evaluation metrics. Supine-supine registrations yielded highest accuracy (0.98 ± 0.01, 0.91 ± 0.04, 0.23 ± 0.19 cm, 1.17 ± 1.18 cm, 0.51 ± 0.26 cm for NCC, DSC, MDA, 95 %HD, and ILRU), followed by prone-prone and supine-prone registrations. Arm positioning had no significant impact on registration performance. For the best DIR strategy, uncertainty of 0.44 and 0.81 cm in the breast and shoulder regions was found.

Conclusions

B-spline algorithm using NMI and registered supine and prone cohorts independently provides the most optimal spatial normalisation strategy for breast IBDM.

Background and purpose

Monte Carlo (MC) based dose calculations are widely used in radiotherapy with a low statistical uncertainty, being accurate but slow. Increasing the uncertainty accelerates the calculation, but reduces quality. In online adaptive planning, however, dose is recalculated every treatment fraction, potentially decreasing the cumulative calculation error. This study aimed to evaluate the effect of higher MC statistical uncertainty in the context of daily online plan adaptation.

Materials and methods

For twenty prostate cancer patients, daily plans were simulated for 5 fractions and three modes of variation: rigid whole body translations, local-rigid prostate translations and local-rigid prostate rotations. For each mode and fraction, adaptive plans were generated from a clinical reference plan using three MC uncertainty values: 1 % (standard), 2 % and 3 % per plan. Dose-volume criteria were evaluated for accumulated doses, checking plan acceptability and comparing higher uncertainty plans to the standard.

Results

Increasing the statistical uncertainty setting from 1 % to 2–3 % caused an accumulated median target D_98% reduction of 0.1 Gy, with interquartile ranges (IQRs) up to 0.12 Gy. Rectum V_35Gy increased in median up to 0.16 cm³ with IQRs up to 0.33 cm³. The bladder V_28Gy and V_32Gy showed median increases up to 0.24 %-point, with IQRs up to 0.54 %-point. Using 2 % uncertainty reduced calculation times by more than a minute for all modes of variation, with no further time gain when increasing to 3 %.

Conclusion

A 2–3 % MC statistical uncertainty was clinically feasible. Using a 2 % uncertainty setting reduced calculation times at the cost of limited relative dose-volume differences.

Background and purpose

In vivo dosimetry is not standard in brachytherapy and some errors go undetected. The aim of this study was to evaluate the accuracy of multi-channel vaginal cylinder pulsed dose-rate brachytherapy using in vivo dosimetry.

Materials and methods

In vivo dosimetry data was collected during the years 2019–2022 for 22 patients (32 fractions) receiving multi-channel cylinder pulsed dose-rate brachytherapy. An inorganic scintillation detector was inserted in a cylinder channel. Each fraction was analysed as independent data sets. In vivo dosimetry-based source-tracking was used to determine the relative source-to-detector position. Measured dose was compared to planned and re-calculated source-tracking based doses. Assuming no change in organ and applicator geometry throughout treatment, the planned and source-tracking based dose distributions were compared in select volumes via γ-index analysis and dose-volume-histograms.

Results

The mean ± SD planned vs. measured dose deviations in the first pulse were 0.8 $\pm$ 5.9 %. In 31/32 fractions the deviation was within the combined in vivo dosimetry uncertainty (averaging 9.7 %, k = 2) and planning dose calculation uncertainty (1.6 %, k = 2). The dwell-position offsets were < 2 mm for 88 % of channels, with the largest being 5.1 mm (4.0 mm uncertainty, k = 2). 3 %/2 mm γ pass-rates averaged 97.0 % (clinical target volume (CTV)), 100.0 % (rectum), 99.9 % (bladder). The mean ± SD deviation was −1.$1$ ± 2.9 % for CTV D98, and −0.2 ± 0.9 % and −1.2 ± 2.5 %, for bladder and rectum D2cm³ respectively, indicating good agreement between intended and delivered dose.

Conclusions

In vivo dosimetry verified accurate and stable dose delivery in multi-channel vaginal cylinder based pulsed dose-rate brachytherapy.

This work investigates the use of a multi-2D cine magnetic resonance imaging-based comprehensive motion monitoring (CMM) system for the assessment of prostate intrafraction 3D drifts. The data of six healthy volunteers were analyzed and the values of a clinically-relevant registration quality factor metric exported by CMM were presented. Additionally, the CMM-derived prostate motion was compared to a 3D-based reference and the 2D-3D tracking agreement was reported. Due to the low quality of SI motion tracking (often $>$2 mm tracking mismatch between anatomical planes) we conclude that further improvements are desirable prior to clinical introduction of CMM for prostate drift corrections.

Background and purpose

The impact of intrafractional motion and deformations on clinical radiotherapy delivery has so far only been investigated by simulations as well as point and planar dose measurements. The aim of this study was to combine anthropomorphic 3D dosimetry with a deformable abdominal phantom to measure the influence of intra-fractional motion and gating in photon radiotherapy and evaluate the applicability in proton therapy.

Material and methods

An abdominal phantom was modified to hold a deformable anthropomorphic 3D dosimeter shaped as a human liver. A liver-specific photon radiotherapy and a proton pencil beam scanning therapy plan were delivered to the phantom without motion as well as with 12 mm sinusoidal motion while using either no respiratory gating or respiratory gating.

Results

Using the stationary irradiation as reference the local 3 %/2 mm 3D gamma index pass rate of the motion experiments in the planning target volume (PTV) was above 97 % (photon) and 78 % (proton) with gating whereas it was below 74 % (photon) and 45 % (proton) without gating.

Conclusions

For the first time a high-resolution deformable anthropomorphic 3D dosimeter embedded in a deformable abdominal phantom was applied for experimental validation of both photon and proton treatments of targets exhibiting respiratory motion. It was experimentally shown that gating improves dose coverage and the geometrical accuracy for both photon radiotherapy and proton therapy.

Proton therapy (PT) is an advancing radiotherapy modality increasingly integrated into clinical settings, transitioning from research facilities to hospital environments. A critical aspect of the commissioning of a proton pencil beam scanning delivery system is the acquisition of experimental beam data for accurate beam modelling within the treatment planning system (TPS). These guidelines describe in detail the acquisition of proton pencil beam modelling data. First, it outlines the intrinsic characteristics of a proton pencil beam—energy distribution, angular-spatial distribution and particle number. Then, it lists the input data typically requested by TPSs. Finally, it describes in detail the set of experimental measurements recommended for the acquisition of proton pencil beam modelling data—integrated depth-dose curves, spot maps in air, and reference dosimetry. The rigorous characterization of these beam parameters is essential for ensuring the safe and precise delivery of proton therapy treatments.

Background and purpose

Accurate and automated segmentation of targets and organs-at-risk (OARs) is crucial for the successful clinical application of online adaptive radiotherapy (ART). Current methods for cone-beam computed tomography (CBCT) auto-segmentation face challenges, resulting in segmentations often failing to reach clinical acceptability. Current approaches for CBCT auto-segmentation overlook the wealth of information available from initial planning and prior adaptive fractions that could enhance segmentation precision.

Materials and methods

We introduce a novel framework that incorporates data from a patient’s initial plan and previous adaptive fractions, harnessing this additional temporal context to significantly refine the segmentation accuracy for the current fraction’s CBCT images. We present LSTM-UNet, an innovative architecture that integrates Long Short-Term Memory (LSTM) units into the skip connections of the traditional U-Net framework to retain information from previous fractions. The models underwent initial pre-training with simulated data followed by fine-tuning on a clinical dataset.

Results

Our proposed model’s segmentation predictions yield an average Dice similarity coefficient of 79% from 8 Head & Neck organs and targets, compared to 52% from a baseline model without prior knowledge and 78% from a baseline model with prior knowledge but no memory.

Conclusions

Our proposed model excels beyond baseline segmentation frameworks by effectively utilizing information from prior fractions, thus reducing the effort of clinicians to revise the auto-segmentation results. Moreover, it works together with registration-based methods that offer better prior knowledge. Our model holds promise for integration into the online ART workflow, offering precise segmentation capabilities on synthetic CT images.

Background and purpose

Squamous cell carcinoma of the anus (SCCA) can recur after chemoradiotherapy (CRT). Early prediction of treatment response is crucial for individualising treatment. Existing data on radiological biomarkers is limited and contradictory. We performed an individual patient data meta-analysis (IPM) of four prospective trials investigating whether diffusion-weighted (DW) magnetic resonance imaging (MRI) in weeks two to three of CRT predicts treatment failure in SCCA.

Material and methods

Individual patient data from four trials, including paired DW-MRI at baseline and during CRT, were combined into one dataset. The association between ADC volume histogram parameters and treatment failure (locoregional and any failure) was assessed using logistic regression. Pre-defined analysis included categorising patients into a change in the mean ADC of the delineated tumour volume above and below 20%.

Results

The study found that among all included 142 patients, 11.3 % (n = 16) had a locoregional treatment failure. An ADC mean change of <20 % and >20 % resulted in a locoregional failure rate of 16.7 % and 8.0 %, respectively. However, no other ADC-based histogram parameter was associated with locoregional or any treatment failure.

Conclusions

DW-MRI standard parameters, as an isolated biomarker, were not found to be associated with increased odds of treatment failure in SCCA in this IPM. Radiological biomarker investigations involve multiple steps and can result in heterogeneous data. In future, it is crucial to include radiological biomarkers in large prospective trials to minimize heterogeneity and maximize learning.

Background and purpose

Deep-learning (DL) models for segmentation of the gross tumor volume (GTV) in radiotherapy are generally based on clinical delineations which suffer from inter-observer variability. The aim of this study was to compare performance of a DL-model based on clinical glioblastoma GTVs to a model based on a single-observer edited version of the same GTVs.

Materials and methods

The dataset included imaging data (Computed Tomography (CT), T1, contrast-T1 (T1C), and fluid-attenuated-inversion-recovery (FLAIR)) of 259 glioblastoma patients treated with post-operative radiotherapy between 2012 and 2019 at a single institute. The clinical GTVs were edited using all imaging data. The dataset was split into 207 cases for training/validation and 52 for testing.

GTV segmentation models (nnUNet) were trained on clinical and edited GTVs separately and compared using Surface Dice with 1 mm tolerance (sDSC_1mm). We also evaluated model performance with respect to extent of resection (EOR), and different imaging combinations (T1C/T1/FLAIR/CT, T1C/FLAIR/CT, T1C/FLAIR, T1C/CT, T1C/T1, T1C). A Wilcoxon test was used for significance testing.

Results

The median (range) sDSC_1mm of the clinical-GTV-model and edited-GTV-model both evaluated with the edited contours, was 0.76 (0.43–0.94) vs. 0.92 (0.60–0.98) respectively (p < 0.001). sDSC_1mm was not significantly different between patients with a biopsy, partial, and complete resection. T1C as single input performed as good as use of imaging combinations.

Conclusions

High segmentation accuracy was obtained by the DL-models. Editing of the clinical GTVs significantly increased DL performance with a relevant effect size. DL performance was robust for EOR and highly accurate using only T1C.