Reproducibility in Radiomics: A Comparison of Feature Extraction Methods and Two Independent Datasets.

Radiomics involves the extraction of information from medical images that are not visible to the human eye. There is evidence that these features can be used for treatment stratification and outcome prediction. However, there is much discussion about the reproducibility of results between different studies. This paper studies the reproducibility of CT texture features used in radiomics, comparing two feature extraction implementations, namely the MATLAB toolkit and Pyradiomics, when applied to independent datasets of CT scans of patients: (i) the open access RIDER dataset containing a set of repeat CT scans taken 15 min apart for 31 patients (RIDER Scan 1 and Scan 2, respectively) treated for lung cancer; and (ii) the open access HN1 dataset containing 137 patients treated for head and neck cancer. Gross tumor volume (GTV), manually outlined by an experienced observer available on both datasets, was used. The 43 common radiomics features available in MATLAB and Pyradiomics were calculated using two intensity-level quantization methods with and without an intensity threshold. Cases were ranked for each feature for all combinations of quantization parameters, and the Spearman's rank coefficient, rs, calculated. Reproducibility was defined when a highly correlated feature in the RIDER dataset also correlated highly in the HN1 dataset, and vice versa. A total of 29 out of the 43 reported stable features were found to be highly reproducible between MATLAB and Pyradiomics implementations, having a consistently high correlation in rank ordering for RIDER Scan 1 and RIDER Scan 2 (rs > 0.8). 18/43 reported features were common in the RIDER and HN1 datasets, suggesting they may be agnostic to disease site. Useful radiomics features should be selected based on reproducibility. This study identified a set of features that meet this requirement and validated the methodology for evaluating reproducibility between datasets.

The Application of Radiomics and AI to Molecular Imaging for Prostate Cancer.

Molecular imaging is a key tool in the diagnosis and treatment of prostate cancer (PCa). Magnetic Resonance (MR) plays a major role in this respect with nuclear medicine imaging, particularly, Prostate-Specific Membrane Antigen-based, (PSMA-based) positron emission tomography with computed tomography (PET/CT) also playing a major role of rapidly increasing importance. Another key technology finding growing application across medicine and specifically in molecular imaging is the use of machine learning (ML) and artificial intelligence (AI). Several authoritative reviews are available of the role of MR-based molecular imaging with a sparsity of reviews of the role of PET/CT. This review will focus on the use of AI for molecular imaging for PCa. It will aim to achieve two goals: firstly, to give the reader an introduction to the AI technologies available, and secondly, to provide an overview of AI applied to PET/CT in PCa. The clinical applications include diagnosis, staging, target volume definition for treatment planning, outcome prediction and outcome monitoring. ML and AL techniques discussed include radiomics, convolutional neural networks (CNN), generative adversarial networks (GAN) and training methods: supervised, unsupervised and semi-supervised learning.

Early dietitian referral in lung cancer: use of machine learning.

OBJECTIVES: The Dietetic Assessment and Intervention in Lung Cancer (DAIL) study was an observational cohort study. It triaged the need for dietetic input in patients with lung cancer, using questionnaires with 137 responses. This substudy tested if machine learning could predict need to see a dietitian (NTSD) using 5 or 10 measures. METHODS: 76 cases from DAIL were included (Royal Surrey NHS Foundation Trust; RSH: 56, Frimley Park Hospital; FPH 20). Univariate analysis was used to find the strongest correlates with NTSD and 'critical need to see a dietitian' CNTSD. Those with a Spearman correlation above ±0.4 were selected to train a support vector machine (SVM) to predict NTSD and CNTSD. The 10 and 5 best correlates were evaluated. RESULTS: 18 and 13 measures had a correlation above ±0.4 for NTSD and CNTSD, respectively, producing SVMs with 3% and 7% misclassification error. 10 measures yielded errors of 7% (NTSD) and 9% (CNTSD). 5 measures yielded between 7% and 11% errors. SVM trained on the RSH data and tested on the FPH data resulted in errors of 20%. CONCLUSIONS: Machine learning can predict NTSD producing misclassification errors <10%. With further work, this methodology allows integrated early referral to a dietitian independently of a healthcare professional.

Statistical limitations in ion imaging.

In this study, we investigated the capacity of various ion beams available for radiotherapy to produce high quality relative stopping power map acquired from energy-loss measurements. The image quality metrics chosen to compare the different ions were signal-to-noise ratio (SNR) as a function of dose and spatial resolution. Geant4 Monte Carlo simulations were performed for: hydrogen, helium, lithium, boron and carbon ion beams crossing a 20 cm diameter water phantom to determine SNR and spatial resolution. It has been found that protons possess a significantly larger SNR when compared with other ions at a fixed range (up to 36% higher than helium) due to the proton nuclear stability and low dose per primary. However, it also yields the lowest spatial resolution against all other ions, with a resolution lowered by a factor 4 compared to that of carbon imaging, for a beam with the same initial range. When comparing for a fixed spatial resolution of 10 lp cm-1, carbon ions produce the highest image quality metrics with proton ions producing the lowest. In conclusion, it has been found that no ion can maximize all image quality metrics simultaneously and that a choice must be made between spatial resolution, SNR, and dose.

Statistical limitations in proton imaging.

Proton imaging is a promising technology for proton radiotherapy as it can be used for: (1) direct sampling of the tissue stopping power, (2) input information for multi-modality RSP reconstruction, (3) gold-standard calibration against concurrent techniques, (4) tracking motion and (5) pre-treatment positioning. However, no end-to-end characterization of the image quality (signal-to-noise ratio and spatial resolution, blurring uncertainty) against the dose has been done. This work aims to establish a model relating these characteristics and to describe their relationship with proton energy and object size. The imaging noise originates from two processes: the Coulomb scattering with the nucleus, producing a path deviation, and the energy loss straggling with electrons. The noise is found to increases with thickness crossed and, independently, decreases with decreasing energy. The scattering noise is dominant around high-gradient edge whereas the straggling noise is maximal in homogeneous regions. Image quality metrics are found to behave oppositely against energy: lower energy minimizes both the noise and the spatial resolution, with the optimal energy choice depending on the application and location in the imaged object. In conclusion, the model presented will help define an optimal usage of proton imaging to reach the promised application of this technology and establish a fair comparison with other imaging techniques.

Dosimetric accuracy of dynamic couch rotation during volumetric modulated arc therapy (DCR-VMAT) for primary brain tumours.

Radiotherapy treatment plans using dynamic couch rotation during volumetric modulated arc therapy (DCR-VMAT) reduce the dose to organs at risk (OARs) compared to coplanar VMAT, while maintaining the dose to the planning target volume (PTV). This paper seeks to validate this finding with measurements. DCR-VMAT treatment plans were produced for five patients with primary brain tumours and delivered using a commercial linear accelerator (linac). Dosimetric accuracy was assessed using point dose and radiochromic film measurements. Linac-recorded mechanical errors were assessed by extracting deviations from log files for multi-leaf collimator (MLC), couch, and gantry positions every 20 ms. Dose distributions, reconstructed from every fifth log file sample, were calculated and used to determine deviations from the treatment plans. Median (range) treatment delivery times were 125 s (123-133 s) for DCR-VMAT, compared to 78 s (64-130 s) for coplanar VMAT. Absolute point doses were 0.8% (0.6%-1.7%) higher than prediction. For coronal and sagittal films, respectively, 99.2% (96.7%-100%) and 98.1% (92.9%-99.0%) of pixels above a 20% low dose threshold reported gamma <1 for 3% and 3 mm criteria. Log file analysis showed similar gantry rotation root-mean-square error (RMSE) for VMAT and DCR-VMAT. Couch rotation RMSE for DCR-VMAT was 0.091° (0.086-0.102°). For delivered dose reconstructions, 100% of pixels above a 5% low dose threshold reported gamma <1 for 2% and 2 mm criteria in all cases. DCR-VMAT, for the primary brain tumour cases studied, can be delivered accurately using a commercial linac.

Recent developments in non-coplanar radiotherapy.

This paper gives an overview of recent developments in non-coplanar intensity modulated radiotherapy (IMRT) and volumetric modulated arc therapy (VMAT). Modern linear accelerators are capable of automating motion around multiple axes, allowing efficient delivery of highly non-coplanar radiotherapy techniques. Novel techniques developed for C-arm and non-standard linac geometries, methods of optimization, and clinical applications are reviewed. The additional degrees of freedom are shown to increase the therapeutic ratio, either through dose escalation to the target or dose reduction to functionally important organs at risk, by multiple research groups. Although significant work is still needed to translate these new non-coplanar radiotherapy techniques into the clinic, clinical implementation should be prioritized. Recent developments in non-coplanar radiotherapy demonstrate that it continues to have a place in modern cancer treatment.

PRaVDA: The first solid-state system for proton computed tomography.

PURPOSE: Proton CT is widely recognised as a beneficial alternative to conventional X-ray CT for treatment planning in proton beam radiotherapy. A novel proton CT imaging system, based entirely on solid-state detector technology, is presented. Compared to conventional scintillator-based calorimeters, positional sensitive detectors allow for multiple protons to be tracked per read out cycle, leading to a potential reduction in proton CT scan time. Design and characterisation of its components are discussed. An early proton CT image obtained with a fully solid-state imaging system is shown and accuracy (as defined in Section IV) in Relative Stopping Power to water (RSP) quantified. METHOD: A solid-state imaging system for proton CT, based on silicon strip detectors, has been developed by the PRaVDA collaboration. The system comprises a tracking system that infers individual proton trajectories through an imaging phantom, and a Range Telescope (RT) which records the corresponding residual energy (range) for each proton. A back-projection-then-filtering algorithm is used for CT reconstruction of an experimentally acquired proton CT scan. RESULTS: An initial experimental result for proton CT imaging with a fully solid-state system is shown for an imaging phantom, namely a 75 mm diameter PMMA sphere containing tissue substitute inserts, imaged with a passively-scattered 125 MeV beam. Accuracy in RSP is measured to be ⩽1.6% for all the inserts shown. CONCLUSIONS: A fully solid-state imaging system for proton CT has been shown capable of imaging a phantom with protons and successfully improving RSP accuracy. These promising results, together with system the capability to cope with high proton fluences (2×108 protons/s), suggests that this research platform could improve current standards in treatment planning for proton beam radiotherapy.

Mean heart dose variation over a course of breath-holding breast cancer radiotherapy.

OBJECTIVE: The purpose of the work was to estimate the dose received by the heart throughout a course of breath-holding breast radiotherapy. METHODS: 113 cone-beam CT (CBCT) scans were acquired for 20 patients treated within the HeartSpare 1A study, in which both an active breathing control (ABC) device and a voluntary breath-hold (VBH) method were used. Predicted mean heart doses were obtained from treatment plans. CBCT scans were imported into a treatment planning system, heart outlines defined, images registered to the CT planning scan and mean heart dose recorded. Two observers outlined two cases three times each to assess interobserver and intraobserver variation. RESULTS: There were no statistically significant differences between ABC and VBH heart dose data from CT planning scans, or in the CBCT-based estimates of heart dose, and no effect from the order of the breath-hold method. Variation in mean heart dose per fraction over the three imaged fractions was <6 cGy without setup correction, decreasing to 3.3 cGy with setup correction. If scaled to 15 fractions, all differences between predicted and estimated mean heart doses were <0.5 Gy and in 80% of cases, they were <0.25 Gy. CONCLUSION: Variation in mean heart dose was at an acceptable level over the duration of breath-holding radiotherapy and was well predicted by the planning system. Advances in knowledge: Mean heart dose was not adversely affected by fraction-to-fraction variations throughout a course of heart-sparing radiotherapy using two well-established breath-holding methods.

Non-coplanar trajectories to improve organ at risk sparing in volumetric modulated arc therapy for primary brain tumors.

BACKGROUND AND PURPOSE: To evaluate non-coplanar volumetric modulated arc radiotherapy (VMAT) trajectories for organ at risk (OAR) sparing in primary brain tumor radiotherapy. MATERIALS AND METHODS: Fifteen patients were planned using coplanar VMAT and compared against non-coplanar VMAT plans for three trajectory optimization techniques. A geometric heuristic technique (GH) combined beam scoring and Dijkstra's algorithm to minimize the importance-weighted sum of OAR volumes irradiated. Fluence optimization was used to perform a local search around coplanar and GH trajectories, producing fluence-based local search (FBLS) and FBLS+GH trajectories respectively. RESULTS: GH, FBLS, and FBLS+GH trajectories reduced doses to the contralateral globe, optic nerve, hippocampus, temporal lobe, and cochlea. However, FBLS increased dose to the ipsilateral lens, optic nerve and globe. Compared to GH, FBLS+GH increased dose to the ipsilateral temporal lobe and hippocampus, contralateral optics, and the brainstem and body. GH and FBLS+GH trajectories reduced bilateral hippocampi normal tissue complication probability (p=0.028 and p=0.043, respectively). All techniques reduced PTV conformity; GH and FBLS+GH trajectories reduced homogeneity but less so for FBLS+GH. CONCLUSIONS: The geometric heuristic technique best spared OARs and reduced normal tissue complication probability, however incorporating fluence information into non-coplanar trajectory optimization maintained PTV homogeneity.

A multicentre study of the evidence for customized margins in photon breast boost radiotherapy.

OBJECTIVE: To determine if subsets of patients may benefit from smaller or larger margins when using laser setup and bony anatomy verification of breast tumour bed (TB) boost radiotherapy (RT). METHODS: Verification imaging data acquired using cone-beam CT, megavoltage CT or two-dimensional kilovoltage imaging on 218 patients were used (1574 images). TB setup errors for laser-only setup (dlaser) and for bony anatomy verification (dbone) were determined using clips implanted into the TB as a gold standard for the TB position. Cases were grouped by centre-, patient- and treatment-related factors, including breast volume, TB position, seroma visibility and surgical technique. Systematic (Σ) and random (σ) TB setup errors were compared between groups, and TB planning target volume margins (MTB) were calculated. RESULTS: For the study population, Σlaser was between 2.8 and 3.4 mm, and Σbone was between 2.2 and 2.6 mm, respectively. Females with larger breasts (p = 0.03), easily visible seroma (p ≤ 0.02) and open surgical technique (p ≤ 0.04) had larger Σlaser. Σbone was larger for females with larger breasts (p = 0.02) and lateral tumours (p = 0.04). Females with medial tumours (p < 0.01) had smaller Σbone. CONCLUSION: If clips are not used, margins should be 8 and 10 mm for bony anatomy verification and laser setup, respectively. Individualization of TB margins may be considered based on breast volume, TB and seroma visibility. ADVANCES IN KNOWLEDGE: Setup accuracy using lasers and bony anatomy is influenced by patient and treatment factors. Some patients may benefit from clip-based image guidance more than others.

The UK HeartSpare Study (Stage IB): randomised comparison of a voluntary breath-hold technique and prone radiotherapy after breast conserving surgery.

PURPOSE: To compare mean heart and left anterior descending coronary artery (LAD) doses (NTDmean) and positional reproducibility in larger-breasted women receiving left breast radiotherapy using supine voluntary deep-inspiratory breath-hold (VBH) and free-breathing prone techniques. MATERIALS AND METHODS: Following surgery for early breast cancer, patients with estimated breast volumes >750 cm(3) underwent planning-CT scans in supine VBH and free-breathing prone positions. Radiotherapy treatment plans were prepared, and mean heart and LAD doses were calculated. Patients were randomised to receive one technique for fractions 1-7, before switching techniques for fractions 8-15 (40 Gy/15 fractions total). Daily electronic portal imaging and alternate-day cone-beam CT (CBCT) imaging were performed. The primary endpoint was the difference in mean LAD NTDmean between techniques. Population systematic (Σ) and random errors (σ) were estimated. Within-patient comparisons between techniques used Wilcoxon signed-rank tests. RESULTS: 34 patients were recruited, with complete dosimetric data available for 28. Mean heart and LAD NTDmean doses for VBH and prone treatments respectively were 0.4 and 0.7 (p<0.001) and 2.9 and 7.8 (p<0.001). Clip-based CBCT errors for VBH and prone respectively were ⩽3.0 mm and ⩽6.5 mm (Σ) and ⩽3.5 mm and ⩽5.4 mm (σ). CONCLUSIONS: In larger-breasted women, supine VBH provided superior cardiac sparing and reproducibility than a free-breathing prone position.

Voluntary breath-hold technique for reducing heart dose in left breast radiotherapy.

Breath-holding techniques reduce the amount of radiation received by cardiac structures during tangential-field left breast radiotherapy. With these techniques, patients hold their breath while radiotherapy is delivered, pushing the heart down and away from the radiotherapy field. Despite clear dosimetric benefits, these techniques are not yet in widespread use. One reason for this is that commercially available solutions require specialist equipment, necessitating not only significant capital investment, but often also incurring ongoing costs such as a need for daily disposable mouthpieces. The voluntary breath-hold technique described here does not require any additional specialist equipment. All breath-holding techniques require a surrogate to monitor breath-hold consistency and whether breath-hold is maintained. Voluntary breath-hold uses the distance moved by the anterior and lateral reference marks (tattoos) away from the treatment room lasers in breath-hold to monitor consistency at CT-planning and treatment setup. Light fields are then used to monitor breath-hold consistency prior to and during radiotherapy delivery.

4D ultrasound speckle tracking of intra-fraction prostate motion: a phantom-based comparison with x-ray fiducial tracking using CyberKnife.

This study investigates the use of a mechanically-swept 3D ultrasound (3D-US) probe for soft-tissue displacement monitoring during prostate irradiation, with emphasis on quantifying the accuracy relative to CyberKnife® x-ray fiducial tracking. An US phantom, implanted with x-ray fiducial markers was placed on a motion platform and translated in 3D using five real prostate motion traces acquired using the Calypso system. Motion traces were representative of all types of motion as classified by studying Calypso data for 22 patients. The phantom was imaged using a 3D swept linear-array probe (to mimic trans-perineal imaging) and, subsequently, the kV x-ray imaging system on CyberKnife. A 3D cross-correlation block-matching algorithm was used to track speckle in the ultrasound data. Fiducial and US data were each compared with known phantom displacement. Trans-perineal 3D-US imaging could track superior-inferior (SI) and anterior-posterior (AP) motion to ≤0.81 mm root-mean-square error (RMSE) at a 1.7 Hz volume rate. The maximum kV x-ray tracking RMSE was 0.74 mm, however the prostate motion was sampled at a significantly lower imaging rate (mean: 0.04 Hz). Initial elevational (right-left; RL) US displacement estimates showed reduced accuracy but could be improved (RMSE <2.0 mm) using a correlation threshold in the ultrasound tracking code to remove erroneous inter-volume displacement estimates. Mechanically-swept 3D-US can track the major components of intra-fraction prostate motion accurately but exhibits some limitations. The largest US RMSE was for elevational (RL) motion. For the AP and SI axes, accuracy was sub-millimetre. It may be feasible to track prostate motion in 2D only. 3D-US also has the potential to improve high tracking accuracy for all motion types. It would be advisable to use US in conjunction with a small (∼2.0 mm) centre-of-mass displacement threshold in which case it would be possible to take full advantage of the accuracy and high imaging rate capability.

Trajectory optimization for dynamic couch rotation during volumetric modulated arc radiotherapy.

Non-coplanar radiation beams are often used in three-dimensional conformal and intensity modulated radiotherapy to reduce dose to organs at risk (OAR) by geometric avoidance. In volumetric modulated arc radiotherapy (VMAT) non-coplanar geometries are generally achieved by applying patient couch rotations to single or multiple full or partial arcs. This paper presents a trajectory optimization method for a non-coplanar technique, dynamic couch rotation during VMAT (DCR­VMAT), which combines ray tracing with a graph search algorithm. Four clinical test cases (partial breast, brain, prostate only, and prostate and pelvic nodes) were used to evaluate the potential OAR sparing for trajectory-optimized DCR­VMAT plans, compared with standard coplanar VMAT. In each case, ray tracing was performed and a cost map reflecting the number of OAR voxels intersected for each potential source position was generated. The least-cost path through the cost map, corresponding to an optimal DCR­VMAT trajectory, was determined using Dijkstra's algorithm. Results show that trajectory optimization can reduce dose to specified OARs for plans otherwise comparable to conventional coplanar VMAT techniques. For the partial breast case, the mean heart dose was reduced by 53%. In the brain case, the maximum lens doses were reduced by 61% (left) and 77% (right) and the globes by 37% (left) and 40% (right). Bowel mean dose was reduced by 15% in the prostate only case. For the prostate and pelvic nodes case, the bowel V50 Gy and V60 Gy were reduced by 9% and 45% respectively. Future work will involve further development of the algorithm and assessment of its performance over a larger number of cases in site-specific cohorts.

Normal tissue complication probability (NTCP) parameters for breast fibrosis: pooled results from two randomised trials.

INTRODUCTION: The dose-volume effect of radiation therapy on breast tissue is poorly understood. We estimate NTCP parameters for breast fibrosis after external beam radiotherapy. MATERIALS AND METHODS: We pooled individual patient data of 5856 patients from 2 trials including whole breast irradiation followed with or without a boost. A two-compartment dose volume histogram model was used with boost volume as the first compartment and the remaining breast volume as second compartment. Results from START-pilot trial (n=1410) were used to test the predicted models. RESULTS: 26.8% patients in the Cambridge trial (5 years) and 20.7% patients in the EORTC trial (10 years) developed moderate-severe breast fibrosis. The best fit NTCP parameters were BEUD3(50)=136.4 Gy, γ50=0.9 and n=0.011 for the Niemierko model and BEUD3(50)=132 Gy, m=0.35 and n=0.012 for the Lyman Kutcher Burman model. The observed rates of fibrosis in the START-pilot trial agreed well with the predicted rates. CONCLUSIONS: This large multi-centre pooled study suggests that the effect of volume parameter is small and the maximum RT dose is the most important parameter to influence breast fibrosis. A small value of volume parameter 'n' does not fit with the hypothesis that breast tissue is a parallel organ. However, this may reflect limitations in our current scoring system of fibrosis.

The UK HeartSpare Study: randomised evaluation of voluntary deep-inspiratory breath-hold in women undergoing breast radiotherapy.

PURPOSE: To determine whether voluntary deep-inspiratory breath-hold (v_DIBH) and deep-inspiratory breath-hold with the active breathing coordinator™ (ABC_DIBH) in patients undergoing left breast radiotherapy are comparable in terms of normal-tissue sparing, positional reproducibility and feasibility of delivery. METHODS: Following surgery for early breast cancer, patients underwent planning-CT scans in v_DIBH and ABC_DIBH. Patients were randomised to receive one technique for fractions 1-7 and the second technique for fractions 8-15 (40 Gy/15 fractions total). Daily electronic portal imaging (EPI) was performed and matched to digitally-reconstructed radiographs. Cone-beam CT (CBCT) images were acquired for 6/15 fractions and matched to planning-CT data. Population systematic (Σ) and random errors (σ) were estimated. Heart, left-anterior-descending coronary artery, and lung doses were calculated. Patient comfort, radiographer satisfaction and scanning/treatment times were recorded. Within-patient comparisons between the two techniques used the paired t-test or Wilcoxon signed-rank test. RESULTS: Twenty-three patients were recruited. All completed treatment with both techniques. EPI-derived Σ were ≤ 1.8mm (v_DIBH) and ≤ 2.0mm (ABC_DIBH) and σ ≤ 2.5mm (v_DIBH) and ≤ 2.2mm (ABC_DIBH) (all p non-significant). CBCT-derived Σ were ≤ 3.9 mm (v_DIBH) and ≤ 4.9 mm (ABC_DIBH) and σ ≤ 4.1mm (v_DIBH) and ≤ 3.8mm (ABC_DIBH). There was no significant difference between techniques in terms of normal-tissue doses (all p non-significant). Patients and radiographers preferred v_DIBH (p=0.007, p=0.03, respectively). Scanning/treatment setup times were shorter for v_DIBH (p=0.02, p=0.04, respectively). CONCLUSIONS: v_DIBH and ABC_DIBH are comparable in terms of positional reproducibility and normal tissue sparing. v_DIBH is preferred by patients and radiographers, takes less time to deliver, and is cheaper than ABC_DIBH.

Towards real-time VMAT verification using a prototype, high-speed CMOS active pixel sensor.

This work investigates the feasibility of using a prototype complementary metal oxide semiconductor active pixel sensor (CMOS APS) for real-time verification of volumetric modulated arc therapy (VMAT) treatment. The prototype CMOS APS used region of interest read out on the chip to allow fast imaging of up to 403.6 frames per second (f/s). The sensor was made larger (5.4 cm × 5.4 cm) using recent advances in photolithographic technique but retains fast imaging speed with the sensor's regional read out. There is a paradigm shift in radiotherapy treatment verification with the advent of advanced treatment techniques such as VMAT. This work has demonstrated that the APS can track multi leaf collimator (MLC) leaves moving at 18 mm s(-1) with an automatic edge tracking algorithm at accuracy better than 1.0 mm even at the fastest imaging speed. Evaluation of the measured fluence distribution for an example VMAT delivery sampled at 50.4 f/s was shown to agree well with the planned fluence distribution, with an average gamma pass rate of 96% at 3%/3 mm. The MLC leaves motion and linac pulse rate variation delivered throughout the VMAT treatment can also be measured. The results demonstrate the potential of CMOS APS technology as a real-time radiotherapy dosimeter for delivery of complex treatments such as VMAT.

Adaptive breast radiation therapy using modeling of tissue mechanics: a breast tissue segmentation study.

PURPOSE: To validate and compare the accuracy of breast tissue segmentation methods applied to computed tomography (CT) scans used for radiation therapy planning and to study the effect of tissue distribution on the segmentation accuracy for the purpose of developing models for use in adaptive breast radiation therapy. METHODS AND MATERIALS: Twenty-four patients receiving postlumpectomy radiation therapy for breast cancer underwent CT imaging in prone and supine positions. The whole-breast clinical target volume was outlined. Clinical target volumes were segmented into fibroglandular and fatty tissue using the following algorithms: physical density thresholding; interactive thresholding; fuzzy c-means with 3 classes (FCM3) and 4 classes (FCM4); and k-means. The segmentation algorithms were evaluated in 2 stages: first, an approach based on the assumption that the breast composition should be the same in both prone and supine position; and second, comparison of segmentation with tissue outlines from 3 experts using the Dice similarity coefficient (DSC). Breast datasets were grouped into nonsparse and sparse fibroglandular tissue distributions according to expert assessment and used to assess the accuracy of the segmentation methods and the agreement between experts. RESULTS: Prone and supine breast composition analysis showed differences between the methods. Validation against expert outlines found significant differences (P<.001) between FCM3 and FCM4. Fuzzy c-means with 3 classes generated segmentation results (mean DSC = 0.70) closest to the experts' outlines. There was good agreement (mean DSC = 0.85) among experts for breast tissue outlining. Segmentation accuracy and expert agreement was significantly higher (P<.005) in the nonsparse group than in the sparse group. CONCLUSIONS: The FCM3 gave the most accurate segmentation of breast tissues on CT data and could therefore be used in adaptive radiation therapy-based on tissue modeling. Breast tissue segmentation methods should be used with caution in patients with sparse fibroglandular tissue distribution.

Cone beam computed tomography number errors and consequences for radiotherapy planning: an investigation of correction methods.

PURPOSE: The potential of keV cone beam computed tomography (CBCT) for guiding adaptive replanning is well-known. There are impediments to this, one being CBCT number accuracy. The purpose of this study was to investigate CBCT number correction methods and the affect of residual inaccuracies on dose deposition. Four different correction strategies were applied to the same patient data to compare performance and the sophistication of correction-method needed for acceptable dose errors. METHODS AND MATERIALS: Planning CT and CBCT reconstructions were used for 12 patients (6 brain, 3 prostate, and 3 bladder cancer patients). All patients were treated using Elekta linear accelerators and XVI imaging systems. Two of the CBCT number correction methods investigated were based on an algorithm previously proposed by the authors but only previously applied to phantoms. Two further methods, based on an approach previously suggested in the research literature, were also examined. Dose calculations were performed using scans of a "worst" subset of patients using the Pinnacle³ version 9.0 treatment planning system and the patients' clinical plans. RESULTS: All mean errors in CBCT number were <50 HU, and all correction methods performed well or adequately in dose calculations. The worst single dose discrepancy identified for any of the examined methods or patients was 3.0%. Mean errors in the doses to treatment volumes or organs at risk were negatively correlated with the mean error in CT number. That is, a mean CT number that was too large, averaged over the entire CBCT volume, implied an underdosing in a volume-of-interest and vice versa. CONCLUSIONS: Results suggest that (1) the correction of CBCT numbers to within a mean error of 50 HU in the scan volume provides acceptable discrepancies in dose (<3%) and (2) this is achievable with even quite unsophisticated correction methods.