Search

BackgroundIn cone-beam computed tomography (CBCT)-guided radiotherapy, off-by-one vertebral-body misalignments are rare but serious errors that lead to wrong-site treatments.PurposeAn automatic error detection algorithm was developed that uses a three-branch convolutional neural network error detection model (EDM) to detect off-by-one vertebral-body misalignments using planning computed tomography (CT) images and setup CBCT images.MethodsAlgorithm training and test data consisted of planning CTs and CBCTs from 480 patients undergoing radiotherapy treatment in the thoracic and abdominal regions at two radiotherapy clinics. The clinically applied registration was used to derive true-negative (no error) data. The setup and planning images were then misaligned by one vertebral-body in both the superior and inferior directions, simulating the most likely misalignment scenarios. For each of the aligned and misaligned 3D image pairs, 2D slice pairs were automatically extracted in each anatomical plane about a point within the vertebral column. The three slice pairs obtained were then inputted to the EDM that returned a probability of vertebral misalignment. One model (EDM1 ) was trained solely on data from institution 1. EDM1 was further trained using a lower learning rate on a dataset from institution 2 to produce a fine-tuned model, EDM2 . Another model, EDM3 , was trained from scratch using a training dataset composed of data from both institutions. These three models were validated on a randomly selected and unseen dataset composed of images from both institutions, for a total of 303 image pairs. The model performances were quantified using a receiver operating characteristic analysis. Due to the rarity of vertebral-body misalignments in the clinic, a minimum threshold value yielding a specificity of at least 99% was selected. Using this threshold, the sensitivity was calculated for each model, on each institution's test set separately.ResultsWhen applied to the combined test set, EDM1 , EDM2 , and EDM3 resulted in an area under curve of 99.5%, 99.4%, and 99.5%, respectively. EDM1 achieved a sensitivity of 96% and 88% on Institution 1 and Institution 2 test set, respectively. EDM2 obtained a sensitivity of 95% on each institution's test set. EDM3 achieved a sensitivity of 95% and 88% on Institution 1 and Institution 2 test set, respectively.ConclusionThe proposed algorithm demonstrated accuracy in identifying off-by-one vertebral-body misalignments in CBCT-guided radiotherapy that was sufficiently high to allow for practical implementation. It was found that fine-tuning the model on a multi-facility dataset can further enhance the generalizability of the algorithm.

PurposeAccurate and robust auto-segmentation of highly deformable organs (HDOs), for example, stomach or bowel, remains an outstanding problem due to these organs' frequent and large anatomical variations. Yet, time-consuming manual segmentation of these organs presents a particular challenge to time-limited modern radiotherapy techniques such as on-line adaptive radiotherapy and high-dose-rate brachytherapy. We propose a machine-assisted interpolation (MAI) that uses prior information in the form of sparse manual delineations to facilitate rapid, accurate segmentation of the stomach from low field magnetic resonance images (MRI) and the bowel from computed tomography (CT) images.MethodsStomach MR images from 116 patients undergoing 0.35T MRI-guided abdominal radiotherapy and bowel CT images from 120 patients undergoing high dose rate pelvic brachytherapy treatment were collected. For each patient volume, the manual delineation of the HDO was extracted from every 8th slice. These manually drawn contours were first interpolated to obtain an initial estimate of the HDO contour. A two-channel 64 × 64 pixel patch-based convolutional neural network (CNN) was trained to localize the position of the organ's boundary on each slice within a five-pixel wide road using the image and interpolated contour estimate. This boundary prediction was then input, in conjunction with the image, to an organ closing CNN which output the final organ segmentation. A Dense-UNet architecture was used for both networks. The MAI algorithm was separately trained for the stomach segmentation and the bowel segmentation. Algorithm performance was compared against linear interpolation (LI) alone and against fully automated segmentation (FAS) using a Dense-UNet trained on the same datasets. The Dice Similarity Coefficient (DSC) and mean surface distance (MSD) metrics were used to compare the predictions from the three methods. Statistically significance was tested using Student's t test.ResultsFor the stomach segmentation, the mean DSC from MAI (0.91 ± 0.02) was 5.0% and 10.0% higher as compared to LI and FAS, respectively. The average MSD from MAI (0.77 ± 0.25 mm) was 0.54 and 3.19 mm lower compared to the two other methods. Only 7% of MAI stomach predictions resulted in a DSC

PurposeOn-board magnetic resonance imaging (MRI) greatly enhances real-time target tracking capability during radiotherapy treatments. However, multislice and volumetric MRI techniques are frame rate limited and introduce unacceptable latency between the target moving out of position and the beam being turned off. We present a technique to estimate continuous volumetric tissue motion using motion models built from a repeated acquisition of a stack of MR slices. Applications including multislice target visualization and out-of-slice motion estimation during MRI-guided radiotherapy are demonstrated.MethodsEight healthy volunteer studies were performed using a 0.35 T MRI-guided radiotherapy system. Images were acquired at three frames per second in an interleaved fashion across ten adjacent sagittal slice positions covering 4.5 cm using a balanced steady-state-free precession sequence. A previously published five-dimensional (5D) linear motion model used for MRI-guided radiotherapy gating was extended to include multiple slices. This model utilizes an external respiratory bellows signal recorded during imaging to simultaneously estimate motion across all imaged slices. For comparison to an image-based approach, the manifold learning technique local linear embedding (LLE) was used to derive a respiratory surrogate for motion modeling. Manifolds for every slice were aligned during LLE in a group-wise fashion, enabling motion estimation outside the current imaged slice using a motion model, a process which we denote as mSGA. Additionally, a method is developed to evaluate out-of-slice motion estimates. The multislice motion model was evaluated in a single slice with each newly acquired image using a leave-one-out approach. Model-generated gating decision accuracy and beam-on positive predictive value (PPV) are reported along with the median and 95th percentile distance between model and ground truth target centroids.ResultsThe average model gating decision accuracy and PPV across all volunteer studies was 93.7% and 92.8% using the 5D model, and 96.8% and 96.1% using the mSGA model, respectively. The median and 95th percentile distance between model and ground truth target centroids was 0.91 and 2.90 mm, respectively, using the 5D model and 0.58 and 1.49 mm using the mSGA model, averaged over all eight subjects. The mSGA motion model provided a statistically significant improvement across all evaluation metrics compared to the external surrogate-based 5D model.ConclusionThe proposed techniques for out-of-slice target motion estimation demonstrated accuracy likely sufficient for clinical use. Results indicate the mSGA model may provide higher accuracy, however, the external surrogate-based model allows for unbiased in vivo accuracy evaluation.

PurposeTo demonstrate the preliminary feasibility of a longitudinal diffusion magnetic resonance imaging (MRI) strategy for assessing patient response to radiotherapy at 0.35 T using an MRI-guided radiotherapy system (ViewRay).MethodsSix patients (three head and neck cancer, three sarcoma) who underwent fractionated radiotherapy were enrolled in this study. A 2D multislice spin echo single-shot echo planar imaging diffusion pulse sequence was implemented on the ViewRay system and tested in phantom studies. The same pulse sequence was used to acquire longitudinal diffusion data (every 2-5 fractions) on the six patients throughout the entire course of radiotherapy. The reproducibility of the apparent diffusion coefficient (ADC) measurements was assessed using reference regions and the temporal variations of the tumor ADC values were evaluated.ResultsIn diffusion phantom studies, the ADC values measured on the ViewRay system matched well with reference ADC values with

PurposeTo determine if and by how much the commercial 4DCT protocols under- and overestimate tumor breathing motion.Methods1D simulations were conducted that modeled a 16-slice CT scanner and tumors moving proportionally to breathing amplitude. External breathing surrogate traces of at least 5-min duration for 50 patients were used. Breathing trace amplitudes were converted to motion by relating the nominal tumor motion to the 90th percentile breathing amplitude, reflecting motion defined by the more recent 5DCT approach. Based on clinical low-pitch helical CT acquisition, the CT detector moved according to its velocity while the tumor moved according to the breathing trace. When the CT scanner overlapped the tumor, the overlapping slices were identified as having imaged the tumor. This process was repeated starting at successive 0.1 s time bin in the breathing trace until there was insufficient breathing trace to complete the simulation. The tumor size was subtracted from the distance between the most superior and inferior tumor positions to determine the measured tumor motion for that specific simulation. The effect of the scanning parameter variation was evaluated using two commercial 4DCT protocols with different pitch values. Because clinical 4DCT scan sessions would yield a single tumor motion displacement measurement for each patient, errors in the tumor motion measurement were considered systematic. The mean of largest 5% and smallest 5% of the measured motions was selected to identify over- and underdetermined motion amplitudes, respectively. The process was repeated for tumor motions of 1-4 cm in 1 cm increments and for tumor sizes of 1-4 cm in 1 cm increments.ResultsIn the examined patient cohort, simulation using pitch of 0.06 showed that 30% of the patients exhibited a 5% chance of mean breathing amplitude overestimations of 47%, while 30% showed a 5% chance of mean breathing amplitude underestimations of 36%; with a separate simulation using pitch of 0.1 showing, respectively, 37% overestimation and 61% underestimation.ConclusionsThe simulation indicates that commercial low-pitch helical 4DCT processes potentially yield large tumor motion measurement errors, both over- and underestimating the tumor motion.

PurposeTo demonstrate that a "5DCT" technique which utilizes fast helical acquisition yields the same respiratory-gated images as a commercial technique for regular, mechanically produced breathing cycles.MethodsRespiratory-gated images of an anesthetized, mechanically ventilated pig were generated using a Siemens low-pitch helical protocol and 5DCT for a range of breathing rates and amplitudes and with standard and low dose imaging protocols. 5DCT reconstructions were independently evaluated by measuring the distances between tissue positions predicted by a 5D motion model and those measured using deformable registration, as well by reconstructing the originally acquired scans. Discrepancies between the 5DCT and commercial reconstructions were measured using landmark correspondences.ResultsThe mean distance between model predicted tissue positions and deformably registered tissue positions over the nine datasets was 0.65 ± 0.28 mm. Reconstructions of the original scans were on average accurate to 0.78 ± 0.57 mm. Mean landmark displacement between the commercial and 5DCT images was 1.76 ± 1.25 mm while the maximum lung tissue motion over the breathing cycle had a mean value of 27.2 ± 4.6 mm. An image composed of the average of 30 deformably registered images acquired with a low dose protocol had 6 HU image noise (single standard deviation) in the heart versus 31 HU for the commercial images.ConclusionsAn end to end evaluation of the 5DCT technique was conducted through landmark based comparison to breathing gated images acquired with a commercial protocol under highly regular ventilation. The techniques were found to agree to within 2 mm for most respiratory phases and most points in the lung.

PurposeThe purpose of this work is to develop a cardiac-induced lung motion model to be integrated into an existing breathing motion model.MethodsThe authors' proposed cardiac-induced lung motion model represents the lung tissue's specific response to the subject's cardiac cycle. The model is mathematically defined as a product of a converging polynomial function h of the cardiac phase (c) and the maximum displacement y(X0) of each voxel (X0) among all the cardiac phases. The function h(c) was estimated from cardiac-gated MR imaging of ten healthy volunteers using an Akaike Information Criteria optimization algorithm. For each volunteer, a total of 24 short-axis and 18 radial planar views were acquired on a 1.5 T MR scanner during a series of 12-15 s breath-hold maneuvers. Each view contained 30 temporal frames of equal time-duration beginning with the end-diastolic cardiac phase. The frames in each of the planar views were resampled to create a set of three-dimensional (3D) anatomical volumes representing thoracic anatomy at different cardiac phases. A 3D multiresolution optical flow deformable image registration algorithm was used to quantify the difference in tissue position between the end-diastolic cardiac phase and the remaining cardiac phases. To account for image noise, voxel displacements whose maximum values were less than 0.3 mm, were excluded. In addition, the blood vessels were segmented and excluded in order to eliminate registration artifacts caused by blood-flow.ResultsThe average cardiac-induced lung motions for displacements greater than 0.3 mm were found to be 0.86 ± 0.74 and 0.97 ± 0.93 mm in the left and right lungs, respectively. The average model residual error for the ten healthy volunteers was found to be 0.29 ± 0.08 mm in the left lung and 0.38 ± 0.14 mm in the right lung for tissue displacements greater than 0.3 mm. The relative error decreased with increasing cardiac-induced lung tissue motion. While the relative error was > 60% for submillimeter cardiac-induced lung tissue motion, the relative error decreased to < 5% for cardiac-induced lung tissue motion that exceeded 10 mm in displacement.ConclusionsThe authors' studies implied that modeling and including cardiac-induced lung motion would improve breathing motion model accuracy for tissues with cardiac-induced motion greater than 0.3 mm.

PurposeTo characterize radiation therapy patient breathing patterns based on measured external surrogate information.MethodsBreathing surrogate data were collected during 4DCT from a cohort of 50 patients including 28 patients with lung cancer and 22 patients without lung cancer. A spirometer and an abdominal pneumatic bellows were used as the surrogates. The relationship between these measurements was assumed to be linear within a small phase difference. The signals were correlated and drift corrected using a previously published method to convert the signal into tidal volume. The airflow was calculated with a first order time derivative of the tidal volume using a window centered on the point of interest and with a window length equal to the CT gantry rotation period. The airflow was compared against the tidal volume to create ellipsoidal patterns that were binned into 25 ml × 25 ml∕s bins to determine the relative amount of time spent in each bin. To calculate the variability of the maximum inhalation tidal volume within a free-breathing scan timeframe, a metric based on percentile volume ratios was defined. The free breathing variability metric (κ) was defined as the ratio between extreme inhalation tidal volumes (defined as >93 tidal volume percentile of the measured tidal volume) and normal inhalation tidal volume (defined as >80 tidal volume percentile of the measured tidal volume).ResultsThere were three observed types of volume-flow curves, labeled Types 1, 2, and 3. Type 1 patients spent a greater duration of time during exhalation with κ = 1.37 ± 0.11. Type 2 patients had equal time duration spent during inhalation and exhalation with κ = 1.28 ± 0.09. The differences between the mean peak exhalation to peak inhalation tidal volume, breathing period, and the 85th tidal volume percentile for Type 1 and Type 2 patients were statistically significant at the 2% significance level. The difference between κ and the 98th tidal volume percentile for Type 1 and Type 2 patients was found to be statistically significant at the 1% significance level. Three patients did not display a breathing stability curve that could be classified as Type 1 or Type 2 due to chaotic breathing patterns. These patients were classified as Type 3 patients.ConclusionsBased on an observed volume-flow curve pattern, the cohort of 50 patients was divided into three categories called Type 1, Type 2, and Type 3. There were statistically significant differences in breathing characteristics between Type 1 and Type 2 patients. The use of volume-flow curves to classify patients has been demonstrated as a physiological characterization metric that has the potential to optimize gating windows in radiation therapy.

Purpose: To develop and evaluate a novel motion prediction method for magnetic resonance image (MRI)‐guided radiotherapy applications. This method, which we deem "image regression," predicts future tissue motion based on a weighted combination of previously observed motion states. Motion predictions are derived from a sliding window of recent motion states which are defined by a temporal sequence of images. A key advantage of this method compared to other motion prediction methods is that its computational complexity scales weekly with the number of spatial points predicted. Applications of gating latency reduction and improvement in deformable registration‐based target tracking are demonstrated. Methods: The image regression (IR) motion prediction method was developed and evaluated using 26.9 h of real‐time imaging acquired from eight healthy volunteers and 13 patients using a 0.35 T MRI‐guided radiotherapy system. Motion predictions were performed 0.25–0.33 s into the future using a weighted sum of previously observed motion states with image similarity‐derived weights. The set of previously observed motion states were continuously updated to incorporate the changes in breathing patterns. The accuracy of the predicted radiotherapy gating decision, beam‐on positive predictive value (PPV), and predicted vs ground‐truth target centroid position errors are reported. The IR technique was compared against no prediction, linear extrapolation, and an established autoregressive linear prediction algorithm. The usage of IR to initialize the deformable registration and enhance the target tracking was demonstrated in the healthy volunteer studies. Deformable registration with IR initialization was compared to the initialization performed by current clinical software: no initialization, previous image registration initialization and linear motion extrapolation initialization. Results: The average IR‐predicted radiation beam gating decision accuracy was 95.8%, with a PPV of 95.7%, and median and 95th percentile centroid position errors of 0.63 and 2.08 mm, respectively. Compared to the autoregressive linear prediction method, gating accuracy was 1.15% greater, PPV was 1.61% greater, and median and 95th percentile centroid distances were 0.21 and 0.23 mm smaller. The IR‐initialized registration on average converged within 0.50 mm of the ground‐truth position in fewer than 10 iterations whereas the next best initialization method required more than 25 iterations. Conclusions: Image regression motion prediction has the potential to reduce the gating latencies and improve the speed and accuracy of deformable registration‐based target tracking in MRI‐guided radiotherapy. [ABSTRACT FROM AUTHOR]