Eugene Yip

Evaluation of a lung tumor autocontouring algorithm for intrafractional tumor tracking using low‐field MRI: A phantom study

PURPOSE The first aim of this study is to investigate the feasibility of online autocontouring of tumor in low field MR images (0.2 and 0.5 T) by means of a phantom and simulation study for tumor-tracking in linac-MR systems. The second aim of this study is to develop an MR compatible, lung tumor motion phantom. METHODS An autocontouring algorithm was developed to determine both the position and shape of a lung tumor from each intra fractional MR image. To initiate the algorithm, an expert user contours the tumor and its maximum anticipated range of motion (herein termed the Background) using pretreatment scan data. During treatment, the algorithm processes each intrafractional MR image and automatically contours the tumor. To evaluate this algorithm, the authors built a phantom that replicates the low field contrast parameters (proton density, T(1), T(2)) of lung tumors and healthy lung parenchyma. This phantom allows simulation of MR images with the expected lung tumor CNR at 0.2 and 0.5 T by using a single 3 T scanner. Dynamic bSSFP images (approximately 4 images per second) are acquired while the phantom undergoes a series of preprogrammed motions based on patient lung tumor motion data. These images are autocontoured off-line using our algorithm. The fidelity of autocontouring is assessed by comparing autocontoured tumor shape and its centroid position to the actual tumor shape and its position. RESULTS The algorithm successfully contoured the shape of a moving tumor model from dynamic MR images acquired every 275 ms. Dices coefficients of > 0.96 and > 0.93 are achieved in 0.5 and 0.2 T equivalent images, respectively. Also, the algorithm tracked tumor position during dynamic studies, with root mean squared error (RMSE) values of < 0.55 and < 0.92 mm for 0.5 and 0.2 T equivalent images, respectively. Autocontouring speed is approximately 5 ms for each image. CONCLUSIONS Dices coefficients of > 0.96 and > 0.93 are achieved between autocontoured and real tumor shapes, and the position of a tumor can be tracked with RMSE values of < 0.55 and < 0.92 mm in 0.5 and 0.2 T equivalent images, respectively. These results demonstrate the feasibility of lung tumor autocontouring in low field MR images, and, by extension, intrafractional lung tumor tracking with our laboratorys linac-MR system.

Neural-network based autocontouring algorithm for intrafractional lung-tumor tracking using Linac-MR.

PURPOSE To develop a neural-network based autocontouring algorithm for intrafractional lung-tumor tracking using Linac-MR and evaluate its performance with phantom and in-vivo MR images. METHODS An autocontouring algorithm was developed to determine both the shape and position of a lung tumor from each intrafractional MR image. A pulse-coupled neural network was implemented in the algorithm for contrast improvement of the tumor region. Prior to treatment, to initiate the algorithm, an expert user needs to contour the tumor and its maximum anticipated range of motion in pretreatment MR images. During treatment, however, the algorithm processes each intrafractional MR image and automatically generates a tumor contour without further user input. The algorithm is designed to produce a tumor contour that is the most similar to the experts manual one. To evaluate the autocontouring algorithm in the authors Linac-MR environment which utilizes a 0.5 T MRI, a motion phantom and four lung cancer patients were imaged with 3 T MRI during normal breathing, and the image noise was degraded to reflect the image noise at 0.5 T. Each of the pseudo-0.5 T images was autocontoured using the authors algorithm. In each test image, the Dice similarity index (DSI) and Hausdorff distance (HD) between the experts manual contour and the algorithm generated contour were calculated, and their centroid positions were compared (Δd centroid). RESULTS The algorithm successfully contoured the shape of a moving tumor from dynamic MR images acquired every 275 ms. From the phantom study, mean DSI of 0.95-0.96, mean HD of 2.61-2.82 mm, and mean Δd centroid of 0.68-0.93 mm were achieved. From the in-vivo study, the authors algorithm achieved mean DSI of 0.87-0.92, mean HD of 3.12-4.35 mm, as well as Δd centroid of 1.03-1.35 mm. Autocontouring speed was less than 20 ms for each image. CONCLUSIONS The authors have developed and evaluated a lung tumor autocontouring algorithm for intrafractional tumor tracking using Linac-MR. The autocontouring performance in the Linac-MR environment was evaluated using phantom and in-vivo MR images. From the in-vivo study, the authors algorithm achieved 87%-92% of contouring agreement and centroid tracking accuracy of 1.03-1.35 mm. These results demonstrate the feasibility of lung tumor autocontouring in the authors laboratorys Linac-MR environment.

Prior data assisted compressed sensing: A novel MR imaging strategy for real time tracking of lung tumors

PURPOSE Hybrid radiotherapy-MRI devices promise real time tracking of moving tumors to focus the radiation portals to the tumor during irradiation. This approach will benefit from the increased temporal resolution of MRIs data acquisition and reconstruction. In this work, the authors propose a novel spatial-temporal compressed sensing (CS) imaging strategy for the real time MRI--prior data assisted compressed sensing (PDACS), which aims to improve the image quality of the conventional CS without significantly increasing reconstruction times. METHODS Conventional 2D CS requires a random sampling of partial k-space data, as well as an iterative reconstruction that simultaneously enforces the images sparsity in a transform domain as well as maintains the fidelity to the acquired k-space. PDACS method requires the additional acquisition of the prior data, and for reconstruction, it additionally enforces fidelity to the prior k-space domain similar to viewsharing. In this work, the authors evaluated the proposed PDACS method by comparing its results to those obtained from the 2D CS and viewsharing methods when performed individually. All three methods are used to reconstruct images from lung cancer patients whose tumors move and who are likely to benefit from lung tumor tracking. The patients are scanned, using a 3T MRI, under free breathing using the fully sampled k-space with 2D dynamic bSSFP sequence in a sagittal plane containing lung tumor. These images form a reference set for the evaluation of the partial k-space methods. To create partial k-space, the fully sampled k-space is retrospectively undersampled to obtain a range of acquisition acceleration factors, and reconstructed with 2D-CS, PDACS, and viewshare methods. For evaluation, metrics assessing global image artifacts as well as tumor contour shape fidelity are determined from the reconstructed images. These analyses are performed both for the original 3T images and those at a simulated 0.5T equivalent noise level. RESULTS In the 3.0T images, the PDACS strategy is shown to give superior results compared to viewshare and conventional 2D CS using all metrics. The 2D-CS tends to perform better than viewshare at the low acceleration factors, while the opposite is true at the high acceleration factors. At simulated 0.5T images, PDACS method performs only marginally better than the viewsharing method, both of which are superior compared to 2D CS. The PDACS image reconstruction time (0.3 s/image) is similar to that of the conventional 2D CS. CONCLUSIONS The PDACS method can potentially improve the real time tracking of moving tumors by significantly increasing MRIs data acquisition speeds. In 3T images, the PDACS method does provide a benefit over the other two methods in terms of both the overall image quality and the ability to accurately and automatically contour the tumor shape. MRIs data acquisition may be accelerated using the simpler viewsharing strategy at the lower, 0.5T magnetic field, as the marginal benefit of the PDACS method may not justify its additional reconstruction times.

Improved lung tumor autocontouring algorithm for intrafractional tumor tracking using 0.5 T linac-MR

To add an intelligent parameter optimization capability to our autocontouring algorithm, and evaluate its performance using in-vivo data. Methods An autocontouring algorithm for intrafractional lung-tumor tracking using linac-MR was previously developed based on pulse-coupled neural networks. The algorithms contouring performance is dependent on eight parameters (including four integer parameters). Previously, the parameters were optimized using a time-consuming, exhaustive method. To avoid this inefficiency, adaptive particle swarm optimization (APSO) was adopted in this study, which is a stochastic, non-gradient based optimization algorithm that can handle integer variables. For this study, six non-small cell lung cancer patients were imaged with 3T MRI at ~4 frames per second (2D sagittal plane, free breathing). For each patient, an expert delineated a gold standard contour (ROIstd) of the lung tumor in 130 consecutive images. The first 30 ROIstd were used for parameter optimization, and the rest 100 ROIstd were used to validate autocontours (ROIauto). In each image, Dice similarity index, Hausdorff distance, and centroid position difference (Δdcentroid) were calculated between ROIstd and ROIauto to measure their similarity. Results & Conclusion An efficient, fully automatic parameter optimization was added to our autocontouring algorithm. Using the six patients data, approximately 1/24 time reduction was achieved in parameter optimization (63–125 hrs to 2–4 hrs per patient), while maintaining the same or slightly improved performance.

Sliding window prior data assisted compressed sensing for MRI tracking of lung tumors

Purpose: Hybrid magnetic resonance imaging and radiation therapy devices are capable of imaging in real‐time to track intrafractional lung tumor motion during radiotherapy. Highly accelerated magnetic resonance (MR) imaging methods can potentially reduce system delay time and/or improves imaging spatial resolution, and provide flexibility in imaging parameters. Prior Data Assisted Compressed Sensing (PDACS) has previously been proposed as an acceleration method that combines the advantages of 2D compressed sensing and the KEYHOLE view‐sharing technique. However, as PDACS relies on prior data acquired at the beginning of a dynamic imaging sequence, decline in image quality occurs for longer duration scans due to drifts in MR signal. Novel sliding window‐based techniques for refreshing prior data are proposed as a solution to this problem. Methods: MR acceleration is performed by retrospective removal of data from the fully sampled sets. Six patients with lung tumors are scanned with a clinical 3 T MRI using a balanced steady‐state free precession (bSSFP) sequence for 3 min at approximately 4 frames per second, for a total of 650 dynamics. A series of distinct pseudo‐random patterns of partial k‐space acquisition is generated such that, when combined with other dynamics within a sliding window of 100 dynamics, covers the entire k‐space. The prior data in the sliding window are continuously refreshed to reduce the impact of MR signal drifts. We intended to demonstrate two different ways to utilize the sliding window data: a simple averaging method and a navigator‐based method. These two sliding window methods are quantitatively compared against the original PDACS method using three metrics: artifact power, centroid displacement error, and Dices coefficient. The study is repeated with pseudo 0.5 T images by adding complex, normally distributed noise with a standard deviation that reduces image SNR, relative to original 3 T images, by a factor of 6. Results: Without sliding window implemented, PDACS‐reconstructed dynamic datasets showed progressive increases in image artifact power as the 3 min scan progresses. With sliding windows implemented, this increase in artifact power is eliminated. Near the end of a 3 min scan at 3 T SNR and 5× acceleration, implementation of an averaging (navigator) sliding window method improves our metrics by the following ways: artifact power decreases from 0.065 without sliding window to 0.030 (0.031), centroid error decreases from 2.64 to 1.41 mm (1.28 mm), and Dice coefficient agreement increases from 0.860 to 0.912 (0.915). At pseudo 0.5 T SNR, the improvements in metrics are as follows: artifact power decreases from 0.110 without sliding window to 0.0897 (0.0985), centroid error decreases from 2.92 mm to 1.36 mm (1.32 mm), and Dice coefficient agreements increases from 0.851 to 0.894 (0.896). Conclusions: In this work we demonstrated the negative impact of slow changes in MR signal for longer duration PDACS dynamic scans, namely increases in image artifact power and reductions of tumor tracking accuracy. We have also demonstrated sliding window implementations (i.e., refreshing of prior data) of PDACS are effective solutions to this problem at both 3 T and simulated 0.5 T bSSFP images.

TU-H-BRA-09: Relationship Between B0 and the Contrast-To-Noise Ratio (CNR) of Tumour to Background for MRI/Radiotherapy Hybrids

PURPOSE To investigate the relationship in MRI between B0 and the contrast-to-noise ratio (CNR) of various tumour/normal tissue pairs. This study is motivated by the current interest in MRI/radiotherapy hybrids, for which multiple magnetic field strengths have been proposed. CNR is the single most important parameter governing the ability of a system to identify a tumour in real time for treatment guidance. The MRI community has long since recognized that the SNR of a well-designed MR system is roughly proportional to B0 , the polarizing magnetic field. However, the CNR between two tissues is much more complicated - dependent not only on this signal behavior, but also on the different relaxation properties of the tissues. METHODS Experimentally-based models of B0 -dependant relaxation for various tumour and normal tissues from the literature were used in conjunction with signal equations for MR sequences suitable for rapid realtime imaging to develop field-dependent predictions for CNR. These CNR models were developed for liver, lung, breast, glioma, and kidney tumours for spoiled-gradient echo (SGE) and balanced steady-state free precession (bSSFP) sequences. RESULTS In all cases there was an improved CNR at lower fields compared to linear dependency. Further, in some tumour sites, the CNR at lower fields was found to be comparable to, or sometimes higher than those at higher fields (i.e. bSSFP CNR for glioma, kidney and liver tumours). CONCLUSION Due to the variation of tissue relaxation parameters with field, lower B0 fields have been shown to perform as well or better (in terms of CNR) than higher fields for some tumour sites. In other sites this effect was less pronounced. It is the complex relationship between CNR and B0 that leads to greater CNR at 0.5 T for certain tumour types studied here for fast imaging. B. Gino Fallone is a co-founder and CEO of MagnetTx Oncology Solutions (under discussions to license Alberta bi-planar linac MR for commercialization).

SU-G-JeP1-15: Sliding Window Prior Data Assisted Compressed Sensing for MRI Lung Tumor Tracking

PURPOSE Prior Data Assisted Compressed Sensing (PDACS) is a partial k-space acquisition and reconstruction method for mobile tumour (i.e. lung) tracking using on-line MRI in radiotherapy. PDACS partially relies on prior data acquired at the beginning of dynamic scans, and is therefore susceptible to artifacts in longer duration scan due to slow drifts in MR signal. A novel sliding window strategy is presented to mitigate this effect. METHODS MRI acceleration is simulated by retrospective removal of data from the fully sampled sets. Six lung cancer patients were scanned (clinical 3T MRI) using a balanced steady state free precession (bSSFP) sequence for 3 minutes at approximately 4 frames per second, for a total of 650 dynamics. PDACS acceleration is achieved by undersampling of k-space in a single pseudo-random pattern. Reconstruction iteratively minimizes the total variations while constraining the images to satisfy both the currently acquired data and the prior data in missing k-space. Our novel sliding window technique (SW-PDACS), uses a series of distinct pseudo-random under-sampling patterns of partial k-space - with the prior data drawn from a sliding window of the most recent data available. Under-sampled data, simulating 2 - 5x acceleration are reconstructed using PDACS and SW-PDACS. Three quantitative metrics: artifact power, centroid error and Dices coefficient are computed for comparison. RESULTS Quantitively metric values from all 6 patients are averaged in 3 bins, each containing approximately one minute of dynamic data. For the first minute bin, PDACS and SW-PDACS give comparable results. Progressive decline in image quality metrics in bins 2 and 3 are observed for PDACS. No decline in image quality is observed for SW-PDACS. CONCLUSION The novel approach presented (SW-PDACS) is a more robust for accelerating longer duration (>1 minute) dynamic MRI scans for tracking lung tumour motion using on-line MRI in radiotherapy. B.G. Fallone is a co-founder and CEO of MagnetTx Oncology Solutions (under discussions to license Alberta bi-planar linac MR for commercialization).

SU‐E‐J‐151: Evaluation of a Real Time Tumour Autocontouring Algorithm Using In‐Vivo Lung MR Images with Various Contrast to Noise Ratios

PURPOSE To quantitatively evaluate a lung tumour autocontouring algorithm using in-vivo lung cancer patient MR images with varying contrast to noise ratios (CNR) simulating images acquired at various MR field strengths. METHODS A non small cell lung cancer patient with posterior lung tumour is imaged (sagittal plane) in a 3T MRI using a dynamic bSSFP sequence (FOV: 40×40cm2 , voxel size: 3.1×3.1×20mm3 , TE = 1.1ms. TR = 2.2ms, 275ms per image) under free breathing for approximately 3 minutes (650 images). Gaussian random noise is added to the 3T images to approximately simulate the equivalent CNR in images acquired at 1.5T, 1.0T, 0.5T, 0.3T and 0.2T. The moving tumour in all 3T images is contoured by a physician for reference. The first 20 of these manual contours are used for the parameters optimization of auto-contouring algorithm. The automatic contours from the remaining images are quantitatively compared with the physicians contours using the centroids displacement and the Dices coefficient (DC). RESULTS The oncologists contours of the 3T images show a maximum S-I motion of 26mm. Compared to the oncologists contours, automatic contours have an average centroid displacement of 1.37mm, and an average DC of 0.881. The autocontouring algorithms performance with images in the range of 1.5T to 0.5T equivalent CNRs is similar to that of the 3T data. However, for the lowest CNR datasets (0.2, 0.3T) an increase in centroid displacement and decrease in DC is observed, with mean displacements of 1.56mm, 1.71mm and DCs of 0.870, 0.836 for the 0.3T and 0.2T dataset, respectivelyConclusions: With in-vivo MR images, the autocontouring algorithm generated lung tumour contours similar to ones drawn by a physician (DC 〉 0.83). In this patient, additional CNR from 〉0.5T MRIs does not provide statistically significant improvement in the accuracy of our autocontouring software. E.Yip is supported by the Canadian Institutes of Health Research as well as Alberta Innovates - Health Solutions.

SU‐E‐J‐83: Feasiblity of Real Time Tumour Tracking in Low Field MRI ‐ A Phantom Study

Purpose: Real time adaptive radiotherapy of lungtumours is a promising application of the proposed hybrid Linac‐MR system. Several linac‐MR designs use low magnetic field strength (0.2T and 0.5T) imaging. Rapid real time imaging is potentially challenging due to the limited available contrast to noise ratio (CNR), especially at low MR fields. The goal of this phantom study is to quantitatively evaluate our in‐house tumour tracking algorithm with images of a moving lungtumour model with CNR equivalent to those obtained at 0.2T and 0.5T.Methods: A chest phantom with a moving lung compartment capable of 1D programmable motion is built. The lung compartment is loaded with mixtures containing MnCl2 and CuSO4 that simulates lung tumour/ healthy lungtissue by mimicking their relaxation properties at 0.2Tand 0.5T in the available 3T scanner. CNR of the acquired images is scaled down to 0.2T and 0.5T by addition of Gaussian noise. Dynamic bSSFP images (4 frames/s) are acquired with the lung compartment undergoing a series of pre‐programmed motion pattern based on patient data. An optical encoder is used to provide an independent reference measurement of phantom position while the lung compartment undergoes motion. In‐house automatic tumour tracking software is used to contour the tumour off‐line. The automatic contours are compared against user defined contours by evaluation of the centroid position error, and the Dice coefficient.Results: The average RMS errors of the contour centroids are 0.9 at both CNR levels Conclusions: Our auto‐contouring algorithm is able to accurately track a moving tumour with the limited CNR available from real time, low field MR sequences.

SU-GG-J-11: Contrast to Noise Ratio Measurements for Real Time MR Lung Tumour Imaging Sequences at Lower Fields - A Phantom Study

Purpose:Magnetic Resonance Imaging(MRI) has been proposed for real‐time image guidance during lung radiotherapy treatment. Dynamic lungMR studies in the literature have demonstrated the feasibility of real‐time tumour tracking at 1.5T scanners. Lower field magnets offer several advantages over high field magnets, but has lower signal to noise ratio and a different contrast environment due to field strength dependences in T1 and T2*. The purpose of this study was to determine the expected contrast to noise ratio (CNR) at 0.2T for several rapid MRsequences by performing experiments in a 3T MRI.Method and Materials:Lungtumour is simulated by loading solution containing CuSO4 and MnCl2 in a sphere. To simulate the lower relative proton densities (PD) of lung parenchyma, 2mm acrylic beads are uniformly suspended in gelatin. T1, T2 and T2* and relative PD are measured for the phantom in 3T and compared against their expected values at 0.2T from literature. For real time imaging, rapid gradient echo sequences (FLASH and balanced SSFP) are used to acquire images from 3–10 frames per second using acceleration techniques of halfscan and parallel acquisition. A dynamic noise scan is used to estimate noise and is adjusted to reflect the lower SNR at 0.2T. Measurements are repeated using a body coil and a 6 channel thoracic SENSE coil for parallel imaging.Results: The measured T1, T2 and T2* and relative PD of the phantom are similar to the values given in literature. For dynamic lungimages,CNR ranges from 9.4–31.7 for bSSFP and 4.0–13.8 for FLASH. In house auto‐contouring algorithm shows good quality contours of spherical tumours with CNR > 2.5. Conclusion: In this phantom study, dynamic lungimagingsequences are shown to provide sufficient tumour‐tissue CNR and temporal resolution for real time MRlungtumour tracking at 0.2T.