Fiducial Localization Research Articles

High-Accuracy Patient-to-Image Registration for the Facilitation of Image-Guided Robotic Microsurgery on the Head

Image-guided microsurgery requires accuracies an order of magnitude higher than today's navigation systems provide. A critical step toward the achievement of such low-error requirements is a highly accurate and verified patient-to-image registration. With the aim of reducing target registration error to a level that would facilitate the use of image-guided robotic microsurgery on the rigid anatomy of the head, we have developed a semiautomatic fiducial detection technique. Automatic force-controlled localization of fiducials on the patient is achieved through the implementation of a robotic-controlled tactile search within the head of a standard surgical screw. Precise detection of the corresponding fiducials in the image data is realized using an automated model-based matching algorithm on high-resolution, isometric cone beam CT images. Verification of the registration technique on phantoms demonstrated that through the elimination of user variability, clinically relevant target registration errors of approximately 0.1 mm could be achieved.

A Quantitative Assessment of the Accuracy and Reliability of O-Arm Images for Deep Brain Stimulation Surgery

Deep brain stimulation (DBS) surgery has an average accuracy of 2 to 3 mm (range, 0-6 mm). Intraoperative detection of track location may be useful in interpreting physiological results and thus limit the number of brain penetrations as well as decrease the incidence of reoperations. The O-arm has been used to identify the DBS lead position; however, early results have indicated a significant discrepancy with lead position on postoperative imaging. This prospective study was conducted to determine the accuracy and reliability of fiducial and track localization and to assess the accuracy of O-arm image-based registration. The computed tomography (CT) image was considered the gold standard, and so for this study, the locations of all objects on the O-arm image were compared with their CT location. Thirty-three DBS surgeries were performed using the O-arm to image each track with detailed analysis of fiducial and track localization accuracy. Twenty-one subsequent surgeries were performed using O-arm registration. Only the final lead position was assessed in these individuals. The measurement error of the system was 0.7 mm, with a maximum error of 1.9 mm. Twenty-two percent of the parallel tracks through the BenGun exceeded this error and demonstrated the ability of the O-arm to detect these skewed tracks. The accuracy of final lead position was 2.04 mm in procedures with registration based on an O-arm image. This was not significantly different from CT-based registration at 2.16 mm. The O-arm was able to detect skewed tracks and provide registration accuracy equivalent to a CT scan.

Traceable micro‐CT scaling accuracy phantom for applications requiring exact measurement of distances or volumes

Volumetric x-ray microcomputed tomography (CT) can be employed in a variety of quantitative research applications such as image-guided interventions or characterization of medical devices. To ensure the highest geometric fidelity of images for these applications, a phantom and image processing algorithm have been developed to calibrate the scaling accuracy of micro-CT scanners to a traceable standard and provide corrections to image voxel sizing. The calibration phantom contains six borosilicate beads whose separations have been measured to a traceable standard. An image processing algorithm compares the known separations of the beads to their separations in micro-CT images. A least-squares solution is used to determine linear scaling correction factors along each of the three scanner axes to minimize errors in the bead separations within the images by correcting the image voxel size. The correction factors were applied to images of a similar phantom with beads at different positions to evaluate the ability of the correction factors to reduce errors at points independent of the fiducial locations in the calibration phantom. The calibration phantom was used to evaluate the scaling accuracy of five different micro-CT scanners representing four different scanner models. In two of the five scanners evaluated, the correction factors significantly reduced the mean error in bead separations in the images from 0.17% to 0.05% and from 0.37% to 0.07% of the actual bead separations, respectively. Scanners yielding similar voxel sizes possessed comparable geometric errors after correction using the phantom. Although the magnitude of the corrections is small, such corrections can be important for demanding micro-CT applications. Even if no voxel size correction is required, the phantom provides an easily implemented method to verify the geometric fidelity of micro-CT scanners to a traceable standard of measurement.

Tissue characterization using a phantom to validate four‐dimensional tissue deformation

This project proposes using a real tissue phantom for 4D tissue deformation reconstruction (4DTDR) and 4D deformable image registration (DIR) validation, which allows for the complete verification of the motion path rather than limited end-point to end-point of motion. Three electro-magnetic-tracking (EMT) fiducials were implanted into fresh porcine liver that was subsequently animated in a clinically realistic phantom. The animation was previously shown to be similar to organ motion, including hysteresis, when driven using a real patient's breathing pattern. For this experiment, 4DCTs and EMT traces were acquired when the phantom was animated using both sinusoidal and recorded patient-breathing traces. Fiducial were masked prior to 4DTDR for reconstruction. The original 4DCT data (with fiducials) were sampled into 20 CT phase sets and fiducials' coordinates were recorded, resulting in time-resolved fiducial motion paths. Measured values of fiducial location were compared to EMT measured traces and the result calculated by 4DTDR. For the sinusoidal breathing trace, 95% of EMT measured locations were within 1.2 mm of the measured 4DCT motion path, allowing for repeatable accurate motion characterization. The 4DTDR traces matched 95% of the EMT trace within 1.6 mm. Using the more irregular (in amplitude and frequency) patient trace, 95% of the EMT trace points fitted both 4DCT and 4DTDR motion path within 4.5 mm. The average match of the 4DTDR estimation of the tissue hysteresis over all CT phases was 0.9 mm using a sinusoidal signal for animation and 1.0 mm using the patient trace. The real tissue phantom is a tool which can be used to accurately characterize tissue deformation, helping to validate or evaluate a DIR or 4DTDR algorithm over a complete motion path. The phantom is capable of validating, evaluating, and quantifying tissue hysteresis, thereby allowing for full motion path validation.

High-Accuracy Characterization of Ambulatory Holter Electrocardiogram Events

In this paper, three noise-robust high-accuracy methods aiming at the detection and delineation of the electrocardiogram (ECG) events (QRS complex, P-wave, T-wave) were developed. The ECG signal was initially appropriately preprocessed by application of a bandpass FIR filter and Discrete Wavelet Transform (DWT). The first detection-delineation method was the Walsh-Hadamard Transform (WHT). The WHT coefficients were divided into two groups and the signal was reconstructed using the second group coefficients. By this reconstruction, the values of first derivative of events are made stronger rather than the values of other parts of signal. In the second method, a feed forward artificial neural network was implemented to detect all events of the ECG signal. In the third method, the first derivative of signal was computed using a new signal smoothing algorithm with corresponding statistical properties. For decreasing False Positive (FP) errors associated with P-wave detection, a discriminating border was introduced as the post processing stage specified by three QRS parameters: the duration of a QRS complex, the time distance from the former and latter QRS complexes, and the potential difference from former QRS complex J-location and the latter QRS complex fiducial location. The proposed methods were applied to DAY general hospital high resolution holter data.

A Fuzzy Preference Relation Based Method for Face Recognition by Gabor Filters

In this paper we have applied Gabor filter for fiducial point localization. After obtaining the fiducial points the number of fiducial points are reduced using a distance formula. The distance of each of this fiducial point is then calculated by the distance formula and stored in the database of the system. The same methodology is also applied on the input face which is to be matched with the faces available in the database. Then a fuzzy preference relation matrix is obtained . the largest eigen value of this matrix is then determined by algebraic method or numerical method depending on the order of the matrix. To apply the numerical method which is more easier for large order matrices we have used the C programming of this method . Once the largest eigen value is determined the corresponding priority vector can easily be obtained from which we can easily match the input face with the database.

On the accuracy of localization achievable in fiducial-based stereoscopic image registration system using an electronic portal imaging device

Portal imaging using electronic portal imaging device (EPID) is a well-established image-guided radiation therapy (IGRT) technique for external beam radiation therapy. The aims of this study are threefold; (i) to assess the accuracy of isocentre localization in the fiducial-based stereoscopic image registration, (ii) to investigate the impact of errors in the beam collimation device on stereoscopic registration, and (iii) to evaluate the intra- and inter-observer variability in stereoscopic registration. Portal images of a ball bearing phantom were acquired and stereoscopic image registrations were performed based on a point centred in the ball bearing as the surrogate for registration. Experiments were replicated by applying intentional offsets in the beam collimation device to simulate collimation errors. The accuracy of fiducial markers localization was performed by repeating the experiment using three spherical lead shots implanted in a pelvic phantom. Portal images of pelvis phantom were given to four expert users to assess the inter-observer variability in performing registration. The isocentre localization accuracy tested using ball bearing phantom was within 0.3 mm. Gravity-induced systematic errors of beam collimation device by 2 mm resulted in positioning offsets of the order of 2 mm opposing the simulated errors. Relatively large inter-portal pair projection errors ranges from 1.3 mm to 1.8 mm were observed with simulated errors in the beam collimation device. The intra-user and inter-user variabilities were observed to be 0.8 and 0.4 mm respectively. Fiducial-based stereoscopic image registration using EPID is robust for IGRT procedure.

SU-E-J-02: Accuracy of Fiducial Marker Localization Using Axial and Helical CT, Cone-Beam CT and KV Imaging.

To determine localization errors of fiducial markers using axial (ACT) and helical CT (HCT), cone beam CT (CBCT) and kV imaging. A thorax phantom containing markers of various sizes (2.5, 5, 10, 20mm) was imaged using ACT, HCT, CBCT and kV imaging. The phantom was imaged with and without motion (15mm amplitude, 15 cycles/min). CT images were reconstructed at 0.625, 1.25, 2.5 and 5mm thickness. Marker location and length were measured using axial and coronal imaging. The measured marker size increased almost linearly with increased slice thickness used in CT reconstruction with ACT for the stationary phantom. In HCT, the marker size varied non-linearly with increase in slice thickness. Motion of the phantom induced further blurring and shifts of the center of the marker with different sizes. The 10 mm marker was elongated up to 13.5 mm and 16.2 mm in the ACT and HCT, respectively. The markers were elongated up to nearly 26 mm using axial and helical scanning modes by phantom motion. The displacement of maker center was more prominent in the HCT with a shift up to 2 mm due to slice thickness for stationary phantom. The marker center was displaced by up to 18 and 22 mm using ACT and HCT, respectively, which was even higher than 15 mm motion amplitude. KV imaging produced the sharpest marker image with the least difference between actual and measured marker sizes. Increase in slice thickness enlarged the apparent marker size and displaced the maker center in ACT and HCT. Motion led to further enlargement in the maker size and displacement of maker center that depended on the motion amplitude. These effects should be considered in CT-based image-guided radiation therapy to ensure accurate tumor localization and patient positioning with implanted markers.

Automated Fiducial Localization in CT Images Based on Surface Processing and Geometrical Prior Knowledge for Radiotherapy Applications

We propose a novel method for radio-opaque external marker localization in CT scans for infrared (IR) patient set-up in radiotherapy. Efforts were focused on the quantification of uncertainties in marker localization in the CT dataset as a function of algorithm performance. We implemented a 3-D approach to fiducial localization based on surface extraction and marker recognition according to geometrical prior knowledge. The algorithm parameters were optimized on a clinical CT dataset coming from 35 cranial and extra-cranial patients; the localization accuracy was benchmarked at variable image resolution versus laser tracker measurements. The applicability of conventional IR optical tracking systems for localizing external surrogates in daily patient set-up procedures was also investigated in 121 proton therapy treatment sessions. Our study shows that the implemented algorithm features surrogates localization with uncertainties lower than 0.3 mm and with a true positive rate of 90.1%, being this latter mainly influenced by fiducial homogeneity in the CT images. The reported clinical validation in proton therapy confirmed the submillimetric accuracy and the expected algorithm sensitivity. Geometrical prior knowledge allows judging the reliability of the extracted fiducial coordinates, ensuring the highest accuracy in patient set-up.

Quantifying ITV instabilities arising from 4DCT: a simulation study using patient data

Treatment planning for patients undergoing radiation therapy is often performed based on four-dimensional computed tomography (4DCT) when respiratory motion is present, as in lung cancer patients. 4DCT is used to define the internal target volume (ITV) that, ideally, incorporates all potential locations of the tumour. In this work, we use the locations of gold fiducial markers implanted in lung tumours of eight patients to represent tumour motion. These fiducial locations are used in a simulation of a four-slice CT scanner to generate the ITV for 10, 20 and 30 mm diameter model tumours. To demonstrate instabilities in the ITV definition based on 4DCT, the ITV calculation was repeated for the same patients for consecutive scan start times, staggered by 1 s. The volumetric difference in the ITV and the per cent of time that the ITV contains in the tumour are both evaluated. The ITV from a single patient was found to vary by 46%–127% for a tumour diameter of 10 mm. The ITV did not cover the entirety of the tumour 11%–74% of the time for a 10 mm tumour diameter.

Fiducial Optimization for Minimal Target Registration Error in Image-Guided Neurosurgery

This paper presents new methods for the optimal selection of anatomical landmarks and optimal placement of fiducial markers in image-guided neurosurgery. These methods allow the surgeon to optimally plan fiducial marker locations on routine diagnostic images before preoperative imaging and to intraoperatively select the set of fiducial markers and anatomical landmarks that minimize the expected target registration error (TRE). The optimization relies on a novel empirical simulation-based TRE estimation method built on actual fiducial localization error (FLE) data. Our methods take the guesswork out of the registration process and can reduce localization error without additional imaging and hardware. Our clinical experiments on five patients who underwent brain surgery with a navigation system show that optimizing one marker location and the anatomical landmarks configuration reduced the TRE. The average TRE values using the usual fiducials setup and using the suggested method were 4.7 mm and 3.2 mm, respectively. We observed a maximum improvement of 4 mm. Reducing the target registration error has the potential to support safer and more accurate minimally invasive neurosurgical procedures.

Accuracy of rapid prototype models for head and neck reconstruction

Rapid prototype (RP) models are used in craniofacial reconstructions; however, there are no standards or acceptable limits to ensure accuracy of the fabricated models. The purpose of this study was to assess the accuracy of RP models by validating the accuracy of SLA skull models with a coordinate measurement device. Stainless steel spheres were located on a dry cadaver skull as fiducial markers, scanned with Multi Detector Computer Tomography (MDCT), and interpreted with software for rapid prototyping. Seven stereolithographic (SLA) models were fabricated and measured with a coordinate measurement device. An Euler rotation transformation calculation was applied to standardize the coordinate system between the control and the models. A paired standard t test (α=.05) was used to compare fiducial marker locations on SLA models with the control. A significant difference was found between the control and each of the SLA models (P<.001) in the Z axis additive build. Significant dimensional differences were not consistently detected in the X and Y axes. Dimensional deviations fell within the size of the MDCT scans voxel dimensions. The greatest discrepancies of medical model fabrication correspond to the largest dimension of the orthotropic voxel volume of the MDCT scan, which is related to the slice thickness of the scan and the Z axis of the RP model. However, the absolute magnitude of the error was small, well within the generally accepted tolerance for patient treatment.

Avoidance of wrong-level thoracic spine surgery: intraoperative localization with preoperative percutaneous fiducial screw placement

The accurate intraoperative localization of the correct thoracic spine level remains a challenging problem in both open and minimally invasive spine surgery. The authors describe a technique of using preoperatively placed percutaneous fiducial screws to localize the area of interest in the thoracic spine, and they assess the safety and efficacy of the technique. To avoid wrong-level surgery in the thoracic spine, the authors preoperatively placed a percutaneous 5-mm fiducial screw at the level of intended surgery using CT guidance. Plain radiographs and CT images with reconstructed views can then be referenced in the operating room to verify the surgical level, and the fiducial screw is easily identified on intraoperative fluoroscopy. The authors compared a group of 26 patients who underwent preoperative (often outpatient) fiducial screw placement prior to open or minimally invasive thoracic spine surgery to a historical group of 26 patients who had intraoperative localization with fluoroscopy alone. In the treatment group of 26 patients, no complications related to fiducial screw placement occurred, and there was no incidence of wrong-level surgery. In comparison, there were no wrong-level surgeries in the historical cohort of 26 patients who underwent mini-open or open thoracic spine surgery without placement of a fiducial screw. However, the authors found that the intraoperative localization fluoroscopy time was greatly reduced when a fiducial screw localization technique was employed. The aforementioned technique for intraoperative localization is safe, efficient, and accurate for identifying the target level in thoracic spine exposures. The fiducial marker screw can be placed using CT guidance on an outpatient basis. There is a reduction in the amount of intraoperative fluoroscopy time needed for localization in the fiducial screw group.

SU-E-T-531: Comparison of Film and CBCT Based Fiducial Marker Localization for Prostate Patients

Purpose: This study is to compare film and cone-beam computed tomography (CBCT) based fiducial marker localization results for intensity modulated radiation therapy (IMRT) of prostate cancer patients. Methods: From 2005 to 2011, 290 prostate cancer patients with 3 intra-prostatic fiducial markers implanted were treated with IMRT boost following 4-field pelvis radiation. Orthogonal films for 241 patients and CBCT images for 49 patients were taken daily before treatment to localize the prostate glands. The estimated couch shifts were recorded and applied. Total of 4072 fractions were analyzed in this study. Results: The overall standard deviations in the left—right (LR), anterior—posterior (AP), and superior- inferior (SI) directions for film and CBCT based localization are 3.90, 4.65, 3.64 mm and 3.94, 4.27, 3.82 mm, respectively. After removing average shifts of each patient, the standard deviations in LR, AP, SI are 3.06, 3.64, 2.63 mm and 3.08, 3.29, 2.48 mm, respectively. The Pearson coefficients of correlation for the frequency distribution of the couch shifts based on films vs. CBCT are 0.99, 0.98, and 0.98 in the LR, AP and SI directions. Conclusions: The CBCT based fiducial marker localization achieves the equivalent accuracy with the orthogonal film based method. The 3–4 mm standard deviations for both film and CBCT groups represent the natural inter-treatment movements of prostates of this patient population. Any new localization methods should be tested against these standard deviations if the same planned target volume (PTV) margins are used.

SU‐E‐J‐150: Applicability of a Point Fiducial‐Based Model for the Prostatic Bed in Post‐Prostatectomy Radiation Therapy Patients: An Expedited Dosimetric Analysis Using New Functionality in Philipsˈ Pinnacle3 Treatment Planning Systemˈs Research Version

Purpose: The shape of the post‐prostatectomy volume, fiducial placement and isodose coverage is more complex than prostate radiation therapy; but to date, there have been no commercial tools to estimate the delivered dose in patients aligned with fiducials. A fiducial‐based algorithm to determine the dosimetric effects of fiducial localization has been implemented as ˈDelivered Dose Investigational Toolˈ (DiDIT), in Philipsˈ Pinnacle TPSˈ research version (Philips Medical Systems, Fitchburg, WI).Methods: 17 post‐prostatectomy patients were treated with SMLC IMRT, electromagnetic (Calypso Medical Technologies, Seattle, WA) tracking, 5mm CTV‐PTV expansions and 3mm tracking tolerances. The DiDIT tool implemented convolution of the static plan dose distribution with rigid motion (rotations and translations) calculated from the ∼150,000 fiducial tracking points, to generate estimates of the delivered dose. This analysis takes into account fiducial placement eccentricity, rotations, translations and the individual patient anatomy and treatment plan. Treatment adequacy was defined as the delivered minimum dose to the prostatic bed (CTV) &gt;= 98% of planned minimum dose. Tracking data from subsets of the treatment course (first 3, 5 and 10 fractions) were utilized to determine the ability to predict treatment adequacy.Results: Each patient had fiducial motion of less than 2mm. Each DiDIT evaluation took 5 – 15 minutes. Treatment was adequate in 11 of 17 patients, with the CTV having cold spots of 4 to 32%. DiDIT analysis of the initial fractions was not predictive of inadequate dose delivery. Patients with inadequate dosimetry seemed to have more eccentrically placed transponders and/or more rotation though no cut‐off values were noted Conclusions: The DiDIT tool was used to quickly model delivered dose from fiducial localization. Results show that even with continuous fiducial tracking, there may be residual errors in the localization of the prostatic bed. Further work comparing these results with daily volumetric imaging is warranted.Work sponsored by NIH R01 CA134541‐01.

3-D fiducial motion tracking using limited MV projections in arc therapy

In-treatment fiducial tracking has recently received attention as a method for improving treatment accuracy, dose conformity, and sparing of healthy tissue. 3-D fiducial localization in arc-radiotherapy remains challenging due to the motion of targets and the complexity of arc deliveries. We propose a novel statistical method for estimating 3-D fiducial motion using limited 2-D megavoltage (MV) projections. 3-D fiducial motion was estimated by a maximum a posteriori (MAP) approach to integrating information of fiducial projections with prior knowledge of target motion. To obtain the imaging geometries, short sequences of MV projections were selected in which fiducials were continuously visible. The MAP algorithm estimated the 3-D motion by maximizing the probability of displacement of fiducials in the sequences. Prior knowledge of target motion from a large statistical sample was built into the model to enhance the accuracy of estimation. In the case that a motion prior was unavailable, the algorithm can be simplified to the maximum likelihood (ML) approach. To compare tracking performance, a multiprojection geometric method was also presented by extending the typical two-project ion geometric estimation approach. The algorithms were evaluated using clinical prostate motion traces, and the performance was measured in quality indices and statistical evaluation. The results showed that the MAP method significantly outperforms the geometric method in all measures. In our simulations, the MAP method achieved an accuracy of less than 1 mm RMS error using only five continuous projections, whereas the geometric method required 15 projections to achieve a similar result. The approach presented can accurately estimate tumor motion using a limited number of continuous projections. The MAP motion estimation is superior to both the ML and geometric estimation methods.

SU-E-T-525: Evaluating the Tracking Performance of Non-Migrating Fiducial Markers for Lung Lesion Localization in CyberKnife Treatments

Purpose: When treating lung lesions using CyberKnife (Accuray Inc, Sunnyvale, CA), a common tool for patient alignment and target localization is the use of radio‐opaque fiducial markers in conjunction with stereoscopic kV imaging. This requires that there not be any variation in marker location throughout a treatment regimen. The high occurrence of fiducial marker migration (especially in lungs) has yielded the development of markers specifically designed to reduce migration. One such marker is the Gold Anchor (Radiadyne LLC, Houston, TX). This study evaluates the capability of the CyberKnife Radiostereoscopic Target Locating System (TLS) to track these implants. This is the only such investigation of these seeds used in the described capacity. Methods and materials: This study evaluates the GA200‐20 Gold Anchor, a 2.0 mm‐long marker that is designed to be injected through a needle, and upon contact with tissue, will collapse into an irregular shape. This irregular shape prevents the marker from migrating. Three markers are placed in an XLT Lung Phantom (CIRS Inc, Norfolk, VA); three being required by the TLS to correct for translational and rotational target motion. The phantom was then imaged using CT, and a treatment plan was designed. Following fiducial localization and subsequent table corrections, 500 cGy was homogeneously delivered to a small spherical target within the phantom, where orthogonal radiochromic film was placed. Results: Using clinically‐relevant imaging parameters, the Gold Anchor was accurately tracked by the TLS. On average, we were able to deliver the prescribed dose to the target volume with less than 0.5 mm targeting deviation, with 0.7mm being the highest deviation. Conclusions: This study shows that these markers are able to be tracked by the CyberKnife's TLS. This is a required initial step for future clinical studies where these markers are used to aid in the target tracking of lung lesions.

WE‐E‐BRC‐05: Voxel Based Four Dimensional Tissue Deformation Reconstruction (4DTDR) Validation Using a Real Tissue Phantom

Purpose: Calculation of four dimensional (4D) dose distributions requires the remapping of dose calculated on each available binned phases of the 4DCT onto a reference phase for summation. Deformable image registration (DIR) is usually used for this task. This project validates a type of new algorithm, using 4D tissue deformation reconstruction (4DTDR) from all 4DCT generated images (or projections) to confirm voxel to voxel accuracy for the complete motion path and future 4D dose calculations. Methods: Three electro‐magnetic‐tracking (EMT) fiducial were implanted into fresh porcine liver and animated in a clinically realistic phantom with highly reproducible hysteresis motion pattern previously reported on. After 4DCT acquisition, all data was copied and fiducial locations were masked in the copied set. All images were sent to the 4DTDR for reconstruction. The original 4DCT data (with fiducials) was sampled in 20 CT phase sets and fiducialsˈ coordinates were recorded, resulting in time‐resolved fiducial motion paths. Measured values of fiducial location were compared to that measured through EMT and those calculated by 4DTDR. Results: The performance of the 4DTDR was measured to have average accuracy of 1.86 mm over all 20 phases (4.54 mm max error) in time and location, calculating a motion path that correctly represents measured hysteresis. EMT measurement confirms 4DCT and 4DTDR values. Conclusion: Accuracy of deformation within tissue necessitates the understanding of motion accuracy over the whole motion path of deformation. Validation of the actual voxel path over the period adds a higher level of accuracy, leading away from typically quoted end‐to‐end accuracy to a definition that includes all motion locations at any moment or any location.

MO-D-BRB-04: EPID-Based 4D Patient Dose Verification for Lung SBRT Treatment

Purpose: To investigate the feasibility of a real-time EPID-based 4D dose verification method for adaptive SBRT lung treatment. Methods: Patients with fiducial markers were scanned with 4D-CT for SBRT planning. Before the treatment, an in-room 4D-CT scan (CT-on-rail) was obtained. Both the fiducial markers and MLC movements were tracked by continuously acquiring EPID images during the treatment. Instead of using the directly measured photon fluence, we reconstructed the incident fluence using the MLC apertures measured by the EPID and the MUs recorded by the R&V system. The delivered dose was then calculated by Monte Carlo simulations for every breathing phase using the in-room 4D-CT. The EPID-measured fiducial marker locations were correlated with the corresponding respiratory phases and the total dose was the sum of each breathing phase with a proper weighting factor after deformable registration. The accuracy of MLC aperture detection and fluence reconstruction was evaluated. A solid water phantom and a dynamic thorax phantom were used to validate the reconstructed 4D dose distributions. Clinical SBRT plans of 10 patients were selected to demonstrate the proposed method. Results: With patient heterogeneity, the field edges for 1×1 to 5×5 cm2 field sizes are still correctly defined (± 0.8 mm). The reconstructed doses agree well with measurements and TPS calculations (0.87%). The EPID measured tumor movements represent the pre-set movements. The dose calculation with detected tumor movements larger than the PTV margin showed poorer coverage to the PTV and the ITV. Conclusions: The proposed method provides a practical way to reconstruct the incident fluence from on-line EPID-measured transit images and R&V recorded MUs and to calculate the fractional 4D doses received by a SBRT patient. With superior geometric and dose calculation accuracy, this method will enable adaptive SBRT to verify and correct the fractional doses received by the patient.

SU‐C‐211‐04: Micro‐CT Geometric Accuracy Phantom for Improved Fiducial Localization in Image‐Guided Needle Positioning Systems

Purpose: To quantify and correct the geometric accuracy of an x‐ray microCT scanner to improve fiducial localization in image‐guided needle positioning systems. Methods: A geometric accuracy phantom was constructed by creating a one‐inch cube consisting of a 4×4×4 3D matrix of calibrated ¼” diameter spherical beads alternating between two materials. The phantom was imaged at both the isocenter and at the edge of the field of view of a micro‐CT scanner. All bead positions were measured in sets of five images and averaged. An automated algorithm evaluated the geometric accuracy of the scanner by comparing the known position of the beads in the phantom to their positions in the images. The algorithm used a least‐squares solution to determine scaling factors in the horizontal, vertical and axial directions of the images to correct the geometric inaccuracies. The accuracy of localizing a fiducial marker mounted onto a needle positioning system was determined with and without using the calculated scaling factors. Results: Use of the scaling factors reduced the geometric error of the measured bead positions within the phantom by a factor of two at both the isocenter and the offset position. The mean bead position errors were reduced from 54 μm to 27 μm at the isocenter and from 68 μm to 26 μm at the offset position. Localization of the fiducial mounted onto the needle positioning system was also improved by applying the scaling factors. The mean error of the measured fiducial position was reduced from 108 μm to 38 μm. Conclusions: Application of a geometric accuracy phantom in micro‐CT image‐guided needle positioning systems can improve fiducial localization and should reduce the overall needle positioning error of these systems. Further study will be required to determine optimal use of the phantom and its utility in image‐guided interventions.