Registration Of Portal Images Research Articles

SU-E-J-237: Image Feature Based DRR and Portal Image Registration

Purpose: Two-dimensional (2D) matching of the kV X-ray and digitally reconstructed radiography (DRR) images is an important setup technique for image-guided radiotherapy (IGRT). In our clinics, mutual information based methods are used for this purpose on commercial linear accelerators, but with often needs for manual corrections. This work proved the feasibility that feature based image transform can be used to register kV and DRR images. Methods: The scale invariant feature transform (SIFT) method was implemented to detect the matching image details (or key points) between the kV and DRR images. These key points represent high image intensity gradients, and thus the scale invariant features. Due to the poor image contrast from our kV image, direct application of the SIFT method yielded many detection errors. To assist the finding of key points, the center coordinates of the kV and DRR images were read from the DICOM header, and the two groups of key points with similar relative positions to their corresponding centers were paired up. Using these points, a rigid transform (with scaling, horizontal and vertical shifts) was estimated. We also artificially introduced vertical and horizontal shifts to test the accuracy of our registration method on anterior-posterior (AP) and lateral pelvic images. Results: The results provided a satisfactory overlay of the transformed kV onto the DRR image. The introduced vs. detected shifts were fit into a linear regression. In the AP image experiments, linear regression analysis showed a slope of 1.15 and 0.98 with an R2 of 0.89 and 0.99 for the horizontal and vertical shifts, respectively. The results are 1.2 and 1.3 with R2 of 0.72 and 0.82 for the lateral image shifts. Conclusion: This work provided an alternative technique for kV to DRR alignment. Further improvements in the estimation accuracy and image contrast tolerance are underway.

DRR and Portal Image Registration for Automatic Patient Positioning in Radiotherapy Treatment

Image processing turns out to be essential in the planning and verification of radiotherapy treatments. Before applying a radiotherapy treatment, a dosimetry planning must be performed. Usually, the planning is done by means of an X-ray volumetric analysis using computerized tomography, where the area to be radiated is marked out. During the treatment phase, it is necessary to place the patient under the particle accelerator exactly as considered in the dosimetry stage. Coarse alignment is achieved using fiduciary markers placed over the patient's skin as external references. Later, fine alignment is provided by comparing a digitally reconstructed radiography (DRR) from the planning stage and a portal image captured by the accelerator in the treatment stage. The preprocessing of DRR and portal images, as well as the minimization of the non-shared information between both kinds of images, is mandatory for the correct operation of the image registration algorithm. With this purpose, mathematical morphology and image processing techniques have been used. The present work describes a fully automatic method to calculate more accurately the necessary displacement of the couch to place the patient exactly at the planned position. The proposed method to achieve the correct positioning of the patient is based on advanced image registration techniques. Preliminary results show a perfect match with the displacement estimated by the physician.

Generic Planning Target Margin for Rectal Cancer Treatment Setup Variation

To calculate the generic planning target margin (GPTM) for patients receiving radiation therapy (RT) for rectal cancer placed in a prone position with a customized cradle for small-bowel exclusion. A total of 25 consecutive rectal cancer patients were treated for 25 or 28 fractions in a prone position using a cradle to maximize small bowel exclusion. Treatment planning computed tomography (CT) scans were used to create orthogonally digitally reconstructed radiographs (DRRs) for portal image registration, which were compared with daily portal images from an electronic portal-imaging device (EPID). Translation values needed to align the DRRs and EPIDs were recorded for the superior to inferior (SI), right to left (RL), and anterior to posterior (AP) directions, and used to calculate the GPTM using the four-parameter model. Age, weight, and body mass index were tested compared with the setup variation using a Pearson correlation and a t test for significance. Gender versus setup variation was compared with a t test. A total of 1,723 EPID images were reviewed. The GPTM was 10 mm superior, 8 mm inferior, 7 mm RL and 10 mm AP. Age and gender were unrelated to setup variation. Weight was significantly associated with systematic AP variation (p < 0.05). BMI was significantly associated with systematic SI (p < 0.05) and AP (p < 0.01) variation and random RL variation (p < 0.05). The GPTM for rectal cancer is asymmetric with a maximum of 10 mm in the superior, anterior and posterior dimensions. Body mass index may effect setup variation. Research using advanced treatment planning should include these margins in the planning target volume definition.

Automated 2D–3D registration of portal images and CT data using line‐segment enhancement

In prostate radiotherapy, setup errors with respect to the patient's bony anatomy can be reduced by aligning 2D megavoltage (MV) portal images acquired during treatment to a reference 3D kilovoltage (kV) CT acquired for treatment planning purposes. The purpose of this study was to evaluate a fully automated 2D-3D registration algorithm to quantify setup errors in 3D through the alignment of line-enhanced portal images and digitally reconstructed radiographs computed from the CT. The line-enhanced images were obtained by correlating the images with a filter bank of short line segments, or "sticks" at different orientations. The proposed methods were validated on (1) accurately collected gold-standard data consisting of a 3D kV cone-beam CT scan of an anthropomorphic phantom of the pelvis and 2D MV portal images in the anterior-posterior (AP) view acquired at 15 different poses and (2) a conventional 3D kV CT scan and weekly 2D MV AP portal images of a patient over 8 weeks. The mean (and standard deviation) of the absolute registration error for rotations around the right-lateral (RL), inferior-superior (IS), and posterior-anterior (PA) axes were 0.212 degree (0.214 degree), 0.055 degree (0.033 degree) and 0.041 degree (0.039 degree), respectively. The corresponding registration errors for translations along the RL, IS, and PA axes were 0.161 (0.131) mm, 0.096 (0.033) mm, and 0.612 (0.485) mm. The mean (and standard deviation) of the total registration error was 0.778 (0.543) mm. Registration on the patient images was successful in all eight cases as determined visually. The results indicate that it is feasible to automatically enhance features in MV portal images of the pelvis for use within a completely automated 2D-3D registration framework for the accurate determination of patient setup errors. They also indicate that it is feasible to estimate all six transformation parameters from a 3D CT of the pelvis and a single portal image in the AP view.

TU-EE-A3-01: 2D-3D Registration of Portal Images with the Planning CT for Detection of Patient Positioning Errors

Purpose: To compare the use of 2D‐3D automatic registration of portal images with the planning CT for detection of patient positioning errors and the use of 3D‐3D registration of MVCBCT with the planning CT.Method and Materials: Two prototype programs were used to carry out 2D‐3D and 3D‐3D image registrations. To assess the accuracy and robustness of these programs, 25 sets of 2D portal images, 25 sets of megavoltage conebeam CT (MVCBCT) images with known positioning shifts were acquired. A planning CT of the RANDO was also acquired. The known shifts between these image sets were ranged from −17mm to 4 mm, −20mm to 5mm and −12mm to 6mm, with uncertainty of 4.278mm, 5.359mm, 3.396mm along the latitude, longitude and vertical directions Results: The average differences between 2D‐3D method and the known shifts were −0.632±0.318 mm, −0.121±0.437 mm, −0.416±0.346 mm, compared to 3D‐3D method of 1.487±0.342 mm, −0.127±0.528 mm, 0.083±0.48 mm along the latitude, longitude and vertical directions. The average differences between 2D‐3D and 3D‐3D image registration methods were 0.86±0.286 mm, −1.39±0.347 mm, −0.33±0.303 mm Conclusion: Both 3D‐3D and 2D‐3D registration methods can detect positioning errors within 1 mm. For a rigid body, 2D‐3D method is sufficient. Conflict of Interest: This project is partly funded by SIEMENS.

A comparative evaluation of two head and neck immobilization devices using electronic portal imaging

A study was performed to compare the positioning reproducibility and the cost efficiency for two head and neck immobilization devices: the Uvex (Uvex Safety, Smithfield, USA) plastic mask system and the Finesse Frame with Ultraplast System (PLANET Medical, Svendborg, Denmark). 20 patients treated with 3D conformal radiation therapy for head and neck cancers were randomly selected (10 for each of the two different immobilization systems) and electronic portal images acquired during their course of treatment were saved and used in this study. The anatomical landmark coordinates and their shifts in the anteroposterior (AP) and craniocaudal (CC) directions with respect to the digitized simulator films for lateral fields were analysed using an in-house developed portal image registration system. Statistically, no evidence was found to indicate that the systematic components of the displacement for the Uvex system and the Finesse Frame with Ultraplast System were different from each other or from zero. The random component of displacement was slightly smaller in the AP direction for the Uvex than the Ultraplast system (sigma = 1.9 mm and 2.9 mm, respectively, p = 0.007), but larger in the CC direction (sigma = 3.8 mm and 2.2 mm, respectively, p<10(-9)). Production time and required materials for a radiation therapy department were also quantified to assess costs for each system. The overall costs per patient were estimated at 141.50 dollars (CAD) and 82.10 dollars for the Uvex and Ultraplast systems, respectively. The Finesse Frame with Ultraplast System of immobilization for head and neck cancer treatment provides a field placement reproducibility that is equal to, or greater than, that of the Uvex plastic mask immobilization system and, while it requires more expensive materials, the workload and consequently overall cost is greatly reduced.

Automatic registration of portal images and volumetric CT for patient positioning in radiation therapy

The efficacy of radiation therapy treatment depends on the patient setup accuracy at each daily fraction. A significant problem is reproducing the patient position during treatment planning for every fraction of the treatment process. We propose and evaluate an intensity based automatic registration method using multiple portal images and the pre-treatment CT volume. We perform both geometric and radiometric calibrations to generate high quality digitally reconstructed radiographs (DRRs) that can be compared against portal images acquired right before treatment dose delivery. We use a graphics processing unit (GPU) to generate the DRRs in order to gain computational efficiency. We also perform a comparative study on various similarity measures and optimization procedures. Simple similarity measure such as local normalized correlation (LNC) performs best as long as the radiometric calibration is carefully done. Using the proposed method, we achieved better than 1 mm average error in repositioning accuracy for a series of phantom studies using two open field (i.e., 41 cm 2) portal images with 90° vergence angle.

A new algorithm for the registration of portal images to planning images in the verification of radiotherapy, as validated in prostate treatments

The manual verification of a radiotherapy treatment, where a portal image is matched onto a planning image, is very time consuming and subject to inter- and intraobserver variability. Therefore, a fully automatic matching procedure (image registration) is required. Existing automatic matching algorithms are confounded, however, by irrelevant information in the portal images (i.e., air in the intestines). Therefore, we have developed a new method, which is an extension of chamfer matching and uses, apart from the distance to the nearest edge, additional information on the correspondence of the gradient angle and magnitude of the edges, making the method less sensitive to confounding information in the images. To validate the automatic matching procedure in clinical practice, we applied the new method on 157 images of 29 randomly selected patients treated for carcinoma of the prostate. Three experts manually matched these images in consensus. Subsequently, the same observers assessed the results of the automatic registration. When regular chamfer matching is used for the fully automatic matching procedure, only 5% of the image pairs could be matched correctly, whereas the new method successfully registered 80% by using additional information on the angle of the edges. From the results of the validation study it can be concluded that a significant reduction in workload for the physicians and technicians can be achieved with this method.

Evaluation of cost functions for gray value matching of two-dimensional images in radiotherapy.

In external beam radiotherapy, portal imaging is applied for verification of the patient setup. Current automatic methods for portal image registration, which are often based on segmentation of anatomical structures, are especially successful for images of the pelvic region. For portal images of more complicated anatomical structures, e.g., lung, these techniques are less successful. It is desirable to have a method for image registration that is applicable for a wide range of treatment sites. In this study, a registration method for two-dimensional (2D) registration of portal and reference images based on intensity values was tested on portal images of various anatomical sites. Tests were performed with and without preprocessing (unsharp mask filtering followed by histogram equalization) for 96 image pairs and six cost functions. The images were obtained from treatments of the rectum, salivary gland, brain, prostate, and lung. To get insight into the behavior of the various cost functions, cost function values were computed for each portal image for 20,000 transformations of the corresponding reference image, translating the reference image in a range of +/- 1 cm and rotating +/- 10 degrees with respect to the clinical match. The automatic match was defined as the transformation associated with the global minimum (found by an exhaustive search). Without preprocessing, the registration reliability was low (less than 27%). With preprocessing, about 90% of the matches were successful, with a difference with our gold standard (manual registration) of about 1 mm and 1 degree SD. All tested cost functions performed similarly. However, the number of local minima using mutual information was larger than for the other tested cost functions. A cost function based on the mean product of the corresponding pixel values had the least number of local minima. In conclusion, gray value based registration of portal images is applicable for a wide range of treatment sites. However, pre-processing of the images is essential.

1080 Poster the RT navigator portal image registration system — First clinical test results

1080 Poster the RT navigator portal image registration system — First clinical test results

Monte Carlo simulations of the imaging performance of metal plate/phosphor screens used in radiotherapy.

The imaging performance of metal plate/phosphor screens which are used for the creation of portal images in radiotherapy is investigated by using Monte Carlo simulations. To this end the modulation transfer function, the noise power spectrum and the detective quantum efficiency [DQE(f)] are calculated for different metals and phosphors and different thicknesses of metal and phosphor for a range of spatial resolutions. The interaction of x-rays with the metal plate/phosphor screen is modeled with the EGS4 electron gamma shower code. Optical transport in the phosphor is modeled by simulating scattering and reabsorption events of individual optical photons. It is shown that metals with a high atomic number perform better than lighter metals in maximizing the DQE(f). It is furthermore shown that the DQE(f) for the metal plate/phosphor screen alone is nearly x-ray quantum absorption limited up to spatial frequencies of 0.4 cycles/mm. In addition, it is argued that the secondary quantum sink of optical photons imposed by the optical chain (mirror, lenses and video camera) leads to a significant degradation of the signal-to-noise ratio at spatial frequencies which are most important for successful registration of portal images. Therefore, the conclusion is that a replacement of the optical chain by a flat array of photodiodes placed directly under the phosphor will lead to a substantial improvement in image quality of portal images.

Determination of rotations in three dimensions using two-dimensional portal image registration.

The relative relationships among anatomic features visualized on planar radiographic images change due to rotations of the patient out of the imaging plane. These changes can be predicted a priori from a three-dimensional radiographic model of the patient. In this study we assess the feasibility of using that information together with a planar image feature alignment tool to account for out-of-plane rotations in the evaluation of subsequent clinical patient images. A series of digitally reconstructed radiographs (DRRs) with known patient rotations was generated from a computed tomography scan of an anthropomorphic head phantom. Fixed anatomic features were extracted, as seen in the DRRs of rotated anatomy and entered into a database. Alignment of features from test radiographs with those from an entry in this database yielded an estimate of rotation out of plane (database entry that resulted in the best fit via planar transformation) along with the planar components of setup errors in the rotated plane. Tests using DRRs and films show that it is possible to select anatomic features in AP skull radiographs with position and orientation sensitive to out-of-plane rotation.

Erratum: “A grey-level image alignment algorithm for registration of portal images and digitally reconstructed radiographs” [Med. Phys. 23 , 75-84 (1996)

Medical PhysicsVolume 23, Issue 7 p. 1299-1299 Free Access Erratum: “A grey-level image alignment algorithm for registration of portal images and digitally reconstructed radiographs” [Med. Phys. 23, 75–84 (1996)] This article corrects the following: A grey-level image alignment algorithm for registration of portal images and digitally reconstructed radiographs D. H. Hristov, B. G. Fallone, Volume 23Issue 1Medical Physics pages: 75-84 First Published online: June 4, 1998 D. H. Hristov, D. H. Hristov Medical Physics Unit and Department of Physics, McGill University, Montreal General Hospital, 1650 Cedar Ave., Montreal, H3G 1A4, CanadaSearch for more papers by this authorB. G. Fallone, B. G. Fallone Medical Physics Unit and Department of Physics, McGill University, Montreal General Hospital, 1650 Cedar Ave., Montreal, H3G 1A4, CanadaSearch for more papers by this author D. H. Hristov, D. H. Hristov Medical Physics Unit and Department of Physics, McGill University, Montreal General Hospital, 1650 Cedar Ave., Montreal, H3G 1A4, CanadaSearch for more papers by this authorB. G. Fallone, B. G. Fallone Medical Physics Unit and Department of Physics, McGill University, Montreal General Hospital, 1650 Cedar Ave., Montreal, H3G 1A4, CanadaSearch for more papers by this author First published: July 1996 https://doi.org/10.1118/1.597780Citations: 2AboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat No abstract is available for this article.Citing Literature Volume23, Issue7July 1996Pages 1299-1299 RelatedInformation

A grey‐level image alignment algorithm for registration of portal images and digitally reconstructed radiographs

An algorithm for automatic registration of pairs of portal images based on image correlation is presented. It uses a fast-Fourier-transform-based cross-correlation operator to find the optimal registration, accounting for both in-plane translations and rotations. Different cross-correlation operators have been tested: the Pearson linear correlation coefficient has been implemented by fast Fourier transform and its performance has been compared to that of the more conventional normalized cross-correlation. A sequential approach has been applied to speed up the registration considerably without degrading the performance of the algorithm. The algorithm has also been applied to the automatic registration of portal images to digitally reconstructed radiographs (DRRs), which have been modified to resemble megavoltage images. The results are indicative of the feasibility of this approach to the inspection of patient setup in radiation therapy.

The cumulative verification image analysis tool for offline evaluation of portal images

Daily portal images acquired using electronic portal imaging devices contain important information about the setup variation of the individual patient. The data can be used to evaluate the treatment and to derive correction for the individual patient. The large volume of images also require software tools for efficient analysis. This article describes the approach of cumulative verification image analysis (CVIA) specifically designed as an offline tool to extract quantitative information from daily portal images. The user interface, image and graphics display, and algorithms of the CVIA tool have been implemented in ANSCI C using the X Window graphics standards. The tool consists of three major components: (a) definition of treatment geometry and anatomical information; (b) registration of portal images with a reference image to determine setup variation; and (c) quantitative analysis of all setup variation measurements. The CVIA tool is not automated. User interaction is required and preferred. Successful alignment of anatomies on portal images at present remains mostly dependent on clinical judgment. Predefined templates of block shapes and anatomies are used for image registration to enhance efficiency, taking advantage of the fact that much of the tool's operation is repeated in the analysis of daily portal images. The CVIA tool is portable and has been implemented on workstations with different operating systems. Analysis of 20 sequential daily portal images can be completed in less than 1 h. The temporal information is used to characterize setup variation in terms of its systematic, random and time-dependent components. The cumulative information is used to derive block overlap isofrequency distributions (BOIDs), which quantify the effective coverage of the prescribed treatment area throughout the course of treatment. Finally, a set of software utilities is available to facilitate feedback of the information for treatment plan recalculation and to test various decision strategies for treatment adjustment. The CVIA tool provides comprehensive analysis of daily images acquired with electronic portal imaging devices. Its offline approach allows characterization of the nature of setup variation for the individual patient that would have been difficult to deduce using only a few daily or weekly portal images. Distribution of the tool will help establish an important database of setup variation from many clinics. The information derived from CVIA can also serve as the foundation to integrate treatment verification, treatment planning, and treatment delivery.

Core-based portal image registration for automatic radiotherapy treatment verification

Portal imaging is the most important quality assurance procedure for monitoring the reproducibility of setup geometry in radiation therapy. The role of portal imaging has become even more critical in recent years due to the migration of three-dimensional (3D) treatment planning technology, including high-precision conformal therapy, from the research setting to routine clinical practice. Unfortunately, traditional methods for acquiring and interpreting portal images suffer from a number of deficiencies that contribute to the well-documented observation that many setup errors go undetected, and some persist for a clinically significant portion of the prescribed dose. Significant improvements in both accuracy and efficiency of detecting setup errors can, in principle, be achieved by using automatic image registration for on-line screening of images obtained from electronic portal imaging devices (EPIDs). This article presents recent developments in a method called core-based image analysis that shows great promise for achieving the desired improvements in error detection. Core-based image analysis is a fundamental computer vision method that is capable of exploiting the full power of EPIDs by providing for on-line detection of setup errors via automatic registration of user-selected anatomical structures. We describe a robust method for automatic portal image registration based on core analysis and demonstrate an approach for assessing both accuracy and precision of registration methods using realistic, digitally reconstructed portal radiographs (DRPRs) where truth is known. Automatic core-based analysis of a set of 20 DRPRs containing known, random field positioning errors was performed for a patient undergoing treatment for prostate cancer. In all cases, the reported translation was within 1 mm of the actual translation with mean absolute errors of 0.3 mm and standard deviations of 0.3 mm. In all cases, the reported rotation was within 0.6 degree of the actual rotation with a mean absolute error of 0.18 degree and a standard deviation of 0.23 degree. Our results, using digitally reconstructed portal radiographs that closely resemble clinical portal images, suggest that automatic core-based registration is suitable as an on-line screening tool for detecting and quantifying patient setup errors.

The effects of out-of-plane rotations on two dimensional portal image registration in conformal radiotherapy of the prostate

Rotations of the patient out of the image plane can significantly degrade the accuracy of two-dimensional (2D) image registration. This study determines the magnitude of the geometric errors introduced by 2D image registration as a result of out-of-plane rotations, and analyzes the dosimetric effects of these errors. The magnitude of the errors introduced by 2D registration were determined by comparing orthogonal view portal images of a rotated phantom to simulator reference images of the same phantom without rotation. Dosimetric effects were calculated for three-dimensional (3D) conformal prostate treatments by applying the registration errors to patient treatment plans. The calculations were performed using a modified version of the dose calculation software used in our Cancer Center for 3D treatment planning based on computed tomography (CT). A method to detect out-of-plane rotations, specific to pelvic treatments, is introduced that uses the relative displacement of the centers of gravity of the acetabula in lateral images. The inherent uncertainty in the registration algorithm was 0.6 +/- 0.5 mm in translation and 0.7 +/- 0.8 degree in rotation within the image plane. For a 2 degrees out-of-plane rotation, the errors increase to 2.3 +/- 1.0 mm and 1.2 +/- 1.1 degrees. In some clinically realizable treatment scenarios it was observed that the errors introduced by the registration procedure could result in an overdosing of the rectal wall. The method to detect out-of-plane rotations was found to have an accuracy of better than 1 degree for rotations of less than 10 degrees. The errors introduced to the patient position by 2D image registration have dosimetrically significant consequences for out-of-plane rotations of 2 degrees or more. However, when used in conjunction with the method to detect out-of-plane rotations, 2D registration software was found to cause insignificant dose errors and, thus, become a more reliable and accurate clinical tool.

Digital portal image registration by sequential anatomical matchpoint and image correlations for real‐time continuous field alignment verification

Detection of radiotherapy field misalignments with electronic portal imaging devices requires the precise initial registration of the digital portal image with a reference image indicating the prescribed field alignment. Moreover, for real-time continuous detection this registration must be performed rapidly--arguably within 250 ms. The quality of this registration is sensitive to the ability of the user to accurately identify corresponding anatomical landmarks in the image pair. To improve the accuracy of the registration and, ultimately, that of the field misalignment measurement, we have developed a sequential digital portal image registration method using both user-identified anatomical matchpoints and image information. A first pass generates registration parameters from user-provided matchpoint coordinates and explicitly accounts for the uncertainty in matchpoint identification. The second pass uses both the initial registration parameters and image information to further improve the registration quality by maximizing cross correlations between segments of the image pair. As this registration method does not use massive matrix/vector computations common to other algorithms, it is inherently faster and well-suited for real-time field placement error detection. On a platform representative of those controlling many commercial electronic portal imaging devices (486 CPU), this algorithm registers portal images in times of less than 6 ms per matchpoint with errors of less than 2% in magnification, 0.5 degree in in-plane rotation, and less than 1 pixel dimension in in-plane translation. As the algorithm assumes a rigid-body geometry, it is sensitive to out-of-plane rotations. A quantitative analysis of this algorithm is presented, indicates its accuracy, and describes its sensitivity to out-of-plane rotations.

153 Evaluation of the accuracy of a method for automatic portal image registration

153 Evaluation of the accuracy of a method for automatic portal image registration

A semiautomatic method for registration of portal images.

A method to register pairs of portal images quickly and accurately has been developed. The approach uses a cross-correlation operator to find the optimal match between corresponding anatomic regions that have been selected by a user on pairs of portal images. The cross-correlation operator determines the translation that best registers each pair of anatomic features independently, and then the images are translated, rotated, and scaled so that the least squares difference between the coordinates of all of the paired regions is minimized. Tests using simulated images have shown that the accuracy of the algorithm is dependent on (i) the size and shape of the structures within the paired regions; (ii) the subject contrast of the anatomic features being matched; (iii) the rotational difference between the anatomic features; and, (iv) the noise in the image. Tests using contrast-detail phantom images have shown that the semiautomated registration algorithm is only slightly less accurate than human observers, but is considerably faster. The semiautomated algorithm shows great promise as a method of quickly and accurately estimating discrepancies in patient positioning when used in conjunction with an on-line portal imaging system.