Contour-based Registration Research Articles

An Estimate of Uncertainty in Deformable Image Registration-Based Accumulated Biological Equivalent Dose for Re-Irradiation

<h3>Purpose/Objective(s)</h3> We describe a dose volume histogram (DVH) overlay technique to estimate uncertainty in deformable image registration (DIR) based accumulated biologically equivalent dose (BED) for re-irradiation planning. We compare applicability of two DIR methods, contour-based (C-DIR) and image intensity and contour hybrid (H-DIR) DIR for BED accumulation. <h3>Materials/Methods</h3> Patient CT-based re-irradiation plans to the same anatomic regions as in their initial irradiations over a four-year period were analyzed. Among the organs-at-risk (OARs) for which BED was accumulated were stomach, heart, and esophagus. Physical dose distributions were converted to BED using OAR-specific parameters based on a linear-quadratic-linear model. For C-DIR, the registration was based only upon the contours drawn on both images. H-DIR is an image intensity-based DIR technique which includes contour-based registration to augment the DIR results. All prior BEDs were warped to the most recent CT image for accumulation. Data input to the DVH overlay workflow were the two CT images to be registered, the deformation vector field, the BED on the earlier CT, and the contours on the later CT. The later-CT contours were warped to the earlier CT, and the earlier-CT BED was warped to the later CT; BED DVHs from both images were then calculated and overlaid. For a fully invertible DIR, the two DVHs for the OAR should be identical; differences between the DVHs were interpreted as BED uncertainties due to inverse consistency error. For a given DVH point, the uncertainty was obtained from the differences between the overlaid DVHs. <h3>Results</h3> For an abdomen case, the stomach accumulated BED_max is 3% higher using C-DIR versus H-DIR; ΔBED_max is approximately 2% for both DIR methods. The higher isodoses were distal to the stomach for all plans; H-DIR, which accounts for regions distal to the OAR, is appropriate for re-irradiation BED calculations. For a thoracic case, both DIR methods yield similar results for the heart. The difference between H-DIR and C-DIR for accumulated heart BED_max is 2.4%; ΔBED_max is 0.9% and 0.6% for H-DIR and C-DIR, respectively. For the esophagus, accumulated BED_max is 54% higher using H-DIR instead of C-DIR, and ΔBED_max is much lower for C-DIR (0.2%) than for H-DIR (1.7%). Visual inspection showed that the BED distribution on the esophagus using H-DIR is unreasonable, attributed to H-DIR relying too heavily upon image intensity features. For accurate BED accumulation in the esophagus, C-DIR is essential during the re-irradiation planning for this case. <h3>Conclusion</h3> The DVH overlay technique can estimate the uncertainty in accumulated BED for re-irradiation scenarios. Caution should be exercised in using various DIR methods before applying DIR dose accumulation for re-irradiation planning. Contour-based DIR may be the preferred registration method for re-irradiation dose accumulation.

Myocardial extracellular volume fraction quantification based on T1 mapping at 3 T: quality optimization by contour-based registration and segmental analysis.

Myocardial extracellular volume fraction (ECV) assessment can be affected by various technical and subject-related factors. To evaluate the role of contour-based registration in quantification of ECV and investigate normal segment-based myocardial ECV values at 3T. Pre- and post-contrast T1 mapping images of the left ventricular basal, mid-cavity, and apical slices were obtained in 26 healthy volunteers. ECV maps were generated using motion correction with and without contour-based registration. The image quality of all ECV maps was evaluated by a 4-point scale. Slices were dichotomized according to the occurrence of misregistration in the source data. Contour-registered ECVs and standard ECVs were compared within each subgroup using analysis of variance for repeated measurements and generalized linear mixed models. In all three slices, higher quality of ECV maps were found using contour-registered method than using standard method. Standard ECVs were statistically different from contour-registered ECVs in global (26.8% ± 2.8% vs. 25.8% ± 2.4%; P = 0.001), mid-cavity (25.4% ± 3.1% vs. 24.3% ± 2.5%; P = 0.016), and apical slices (28.7% ± 4.1% vs. 27.2% ± 3.4%; P = 0.010). In the misregistration subgroups, contour-registered ECVs were lower with smaller SDs (basal: 25.2% ± 1.8% vs. 26.7% ± 2.6%; P = 0.038; mid-cavity: 24.4% ± 2.3% vs. 26.8% ± 3.1%; P = 0.012; apical: 27.5% ± 3.6% vs. 29.7% ± 4.5%; P = 0.016). Apical (27.2% ± 3.4%) and basal-septal ECVs (25.6% ± 2.6%) were statistically higher than mid-cavity ECV (24.3% ± 2.5%; both P < 0.001). Contour-based registration can optimize image quality and improve the precision of ECV quantification in cases demonstrating ventricular misregistration among source images.

Automatic, global registration in laparoscopic liver surgery

PurposeThe initial registration of a 3D pre-operative CT model to a 2D laparoscopic video image in augmented reality systems for liver surgery needs to be fast, intuitive to perform and with minimal interruptions to the surgical intervention. Several recent methods have focussed on using easily recognisable landmarks across modalities. However, these methods still need manual annotation or manual alignment. We propose a novel, fully automatic pipeline for 3D–2D global registration in laparoscopic liver interventions.MethodsFirstly, we train a fully convolutional network for the semantic detection of liver contours in laparoscopic images. Secondly, we propose a novel contour-based global registration algorithm to estimate the camera pose without any manual input during surgery. The contours used are the anterior ridge and the silhouette of the liver.ResultsWe show excellent generalisation of the semantic contour detection on test data from 8 clinical cases. In quantitative experiments, the proposed contour-based registration can successfully estimate a global alignment with as little as 30% of the liver surface, a visibility ratio which is characteristic of laparoscopic interventions. Moreover, the proposed pipeline showed very promising results in clinical data from 5 laparoscopic interventions.ConclusionsOur proposed automatic global registration could make augmented reality systems more intuitive and usable for surgeons and easier to translate to operating rooms. Yet, as the liver is deformed significantly during surgery, it will be very beneficial to incorporate deformation into our method for more accurate registration.

The higher patient-reported outcome measure group had smaller external rotation of the femur in bicruciate-stabilized total knee arthroplasty.

The hypothesis of this study was that the kinematics of patients with higher patient-reported outcome measures (PROMs) differ from those of patients with lower PROMs after bicruciate-stabilized total knee arthroplasty (BCS-TKA). A total of 32 patients with severe knee osteoarthritis were examined 11.2 ± 3.2months after BCS-TKA. The patients performed squats under single fluoroscopic surveillance in the sagittal plane. To estimate the spatial position and orientation of the femoral and tibial components, a 2D-to-3D registration technique was used. This technique uses a contour-based registration algorithm, single-view fluoroscopic images and 3D computer-aided design models. Knee range of motion, varus-valgus alignment, axial rotation of the femur relative to the tibial component, anteroposterior translation of the medial and lateral femorotibial contact points, kinematic paths, and anterior and posterior post-cam engagement were measured. The patients were divided into two groups using hierarchical cluster analysis based on the 1-year postoperative Knee Injury and Osteoarthritis Outcome Score and 2011 Knee Society Score. The femoral component had significantly more external rotation in the low-score group than in the high-score group (5.1 ± 1.8° vs. 2.2 ± 2.7°, p = 0.02). The high-score group had a medial pivot pattern from 0 to 20° of flexion, without significant movement from 20 to 70°, and final bicondylar rollback beyond 70°. The low-score group had a medial pivot from 0 to 70° of flexion and bicondylar rollback beyond 70°. There were no significant between-group differences in the varus-valgus angle or post-cam engagement. The higher PROM group had smaller external rotation of the femur after BCS-TKA. Level III.

Intra- and interrater reliabilities and a method comparison of 2D and 3D techniques in cadavers to determine sacroiliac screw loosening

Sacroiliac (SI) screw loosening may indicate persistent instability, non-union and contribute to pain. Yet, there is no reliable objective measurement technique to detect and monitor SI screw loosening. In 9 cadaveric pelvises one of two SI screw was turned back approximately 20 mm and subsequently assessed by optical measurement, fluoroscopy and a 3D scan using an image intensifier. CTs were segmented and a contour-based registration of the 3D models and the fluoroscopies was performed to measure SI backing out (X-ray module). Three independent observers performed measurements with three repetitions. Deviation of the measurement techniques to the 3D scan, intra- and interrater reliabilities and method equivalence to the 3D scan were assessed. The X-ray module and two fluoroscopic measurement techniques yielded a difference less than 5 mm compared to the 3D scan and equivalence to the 3D scan. Intrarater reliability was for two observers and almost all techniques very good. Three fluoroscopic measurement techniques and optical measurements displayed a very good interrater reliability. The 3D scan and X-ray module yielded the most precise values for SI screw loosening but only the fluoroscopic measurement of the inlet lateral loosening displayed a good reliability and equivalence to the 3D scan.

Statistical motion modelling for robust evaluation of clinically delivered accumulated dose distributions after curative radiotherapy of locally advanced prostate cancer

Background and purposePlanned doses are used as surrogate for the actually delivered dose in radiotherapy. We have estimated the delivered dose in a dose-escalation trial of locally advanced prostate cancer by statistical dose-accumulation and by DVH-summation, and compared to planned dose. Materials and methodPrescribed dose-escalation to the prostate was 67.5 Gy/25fr., corresponding to 81GyEQD2 assuming α/β = 1.5. The 21 patients had three targets (i.e. CTV67.5 + 2 mm, CTV60 + 5 mm, CTV50 + 10 mm) irradiated by a simultaneous-integrated-boost technique. Analysis was based on 213 CT scans and 5-years of follow-up. For statistical dose-accumulation, we modelled 10000 possible treatment courses based on planned dose and deformation-vector-fields from contour-based registration. For DVH-summation we recalculated dose on repeat-CTs and estimated median D98%/EUD. Groups with/without disease recurrence were compared. ResultsDiscrepancies between planned and accumulated dose were mostly seen for CTV67.5, where under-dosage was found at different locations in the prostate in 12/21 patients. Delivered dose-escalation (D98%) was on average 73.9GyEQD2 (range: 68.3–78.7GyEQD2). No significant difference in accumulated-D98% was found in patients with (n = 8) and without (n = 13) recurrence (p > 0.05). Average D98%/EUD with statistical dose-accumulation vs DVH-summation was significantly different in CTV60, CTV50, rectum and bladder but not in CTV67.5. ConclusionThe planned dose escalation was not received by more than half-of-the patients. Robustness of the prostate target (CTV67.5) should therefore be better prioritized in these patients given the low toxicity profile. Estimates of delivered dose were less conservative for dose-accumulation due to interaction of random organ motion with the dose matrix.

A novel contour-based registration of lateral cephalogram and profile photograph

A contour-based automatic registration method of lateral cephalograms and profile photographs is proposed to facilitate the accurate measurement of anteroposterior (AP) position of maxillary central incisors relative to the forehead. There are mainly three steps in the proposed method. First, a hierarchical contour detection algorithm is employed to obtain contours of forehead and nose in lateral cephalograms and profile photographs, respectively. Curve deviation around Nasion of Soft Tissue (Ns) is further corrected using an iterative polynomial curve fitting algorithm according to the characteristics of cephalograms. Second, the coarse registration is performed based on four landmark pairs using a routine least squares method for each image pair, and then the contours are roughly matched. Finally, the coarse registration is further refined by applying the coherent point drift (CPD) algorithm to all the contour points. Using image pairs of lateral cephalograms and profile photographs from 49 Chinese patients, we show that the proposed method achieves higher performance in terms of accuracy and robustness than the existing methods.

Non-Rigid Contour-Based Registration of Cell Nuclei in 2-D Live Cell Microscopy Images Using a Dynamic Elasticity Model

The analysis of the pure motion of subnuclear structures without influence of the cell nucleus motion and deformation is essential in live cell imaging. In this paper, we propose a 2-D contour-based image registration approach for compensation of nucleus motion and deformation in fluorescence microscopy time-lapse sequences. The proposed approach extends our previous approach, which uses a static elasticity model to register cell images. Compared with that scheme, the new approach employs a dynamic elasticity model for the forward simulation of nucleus motion and deformation based on the motion of its contours. The contour matching process is embedded as a constraint into the system of equations describing the elastic behavior of the nucleus. This results in better performance in terms of the registration accuracy. Our approach was successfully applied to real live cell microscopy image sequences of different types of cells including image data that was specifically designed and acquired for evaluation of cell image registration methods. An experimental comparison with the existing contour-based registration methods and an intensity-based registration method has been performed. We also studied the dependence of the results on the choice of method parameters.

Automatic surface defect detection for mobile phone screen glass based on machine vision

Defect detection using machine vision technology plays an important role in the manufacturing process of mobile phone screen glass (MPSG). This study proposes an improved detection algorithm for MPSG defect recognition and segmentation. Considering the problem of MPSG image misalignment caused by vibrations in the mobile stages, a contour-based registration (CR) method is used to generate the template image used to align the MPSG images. Based on this registration result, the combination of subtraction and projection (CSP) is used to identify defects on the MPSG image, which can eliminate the influence of fluctuation in ambient illumination. To segment the defects with a fuzzy grey boundary from a noisy MPSG image, an improved fuzzy c-means cluster (IFCM) algorithm is developed in this study. A defect detection system is developed, and the proposed algorithms are validated using a number of experimental tests on MPSG images. The testing results demonstrate that the approach proposed in this study can effectively detect various defects on MPSG and that it has better performance than other methods.

뉴로내비게이션 시스템 표면정합에 대한 병변 정합 오차의 회전적 특성 분석: 팬텀 연구

Abstract: In this study, we investigated the rotational characteristics which were comprised of directionality and lin-earity of target registration error (TRE) as a study in advance to enhance the accuracy of contour-based registrationin neuronavigation. For the experiment, two rigid head phantoms that have different faces with specially designedtarget frame fixed inside of the phantoms were used. Three-dimensional coordinates of facial surface point cloud andtarget point of the phantoms were acquired using computed tomography (CT) and 3D scanner. Iterative closest point(ICP) method was used for registration of two different point cloud and the directionality and linearity of TRE in over-all head were calculated by using 3D position of targets after registration. As a result, it was represented that TREhad consistent direction in overall head region and was increased in linear fashion as distance from facial surface,but did not show high linearity. These results indicated that it is possible for decrease TRE by controlling orientationof facial surface point cloud acquired from scanner, and the prediction of TRE from surface registration error candecrease the registration accuracy in lesion. In the further studies, we have to develop the contour-based registrationmethod for improvement of accuracy by considering rotational characteristics of TRE.Key words: Neuronavigation, Image-guided surgery, Contour-based registration, Target registration error, Rota-tional characteristics

SU-E-J-04: Initial Experience with 4D-CBCT for Lung Cancer: Physician Verification of Computed Shifts Remains Necessary.

Four-dimensional cone-beam CT (4D-CBCT) is a novel imaging technique used to guide treatment setup for patients with pulmonary lesions by providing additional information about tumor motion at the time of treatment. This study aimed to evaluate the efficacy of the 4D-CBCT capability in ensuring accurate patient setup during SBRT. Twelve patients with pulmonary lesions were imaged pre-treatment with Elekta XVI4.5 using the Symmetry protocol resulting in a respiratory correlated 4D-CBCT. Reconstruction produced 10 phased-based and one average 3DCT image set. Patient shifts were derived from contour-based(mask) registration driven by the weighted average of shifts from each phased CT(4D shifts). Physicians reviewed registration and manually adjusted shifts based on visual registration. We exported the average 3DCT to MIM Vista Software 5.1.1 in reference volume coordinates and manually fused to the reference CT. All manual fusions were contour-based registrations performed by a single observer. No rotations were permitted in manual fusion to mimic clinical procedure. Translational 3D shifts from manual fusion were compared to 4D(automatic registration) shifts and final physician-corrected shifts. Mean differences between 4D and 3D shifts in lateral, longitudinal, and vertical directions were 1.07mm, 5.92mm, and 1.43mm, respectively. Mean differences between physician-corrected and 3D shifts were 1.41mm, 4.83mm, and 1.61mm. Differences between 4D shifts and 3D shifts increased with increasing tumor motion. One patient had consistently large longitudinal differences between 4D and 3D shifts (mean=3.0cm). Further review revealed poor 4D registration(via mask and clipbox) on the XVI system which was corrected by physician adjustment prior to treatment. 4D-CBCT is a valuable imaging tool in patient setup. Physician review of contour-based registration is imperative in preventing a geometrical miss. Caution must be employed in tumors that exhibit a large degree of motion. Further research is necessary in determining functional limits of the 4D-CBCT system.

Active Contour-Based Segmentation of Head and Neck with Adaptive Atlas Selection

This paper presents automated segmentation of structures in the Head and Neck (H&amp;N) region, using an active contour-based joint registration and segmentation model. A new atlas selection strategy is also used. Segmentation is performed based on the dense deformation field computed from the registration of selected structures in the atlas image that have distinct boundaries, onto the patient’s image. This approach results in robust segmentation of the structures of interest, even in the presence of tumors, or anatomical differences between the atlas and the patient image. For each patient, an atlas image is selected from the available atlas-database, based on the similarity metric value, computed after performing an affine registration between each image in the atlas-database and the patient’s image. Unlike many of the previous approaches in the literature, the similarity metric is not computed over the entire image region; rather, it is computed only in the regions of soft tissue structures to be segmented. Qualitative and quantitative evaluation of the results is presented.

Contour and Paired‐Point Registration in a Model for Image‐Guided Surgery

This study assesses target registration error (TRE) of contour-based registration (CBR) and paired-point registration (PPR) for endoscopic sinus surgery. The experimental registration model consisted of a replica of a human head (Sawbones #1345-27; Pacific Research Laboratories, Vashon, WA). Twelve surface fiducial markers were affixed to the simulated skin on the model, and titanium screws were placed in the regions of the anterior ethmoid (AE) and sphenoid face (SF). An axial computed tomography scan (1-mm slice thickness) was then obtained. Registration was then performed on the InstaTrak 3500 Plus (GE Surgical Navigation & Visualization, Lawrence, MA) with the standard PPR protocol with 12 points and CBR protocol with 500, 250, 125, 50, and 4 points. TRE was then calculated at the AE and SF targets. Target registration error was significantly lower for paired-point registration compared with contour-based registration (AE, 0.5 mm vs. 1.5 mm, P < .0001; SF, 0.8 mm vs. 1.5 mm, P < .0001). Among contour-based registration protocols, target registration error at the sphenoid face was lowest with 50 points (1.5 mm, P < .02). At the anterior ethmoid, contour-based registration with 50 points produced a lower target registration error than contour-based registration with 125 points (1.5 vs. 1.8 mm, P < .01). Other target registration error values for CBR were similar at both regions. The target registration error interquartile range was lowest with CBR-125 at both regions. Paired-point registration provided significantly lower target registration error. Contour-based registration with 125 points (and possibly as few as 50 points) produces clinically acceptable target registration error. This registration model elucidates important concepts about registration for surgical navigation for sinus surgery.

Multi-sensor image registration using multi-resolution shape analysis

Multi-sensor image registration has been widely used in remote sensing and medical image field, but registration performance is degenerated when heterogeneous images are involved. An image registration method based on multi-resolution shape analysis is proposed in this paper, to deal with the problem that the shape of similar objects is always invariant. The contours of shapes are first detected as visual features using an extended contour search algorithm in order to reduce effects of noise, and the multi-resolution shape descriptor is constructed through Fourier curvature representation of the contour’s chain code. Then a minimum distance function is used to judge the similarity between two contours. To avoid the effect of different resolution and intensity distribution, suitable resolution of each image is selected by maximizing the consistency of its pyramid shapes. Finally, the transformation parameters are estimated based on the matched control-point pairs which are the centers of gravity of the closed contours. Multi-sensor Landsat TM imagery and infrared imagery have been used as experimental data for comparison with the classical contour-based registration. Our results have been shown to be superior to the classical ones.