Lung Registration Using Automatically Detected Landmarks

Polzin, Thomas; Rühaak, Jan; Werner, René; Handels, Heinz; Modersitzki, Jan

doi:10.3414/me13-01-0125

Cited by 8 publications

(1 citation statement)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Spatial accuracy assessment of DIR solutions based on large sets of expert-determined landmark point pairs has been shown to be a statistically robust and straightforward framework [20, 25]. The approach has been utilized within many studies to validate novel DIR algorithm performance (see [26–30] for example). Though there is no universal standard that defines an acceptable DIR spatial accuracy for all situations, in the context of inhale/exhale CT image pairs, DIR algorithms are expected to produce spatial accuracies on the order of the voxel dimensions [31].…”

Section: Introductionmentioning

confidence: 99%

The numerical stability of transformation-based CT ventilation

Castillo

Vinogradskiy

et al. 2017

Int J CARS

View full text Add to dashboard Cite

Computed tomography (CT)-derived ventilation imaging utilizes deformable image registration (DIR) to recover respiratory-induced tissue volume changes from inhale/exhale 4DCT phases. While current strategies for validating CT ventilation rely on analyzing its correlation with existing functional imaging modalities, the numerical stability of the CT ventilation calculation has not been characterized.PurposeThe purpose of this study is to examine how small changes in the DIR displacement field can affect the calculation of transformation-based CT ventilation.MethodsFirst, we derive a mathematical theorem, which states that the change in ventilation metric induced by a perturbation to single displacement vector is bounded by the perturbation magnitude. Second, we introduce a novel Jacobian constrained optimization method for computing user-defined CT ventilation images.ResultsUsing the Jacobian constrained method, we demonstrate that for the same inhale/exhale CT pair, it is possible to compute two DIR transformations that have similar spatial accuracies, but generate ventilation images with significantly different physical characteristics. In particular, we compute a CT ventilation image that perfectly correlates with a single-photon emission CT perfusion scan.ConclusionThe analysis and experiments indicate that while transformation-based CT ventilation is a promising modality, small changes in the DIR displacement field can result in large relative changes in the ventilation image. As such, approaches for improving the reproducibility of CT ventilation are still needed.

show abstract

Section: Introductionmentioning

confidence: 99%

The numerical stability of transformation-based CT ventilation

Castillo

Vinogradskiy

et al. 2017

Int J CARS

View full text Add to dashboard Cite

show abstract

Automatic large quantity landmark pairs detection in 4DCT lung images

Thomas

et al. 2019

Medical Physics

View full text Add to dashboard Cite

Purpose To automatically and precisely detect a large quantity of landmark pairs between two lung computed tomography (CT) images to support evaluation of deformable image registration (DIR). We expect that the generated landmark pairs will significantly augment the current lung CT benchmark datasets in both quantity and positional accuracy. Methods A large number of landmark pairs were detected within the lung between the end‐exhalation (EE) and end‐inhalation (EI) phases of the lung four‐dimensional computed tomography (4DCT) datasets. Thousands of landmarks were detected by applying the Harris‐Stephens corner detection algorithm on the probability maps of the lung vasculature tree. A parametric image registration method (pTVreg) was used to establish initial landmark correspondence by registering the images at EE and EI phases. A multi‐stream pseudo‐siamese (MSPS) network was then developed to further improve the landmark pair positional accuracy by directly predicting three‐dimensional (3D) shifts to optimally align the landmarks in EE to their counterparts in EI. Positional accuracies of the detected landmark pairs were evaluated using both digital phantoms and publicly available landmark pairs. Results Dense sets of landmark pairs were detected for 10 4DCT lung datasets, with an average of 1886 landmark pairs per case. The mean and standard deviation of target registration error (TRE) were 0.47 ± 0.45 mm with 98% of landmark pairs having a TRE smaller than 2 mm for 10 digital phantom cases. Tests using 300 manually labeled landmark pairs in 10 lung 4DCT benchmark datasets (DIRLAB) produced TRE results of 0.73 ± 0.53 mm with 97% of landmark pairs having a TRE smaller than 2 mm. Conclusion A new method was developed to automatically and precisely detect a large quantity of landmark pairs between lung CT image pairs. The detected landmark pairs could be used as benchmark datasets for more accurate and informative quantitative evaluation of DIR algorithms.

show abstract

Attaining Human-Level Performance with Atlas Location Autocontext for Anatomical Landmark Detection in 3D CT Data

O’Neil

Kascenas

Henry

et al. 2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

We present an efficient neural network method for locating anatomical landmarks in 3D medical CT scans, using atlas location autocontext in order to learn long-range spatial context. Location predictions are made by regression to Gaussian heatmaps, one heatmap per landmark. This system allows patchwise application of a shallow network, thus enabling multiple volumetric heatmaps to be predicted concurrently without prohibitive GPU memory requirements. Further, the system allows inter-landmark spatial relationships to be exploited using a simple overdetermined affine mapping that is robust to detection failures and occlusion or partial views. Evaluation is performed for 22 landmarks defined on a range of structures in head CT scans. Models are trained and validated on 201 scans. Over the final test set of 20 scans which was independently annotated by 2 human annotators, the neural network reaches an accuracy which matches the annotator variability, with similar human and machine patterns of variability across landmark classes.

show abstract

Lung Registration Using Automatically Detected Landmarks

Cited by 8 publications

References 28 publications

The numerical stability of transformation-based CT ventilation

The numerical stability of transformation-based CT ventilation

Automatic large quantity landmark pairs detection in 4DCT lung images

Attaining Human-Level Performance with Atlas Location Autocontext for Anatomical Landmark Detection in 3D CT Data

Contact Info

Product

Resources

About