Axel Saalbach

Automatic multi-model-based segmentation of the left atrium in cardiac MRI scans

Model-based segmentation approaches have been proven to produce very accurate segmentation results while simultaneously providing an anatomic labeling for the segmented structures. However, variations of the anatomy, as they are often encountered e.g. on the drainage pattern of the pulmonary veins to the left atrium, cannot be represented by a single model. Automatic model selection extends the model-based segmentation approach to handling significant variational anatomies without user interaction. Using models for the three most common anatomical variations of the left atrium, we propose a method that uses an estimation of the local fit of different models to select the best fitting model automatically. Our approach employs the support vector machine for the automatic model selection. The method was evaluated on 42 very accurate segmentations of MRI scans using three different models. The correct model was chosen in 88.1% of the cases. In a second experiment, reflecting average segmentation results, the model corresponding to the clinical classification was automatically found in 78.0% of the cases.

Optimizing GHT-based heart localization in an automatic segmentation chain

With automated image analysis tools entering rapidly the clinical practice, the demands regarding reliability, accuracy, and speed are strongly increasing. Systematic testing approaches to determine optimal parameter settings and to select algorithm design variants become essential in this context. We present an approach to optimize organ localization in a complex segmentation chain consisting of organ localization, parametric organ model adaptation, and deformable adaptation. In particular, we consider the Generalized Hough Transformation (GHT) and 3D heart segmentation in Computed Tomography Angiography (CTA) images. We rate the performance of our GHT variant by the initialization error and by computation time. Systematic parameter testing on a compute cluster allows to identify a parametrization with a good tradeoff between reliability and speed. This is achieved with coarse image sampling, a coarse Hough space resolution and a filtering step that we introduced to remove unspecific edges. Finally we show that optimization of the GHT parametrization results in a segmentation chain with reduced failure rates.

Filter learning and evaluation of the computer aided visualization and analysis (CAVA) paradigm for pulmonary nodules using the LIDC-IDRI database

We present a simple rendering scheme for thoracic CT datasets which yields a color coding based on local differential geometry features rather than Hounsfield densities. The local curvatures are computed on several resolution scales and mapped onto different colors, thereby enhancing nodular and tubular structures. The rendering can be used as a navigation device to quickly access points of possible chest anomalies, in particular lung nodules and lymph nodes. The underlying principle is to use the nodule enhancing overview as a possible alternative to classical CAD approaches by avoiding explicit graphical markers. For performance evaluation we have used the LIDC-IDRI lung nodule data base. Our results indicate that the nodule-enhancing overview correlates well with the projection images produced from the IDRI expert annotations, and that we can use this measure to optimize the combination of differential geometry filters.

A novel bone suppression method that improves lung nodule detection : Suppressing dedicated bone shadows in radiographs while preserving the remaining signal.

PurposeSuppressing thoracic bone shadows in chest radiographs has been previously reported to improve the detection rates for solid lung nodules, however at the cost of increased false detection rates. These bone suppression methods are based on an artificial neural network that was trained using dual-energy subtraction images in order to mimic their appearance.MethodHere, a novel approach is followed where all bone shadows crossing the lung field are suppressed sequentially leaving the intercostal space unaffected. Given a contour delineating a bone, its image region is spatially transferred to separate normal image gradient components from tangential component. Smoothing the normal partial gradient along the contour results in a reconstruction of the image representing the bone shadow only, because all other overlaid signals tend to cancel out each other in this representation.ResultsThe method works even with highly contrasted overlaid objects such as a pacemaker. The approach was validated in a reader study with two experienced chest radiologists, and these images helped improving both the sensitivity and the specificity of the readers for the detection and localization of solid lung nodules. The AUC improved significantly from 0.596 to 0.655 on a basis of 146 images from patients and normals with a total of 123 confirmed lung nodules.ConclusionSubtracting all reconstructed bone shadows from the original image results in a soft image where lung nodules are no longer obscured by bone shadows. Both the sensitivity and the specificity of experienced radiologists increased.

Limited angle c-arm tomography and segmentation for guidance of atrial fibrillation ablation procedures

Angiographic projections of the left atrium (LA) and the pulmonary veins (PV) acquired with a rotational C-arm system are used for 3D image reconstruction and subsequent automatic segmentation of the LA and PV to be used as roadmap in fluoroscopy guided LA ablation procedures. Acquisition of projections at high oblique angulations may be problematic due to increased collision danger of the detector with the right shoulder of the patient. We investigate the accuracy of image reconstruction and model based roadmap segmentation using limited angle C-arm tomography. The reduction of the angular range from 200 degrees to 150 degrees leads only to a moderate increase of the segmentation error from 1.5 mm to 2.0 mm if matched conditions are used in the segmentation, i.e., the model based segmentation is trained on images reconstructed with the same angular range as the test images. The minor decrease in accuracy may be outweighed by clinical workflow improvement, gained when large C-arm angulations can be avoided.

Decomposing the bony thorax in X-ray images

The identification and segmentation of target objects from medical images is often confused by other more salient objects in the image. This is a specific problem for X-ray projection images where the shadows of semi-transparent objects are overlaid. A bone shadow may confuse the automated detection of other crossing bones and of important soft tissue findings in the lung like lung nodules. We present a method to identify and remove such bone shadows from a chest radiograph for the purpose of suppressing all bone shadows overlapping with the lung field in a standard posterior-anterior view. In this context an elegant novel approach to the problem of identifying and segmenting overlaid objects is followed: Disturbing objects are identified first and literally removed from the image, therefore no longer confusing the detection of other more subtle objects. This method allowed the identification, segmentation, and suppression of the clavicles, the posterior and the anterior parts of the ribs - one after another. In a clinical study the detection of lung nodules by experienced radiologists was improved after bone suppression.

Heterogeneity of kinetic curve parameters as indicator for the malignancy of breast lesions in DCE MRI

Dynamic contrast enhanced Breast MRI (DCE BMRI) has emerged as powerful tool in the diagnostic work-up of breast cancer. While DCE BMRI is very sensitive, specificity remains to be an issue. Consequently, there is a need for features that support the classification of enhancing lesions into benign and malignant lesions. Traditional features include the morphology and the texture of a lesion, as well as the kinetic parameters of the time-intensity curves, i.e., the temporal change of image intensity at a given location. The kinetic parameters include initial contrast uptake of a lesion and the type of the kinetic curve. The curve type is usually assigned to one of three classes: persistent enhancement (Type I), plateau (Type II), and washout (Type III). While these curve types show a correlation with the tumor type (benign or malignant), only a small sub-volume of the lesion is taken into consideration and the curve type will depend on the location of the ROI that was used to generate the kinetic curve. Furthermore, it has been shown that the curve type significantly depends on which MR scanner was used as well as on the scan parameters. Recently, it was shown that the heterogeneity of a given lesion with respect to spatial variation of the kinetic curve type is a clinically significant indicator for malignancy of a tumor. In this work we compare four quantitative measures for the degree of heterogeneity of the signal enhancement ratio in a tumor and evaluate their ability of predicting the dignity of a tumor. All features are shown to have an area under the ROC curve of between 0.63 and 0.78 (for a single feature).

Foveal fully convolutional nets for multi-organ segmentation.

Most fully automatic segmentation approaches target a single anatomical structure in a specific combination of image modalities and are often difficult to extend to other modalities and protocols or segmentation tasks. More recently, deep learning-based approaches promise to be readily adaptable to new applications as long as a suitable training set is available, although most deep learning architectures are still tuned towards a specific application and data domain. In this paper, we propose a novel fully convolutional neural network architecture for image segmentation and show that the same architecture with the same learning parameters can be used to train models for 20 different organs on two different protocols, while still achieving segmentation accuracy that is on par with the state-of-the-art. In addition, the architecture was designed to minimize the amount of GPU memory required for processing large images, which facilitates the application to full-resolution whole-body CT scans. We have evaluated our method on the publicly available data set of the VISCERAL multi-organ segmentation challenge and compared the performance of our method with those of the challenge and two recently proposed deep learning-based approaches. We achieved the highest Dice similarity coefficients for 17 out of 20 organs for the contrast enhanced CT scans and for 10 out of 20 organs for the uncontrasted CT scans in a cross-comparison between our method and participating methods.

Optimal C-arm angulation during transcatheter aortic valve replacement: Accuracy of a rotational C-arm computed tomography based three dimensional heart model

AIM To investigate the accuracy of a rotational C-arm CT-based 3D heart model to predict an optimal C-arm configuration during transcatheter aortic valve replacement (TAVR). METHODS Rotational C-arm CT (RCT) under rapid ventricular pacing was performed in 57 consecutive patients with severe aortic stenosis as part of the pre-procedural cardiac catheterization. With prototype software each RCT data set was segmented using a 3D heart model. From that the line of perpendicularity curve was obtained that generates a perpendicular view of the aortic annulus according to the right-cusp rule. To evaluate the accuracy of a model-based overlay we compared model- and expert-derived aortic root diameters. RESULTS For all 57 patients in the RCT cohort diameter measurements were obtained from two independent operators and were compared to the model-based measurements. The inter-observer variability was measured to be in the range of 0°-12.96° of angular C-arm displacement for two independent operators. The model-to-operator agreement was 0°-13.82°. The model-based and expert measurements of aortic root diameters evaluated at the aortic annulus (r = 0.79, P < 0.01), the aortic sinus (r = 0.93, P < 0.01) and the sino-tubular junction (r = 0.92, P < 0.01) correlated on a high level and the Bland-Altman analysis showed good agreement. The interobserver measurements did not show a significant bias. CONCLUSION Automatic segmentation of the aortic root using an anatomical model can accurately predict an optimal C-arm configuration, potentially simplifying current clinical workflows before and during TAVR.

Orientation regression in hand radiographs: a transfer learning approach.

Most radiologists prefer an upright orientation of the anatomy in a digital X-ray image for consistency and quality reasons. In almost half of the clinical cases, the anatomy is not upright orientated, which is why the images must be digitally rotated by radiographers. Earlier work has shown that automated orientation detection results in small error rates, but requires specially designed algorithms for individual anatomies. In this work, we propose a novel approach to overcome time-consuming feature engineering by means of Residual Neural Networks (ResNet), which extract generic low-level and high-level features, and provide promising solutions for medical imaging. Our method uses the learned representations to estimate the orientation via linear regression, and can be further improved by fine-tuning selected ResNet layers. The method was evaluated on 926 hand X-ray images and achieves a state-of-the-art mean absolute error of 2.79°.