André Gooßen

A Stitching Algorithm for Automatic Registration of Digital Radiographs

In digital radiography oversized radiographs have to be assembled from multiple spatially overlapping exposures. We present an algorithm for fast automatic registration of these radiographs. An external feature is brought into the radiographs to facilitate the reconstruction. Pivotal for this algorithm is an actual interpretation of this feature instead of a simple detection. It possesses strong robustness against noise, feature masking and feature displacement. Evaluation has been performed on 2000 pairs of clinical radiographs. The proposed algorithm proved to be a powerful enhancement of established automatic registration algorithms.

Medical x-ray image enhancement by intra-image and inter-image similarity

In medical X-ray examinations, images suffer considerably from severe, signal-dependent noise as a result of the effort to keep applied doses as low as possible. This noise can be seen as an additive signal that degrades image quality and might disguise valuable content. Lost information has to be restored in a post-processing step. The crucial aspect of filtering medical images is preservation of edges and texture on the one hand and removing noise on the other hand. Classical smoothing filters, such as Gaussian or box filtering. are data-independent and equally blur the image content. State-of-the-art methods currently make use of local neighborhoods or global image statistics. However, exploiting global self-similarity within an image and inter-image similarity for subsequent frames of a sequence bears an unused potential for image restoration. We introduce a non-local filter with data-dependent response that closes the gap between local filtering and stochastic methods. The filter is based on the non-local means approach proposed by Buades1 et al. and is similar to bilateral filtering. In order to apply this approach to medical data, we heavily reduce the computational costs incurred by the original approach. Thus it is possible to interactively enhance single frames or selected regions of interest within a sequence. The proposed filter is applicable for time-domain filtering without the need for accurate motion estimation. Hence it can be seen as a general solution for filtering 2D as well as 2D+t X-ray image data.

Ruler-Based Automatic Stitching of Spatially Overlapping Radiographs

We present an algorithm for fast automatic registration of spatially overlapping radiographs. It possesses strong robustness against noise, feature masking and feature displacement. Pivotal for this algorithm is an actual interpretation of the stitching feature instead of a simple detection. The proposed method has been evaluated on 3000 clinical radiographs and proved to be a powerful enhancement of established automatic registration algorithms.

Automatic Joint Alignment Measurements in Pre- and Post-operative Long Leg Standing Radiographs

OBJECTIVES For diagnosis or treatment assessment of knee joint osteoarthritis it is required to measure bone morphometry from radiographic images. We propose a method for automatic measurement of joint alignment from pre-operative as well as post-operative radiographs. METHODS In a two step approach we first detect and segment any implants or other artificial objects within the image. We exploit physical characteristics and avoid prior shape information to cope with the vast amount of implant types. Subsequently, we exploit the implant delineations to adapt the initialization and adaptation phase of a dedicated bone segmentation scheme using deformable template models. Implant and bone contours are fused to derive the final joint segmentation and thus the alignment measurements. RESULTS We evaluated our method on clinical long leg radiographs and compared both the initialization rate, corresponding to the number of images successfully processed by the proposed algorithm, and the accuracy of the alignment measurement. Ground truth has been generated by an experienced orthopedic surgeon. For comparison a second reader reevaluated the measurements. Experiments on two sets of 70 and 120 digital radiographs show that 92% of the joints could be processed automatically and the derived measurements of the automatic method are comparable to a human reader for pre-operative as well as post-operative images with a typical error of 0.7° and correlations of r = 0.82 to r = 0.99 with the ground truth. CONCLUSIONS The proposed method allows deriving objective measures of joint alignment from clinical radiographs. Its accuracy and precision are on par with a human reader for all evaluated measurements.

Spectral volumetric glandularity assessment

Breast density is associated with an increased risk of developing breast cancer, and several methods have been proposed recently for the fully-automatic assessment of volumetric breast density. However, conventional algorithms require an accurate estimation of the breast shape and thickness for the separation into adipose and glandular tissue within the breast. Here, a spectral extension of a recently developed automatic volumetric breast density algorithm is investigated. The proposed approach measures the adipose and glandular tissue content without any additional breast thickness model. The feasibility of the spectral glandularity assessment is illustrated with measurements from an energy-resolving photon-counting mammography system using reference materials including the BR3D phantom.

Auto Exposure Control for Multiple-Slope Cameras

The dynamic range of natural scenes usually exceeds the dynamic range of imaging sensors by several orders of magnitude. To overcome information loss multiple-slope cameras allow acquisition of images at extended dynamic ranges. However the response curve still has to be adapted to the scene. We present a new auto exposure control for multiple-slope cameras. The proposed method derives an optimum response curve in terms of recorded information. It considers dynamic range expansion as well as the resulting coarsening of quantization. We evaluated our method by simulation and implementation for an actual multiple-slope camera.

Automatic volumetric glandularity assessment from full field digital mammograms

Estimation of breast density suffers from high inter-observer variability. A fully automated solution for objective and consistent assessment of breast density from full field digital mammography (FFDM) data is presented. For the computation of glandularity a region of interest (ROI) with a corresponding height model is automatically extracted from the mammograms. Assessment of adipose and glandular tissue volumes is performed by means of calibration data. Volumetric breast density is finally computed as the fraction of glandular tissue volume to overall breast volume with respect to the ROI. The fully automated approach provides volumetric breast density estimates that show strong non-linear correlation with the manual reference (R2=0.80) and high intra-patient consistency (R∈[0.92,0.97]) among mammograms of different orientation or laterality.

Robust estimation of mammographic breast density: a patient-based approach

Breast density has become an established risk indicator for developing breast cancer. Current clinical practice reflects this by grading mammograms patient-wise as entirely fat, scattered fibroglandular, heterogeneously dense, or extremely dense based on visual perception. Existing (semi-) automated methods work on a per-image basis and mimic clinical practice by calculating an area fraction of fibroglandular tissue (mammographic percent density). We suggest a method that follows clinical practice more strictly by segmenting the fibroglandular tissue portion directly from the joint data of all four available mammographic views (cranio-caudal and medio-lateral oblique, left and right), and by subsequently calculating a consistently patient-based mammographic percent density estimate. In particular, each mammographic view is first processed separately to determine a region of interest (ROI) for segmentation into fibroglandular and adipose tissue. ROI determination includes breast outline detection via edge-based methods, peripheral tissue suppression via geometric breast height modeling, and - for medio-lateral oblique views only - pectoral muscle outline detection based on optimizing a three-parameter analytic curve with respect to local appearance. Intensity harmonization based on separately acquired calibration data is performed with respect to compression height and tube voltage to facilitate joint segmentation of available mammographic views. A Gaussian mixture model (GMM) on the joint histogram data with a posteriori calibration guided plausibility correction is finally employed for tissue separation. The proposed method was tested on patient data from 82 subjects. Results show excellent correlation (r = 0.86) to radiologists grading with deviations ranging between -28%, (q = 0.025) and +16%, (q = 0.975).

Implant-Tolerant Orthopaedic Measurements for Post-Operative Radiographs of the Lower Limbs

In this work we present a method for automated orthopaedic measurements for patients that have undergone a partial or full joint replacement in the lower limbs. In contrast to previously published approaches for partially occluded objects, we deal with objects were the major part of the contour is missing, namely the epiphyses of the long bones in the lower limbs, that have been replaced in large parts by artificial joint implants of varying appearance. We present an approach based on the automatic detection and segmentation of implants and a robust adaptation of a segmentation technique based on deformable models. We evaluated our method on a set of clinical images and achieve an accuracy of 0.6 . for angles and 1.3mm for lengths measurements while significantly reducing assessment time and eliminating user interaction.

Fully automatic joint segmentation for computer-aided diagnosis and planning

Orthopaedic examinations are a major reason for radiographic image acquisition. For many diagnostic problems measurements have to be computed from the recorded radiographs. As this task is time-consuming and lacks objectivity, it is desirable to perform these measurements automatically via so-called computational imaging. This requires robust and accurate methods trained and proven on clinical data. We propose a fully automatic technique and present the three complementing stages of our segmentation algorithm. We evaluated the proposed method on more than 200 clinical images and achieve robust and precise delineations, well-suited for automated computation of orthopaedic measurements.