Raphael Schwarz

Spatio-temporal quantification of vocal fold vibrations using high-speed videoendoscopy and a biomechanical model

Pathologic changes within the organic constitution of vocal folds or a functional impairment of the larynx may result in disturbed or even irregular vocal fold vibrations. The consequences are perturbations of the acoustic speech signal which are perceived as a hoarse voice. By means of appropriate image processing techniques, the vocal fold dynamics are extracted from digital high-speed videos. This study addresses the approach to obtain a parametric description of the spatio-temporal characteristics of the vocal fold oscillations for the aim of classification. For this purpose a biomechanical vocal fold model is introduced. An automatic optimization procedure is developed for fitting the model dynamics to the observed vocal fold oscillations. Thus, the resulting parameter values represent a specific vibration pattern and serve as an objective quantification measure. Performance and reliability of the optimization procedure are validated with synthetically generated data sets. The high-speed videos of two normal voice subjects and six patients suffering from different voice disorders are processed. The resulting model parameters represent a rough approximation of physiological parameters along the entire vocal folds.

Spatiotemporal classification of vocal fold dynamics by a multimass model comprising time-dependent parameters

A model-based approach is proposed to objectively measure and classify vocal fold vibrations by left-right asymmetries along the anterior-posterior direction, especially in the case of nonstationary phonation. For this purpose, vocal fold dynamics are recorded in real time with a digital high-speed camera during phonation of sustained vowels as well as pitch raises. The dynamics of a multimass model with time-dependent parameters are matched to vocal fold vibrations extracted at dorsal, medial, and ventral positions by an automatic optimization procedure. The block-based optimization accounts for nonstationary vibrations and compares the vocal fold and model dynamics by wavelet coefficients. The optimization is verified with synthetically generated data sets and is applied to 40 clinical high-speed recordings comprising normal and pathological voice subjects. The resulting model parameters allow an intuitive visual assessment of vocal fold instabilities within an asymmetry diagram and are applicable to an objective quantification of asymmetries.

Focused shape models for hip joint segmentation in 3D magnetic resonance images

Deformable models incorporating shape priors have proved to be a successful approach in segmenting anatomical regions and specific structures in medical images. This paper introduces weighted shape priors for deformable models in the context of 3D magnetic resonance (MR) image segmentation of the bony elements of the human hip joint. The fully automated approach allows the focusing of the shape model energy to a priori selected anatomical structures or regions of clinical interest by preferentially ordering the shape representation (or eigen-modes) within this type of model to the highly weighted areas. This focused shape model improves accuracy of the shape constraints in those regions compared to standard approaches. The proposed method achieved femoral head and acetabular bone segmentation mean absolute surface distance errors of 0.55±0.18mm and 0.75±0.20mm respectively in 35 3D unilateral MR datasets from 25 subjects acquired at 3T with different limited field of views for individual bony components of the hip joint.

Automated 3D Segmentation of Vertebral Bodies and Intervertebral Discs from MRI

Recent developments in high resolution MRI scanning of the human spine are providing increasing opportunities for the development of accurate automated approaches for pathoanatomical assessment of intervertebral discs and vertebrae. We are developing a fully automated 3D segmentation approach for MRI scans of the human spine based on statistical shape analysis and template matching of grey level intensity profiles. The algorithm reported in the present study was validated on a dataset of high resolution volumetric scans of lower thoracic and lumbar spine obtained on a 3T scanner using the relatively new 3D SPACE (T2-weighted) pulse sequence, and on a dataset of axial T1-weighted scans of lumbar spine obtained on a 1.5T system. A 3D spine curve is initially extracted and used to position the statistical shape models for final segmentation. Initial validating experiments show promising results on both MRI datasets.

Calibration of laryngeal endoscopic high-speed image sequences by an automated detection of parallel laser line projections.

High-speed laryngeal endoscopic systems record vocal fold vibrations during phonation in real-time. For a quantitative analysis of vocal fold dynamics a metrical scale is required to get absolute laryngeal dimensions of the recorded image sequence. For the clinical use there is no automated and stable calibration procedure up to now. A calibration method is presented that consists of a laser projection device and the corresponding image processing for the automated detection of the laser calibration marks. The laser projection device is clipped to the endoscope and projects two parallel laser lines with a known distance to each other as calibration information onto the vocal folds. Image processing methods automatically identify the pixels belonging to the projected laser lines in the image data. The line detection bases on a Radon transform approach and is a two-stage process, which successively uses temporal and spatial characteristics of the projected laser lines in the high-speed image sequence. The robustness and the applicability are demonstrated with clinical endoscopic image sequences. The combination of the laser projection device and the image processing enables the calibration of laryngeal endoscopic images within the vocal fold plane and thus provides quantitative metrical data of vocal fold dynamics.

Three-dimensional morphological and signal intensity features for detection of intervertebral disc degeneration from magnetic resonance images

BACKGROUND AND OBJECTIVES Advances in MRI hardware and sequences are continually increasing the amount and complexity of data such as those generated in high-resolution three-dimensional (3D) scanning of the spine. Efficient informatics tools offer considerable opportunities for research and clinically based analyses of magnetic resonance studies. In this work, we present and validate a suite of informatics tools for automated detection of degenerative changes in lumbar intervertebral discs (IVD) from both 3D isotropic and routine two-dimensional (2D) clinical T2-weighted MRI. MATERIALS AND METHODS An automated segmentation approach was used to extract morphological (traditional 2D radiological measures and novel 3D shape descriptors) and signal appearance (extracted from signal intensity histograms) features. The features were validated against manual reference, compared between 2D and 3D MRI scans and used for quantification and classification of IVD degeneration across magnetic resonance datasets containing IVD with early and advanced stages of degeneration. RESULTS AND CONCLUSIONS Combination of the novel 3D-based shape and signal intensity features on 3D (area under receiver operating curve (AUC) 0.984) and 2D (AUC 0.988) magnetic resonance data deliver a significant improvement in automated classification of IVD degeneration, compared to the combination of previously used 2D radiological measurement and signal intensity features (AUC 0.976 and 0.983, respectively). Further work is required regarding the usefulness of 2D and 3D shape data in relation to clinical scores of lower back pain. The results reveal the potential of the proposed informatics system for computer-aided IVD diagnosis from MRI in large-scale research studies and as a possible adjunct for clinical diagnosis.

Automated MR Hip Bone Segmentation

The accurate segmentation of the bone and articular cartilages from magnetic resonance (MR) images of the hip is important for clinical studies and drug trials into conditions like osteoarthritis. In current studies, segmentations are obtained using time-consuming manual or semi-automatic algorithms which have high inter- and intra-observer variabilities. This paper presents an important step towards obtaining automatic and accurate segmentations of the hip cartilages, namely an approach to automatically segment the bones. The segmentation is performed using three-dimensional active shape models, which are initialized using an affine registration to an atlas. The accuracy and robustness of the approach was experimentally validated using an MR database of we VIBE, we DESS and MEDIC MR images. The (left, right) femoral and ace tabular bone segmentation had a median Dice similarity coefficient of (0.921, 0.926) and (0.830, 0.813).

Model‐based quantification of pathological voice production

Hoarseness, the primary symptom of voice disorders, results from irregular vocal fold vibrations. The oncological therapy of laryngeal cancer may even result in a total loss of voice if an excision of the larynx, and thus, the vocal folds, is necessary. State‐of‐the‐art voice rehabilitation technique in this case is the utilization of scarred tissue in the upper part of the esophagus for substitute voice production. The quality of laryngeal voice, as well as the substitute voice, primarily depends on the anatomy and the vibration patterns of the voice‐producing element. Using endoscopic high‐speed recordings, the voice generators are observed during voice production. In this work, a model‐based approach feasible for the analysis and objective quantification of vocal fold vibrations, as well as the PE dynamics, is presented. By means of an automatic parameter optimization, the dynamic of a biomechanical model of the considered voice‐producing element is fitted to the recorded vibration patterns. Thereby sp...

Automatic detection of anatomical landmarks on the knee joint using MRI data

To propose a new automated learning‐based scheme for locating anatomical landmarks on the knee joint using three‐dimensional (3D) MR image data.

Automatic model‐based semantic registration of multimodal MRI knee data

To propose a robust and automated model‐based semantic registration for the multimodal alignment of the knee bone and cartilage from three‐dimensional (3D) MR image data.