Claire Levrier

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.

In vitro characterization of carotid plaque using a clinical ultrasound imaging system

The goal of this work is to demonstrate the feasibility of objective noninvasive methods for in vivo characterization of carotid plaque content. Using a clinical ultrasonic imager connected to a radiofrequency (RF) signal acquisition system, both RF signals and grayscale (GS) images were obtained from a tissue-mimicking reference phantom and from 26 independent segments of formalin-fixed human carotid endarterectomy specimens. Integrated backscatter (IBS) images of each segment were constructed from an FFT-based analysis on individual RF lines using a sliding Hamming window technique (32 point length, 1 point shift). For 82 matched ROIs, average IBS from parameter images was compared to average GS values normalized with respect to GS in the reference phantom. For 70 of the ROIs, plaque content was assessed by histology in terms of collagen, lipid, hemorrhage and calcification content. The current GS and IBS approaches were not found to be equivalent. The IBS values (mean standard error) for collagen, lipid and hemorrhage were 8.1, 1.1, 5.3 /spl plusmn/ 1.0 and 4.9 /spl plusmn/ 0.7 dB, respectively.

Adaptive beamforming using Van Cittert-Zernicke criterion

This paper proposes a phase aberration correction technique, which copes with the issue related to the classical correlation method, which widens images of punctual coherent scatters. This artifact is due to the fact that the correlation maximization process realigns strong echoes coming from these targets even when the beam formed line is adjacent to the target. Our approach chooses to apply this correlation in selected areas, where there is only speckle. To identify these regions, we use a criterion derived from the Van Cittert-Zernicke (VCZ) theorem, which distinguishes between coherent and diffuse scattering. The result of the delay estimation in areas corresponding to punctual targets is then deduces from delays estimated within speckle regions using interpolation. Results of the proposed method, applied on string phantoms as well as on excised breast tissue samples containing microcalcifications, show improvement in contrast and lateral resolution of punctual scatterers.