Pierre Gravel

A method for modeling noise in medical images

We have developed a method to study the statistical properties of the noise found in various medical images. The method is specifically designed for types of noise with uncorrelated fluctuations. Such signal fluctuations generally originate in the physical processes of imaging rather than in the tissue textures. Various types of noise (e.g., photon, electronics, and quantization) often contribute to degrade medical images; the overall noise is generally assumed to be additive with a zero-mean, constant-variance Gaussian distribution. However, statistical analysis suggests that the noise variance could be better modeled by a nonlinear function of the image intensity depending on external parameters related to the image acquisition protocol. We present a method to extract the relationship between an image intensity and the noise variance and to evaluate the corresponding parameters. The method was applied successfully to magnetic resonance images with different acquisition sequences and to several types of X-ray images.

Computer-aided method for quantification of cartilage thickness and volume changes using MRI: validation study using a synthetic model

The primary objective of this study was to develop a computer-aided method for the quantification of three-dimensional (3-D) cartilage changes over time in knees with osteoarthritis (OA). We introduced a local coordinate system (LCS) for the femoral and tibial cartilage boundaries that provides a standardized representation of cartilage geometry, thickness, and volume. The LCS can be registered in different data sets from the same patient so that results can be directly compared. Cartilage boundaries are segmented from 3-D magnetic resonance (MR) slices with a semi-automated method and transformed into offset-maps , defined by the LCS. Volumes and thickness are computed from these offset-maps. Further anatomical labeling allows focal volumes to be evaluated in predefined subregions. The accuracy of the automated behavior of the method was assessed, without any human intervention, using realistic, synthetic 3-D MR images of a human knee. The error in thickness evaluation is lower than 0.12 mm for the tibia and femur. Cartilage volumes in anatomical subregions show a coefficient of variation ranging from 0.11% to 0.32%. This method improves noninvasive 3-D analysis of cartilage thickness and volume and is well suited for in vivo follow-up clinical studies of OA knees.

Physical characteristics of a low-dose gas microstrip detector for orthopedic x-ray imaging.

A new scanning slit gas detector dedicated to orthopedic x-ray imaging is presented and evaluated in terms of its fundamental imaging characteristics. The system is based on the micromesh gaseous structure detector and achieves primary signal amplification through electronic avalanche in the gas. This feature, together with high quantum detection efficiency and fan-beam geometry, allows for imaging at low radiation levels. The system is composed of 1764 channels spanning a width of 44.8 cm and is capable of imaging an entire patient at speeds of up to 15 cm/s. The resolution was found to be anisotropic and significantly affected by the beam quality in the horizontal direction, but otherwise sufficient for orthopedic studies. As a consequence of line-by-line acquisition, the images contain some ripple components due to mechanical vibrations combined with variations in the x-ray tube output power. The reported detective quantum efficiency (DQE) values are relatively low (0.14 to 0.20 at 0.5 mm(-1)) as a consequence of a suboptimal collimation geometry. The DQE values were found to be unaffected by the exposure down to 7 microGy, suggesting that the system is quantum limited even for low radiation levels. A system composed of two orthogonal detectors is already in use and can produce dual-view full body scans at low doses. This device could contribute to reduce the risk of radiation induced cancer in sensitive clientele undergoing intensive x-ray procedures, like young scoliotic women.

Language lateralization in individuals with callosal agenesis: An fMRI study

Since the seminal work of Broca in 1861, it is well established that language is essentially processed in the left hemisphere. However, the origin of hemispheric specialization remains controversial. Some authors posit that language lateralization is genetically determined, while others have suggested that hemispheric specialization develops with age. Tenants of the latter view have further suggested that the adult pattern of left hemispheric specialization is achieved by means of callosal inhibition of homologous speech areas in the right hemisphere during ontogeny. According to this hypothesis, one would expect language to develop bilaterally in the acallosal brain. A recent functional magnetic resonance imaging (fMRI) study in one patient with agenesis of the corpus callosum suggests that this might indeed be the case (Riecker et al., 2007). However, given the large anatomic and functional variability in the population of subjects with agenesis of the corpus callosum, this finding needs to be more extensively replicated. In the present study, we explored language lateralization in six individuals with agenesis of the corpus callosum using an fMRI protocol which included a syntactic decision task and a sub-vocal verbal fluency task. Two neurologically intact control groups, one comparable to the acallosals in terms of IQ, age and education (n=6) and one group with a high IQ (n=5), performed the same tasks. No differences were found between language lateralization of the subjects with agenesis of the corpus callosum and the control groups in the receptive speech task. However, for expressive speech, the groups differed with respect to frontal activations, with the acallosal participants showing a more bilateral pattern of activation than the high-IQ participants only. No differences were found for temporal regions. Overall, these results indicate that the corpus callosum is not essential for the establishment of lateralized language functions.

Surface reconstruction from planar x-ray images using moving least squares

Planar radiographs still are the gold standard for the measurement of the skeletal weight-bearing shape and posture. In this paper, we propose to use an as-rigid-as-possible deformation approach based on moving least squares to obtain 3D personalized bone models from planar x-ray images. Our prototype implementation is capable of performing interactive rate shape editing. The biplane reconstructions of both femur and vertebrae show a good accuracy when compared to CT-scan.

Coxofemoral joint kinematics using video fluoroscopic images of treadmill-walking cats: development of a technique to assess osteoarthritis-associated disability

The objectives of this pilot study were to develop a video fluoroscopy kinematics method for the assessment of the coxofemoral joint in cats with and without osteoarthritis (OA)-associated disability. Two non-OA cats and four cats affected by coxofemoral OA were evaluated by video fluoroscopy. Video fluoroscopic images of the coxofemoral joints were captured at 120 frames/s using a customized C-arm X-ray system while cats walked freely on a treadmill at 0.4 m/s. The angle patterns over time of the coxofemoral joints were extracted using a graphic user interface following four steps: (i) correction for image distortion; (ii) image denoising and contrast enhancement; (iii) frame-to-frame anatomical marker identification; and (iv) statistical gait analysis. Reliability analysis was performed. The cats with OA presented greater intra-subject stride and gait cycle variability. Three cats with OA presented a left–right asymmetry in the range of movement of the coxofemoral joint angle in the sagittal plane (two with no overlap of the 95% confidence interval, and one with only a slight overlap) consistent with their painful OA joint, and a longer gait cycle duration. Reliability analysis revealed an absolute variation in the coxofemoral joint angle of 2º–6º, indicating that the two-dimensional video fluoroscopy technique provided reliable data. Improvement of this method is recommended: variability would likely be reduced if a larger field of view could be recorded, allowing the identification and tracking of each femoral axis, rather than the trochanter landmarks. The range of movement of the coxofemoral joint has the potential to be an objective marker of OA-associated disability.

Resolution enhancement in digital x-ray imaging

We have developed a restoration method for radiographs that enhances image sharpness and reveals bone microstructures that were initially hidden in the soft-tissue glare. The method is two fold: the image is first deconvolved using the Richardson-Lucy algorithm and is then divided with a signal modelling the soft-tissue distribution to increase the overall contrast. Each step has its own merits but the power of the restoration method lies in their combination. The originality of the method is its reliance on a priori information at each step in the processing. We have measured and modelled analytically the point-spread function of a low-dose gas microstrip x-ray detector at several beam energies. We measured the relationship between the local image intensity and the noise variance for these images. The soft-tissue signal was also modelled using a minimum-curvature filtering technique. These results were then combined into an image deconvolution procedure that uses wavelet filtering to reduce restoration noise while keeping the enhanced small-scale features. The method was applied successfully to images of a human-torso phantom and improved the contrast of small details on the bones and in the soft tissues. We measured a mean 54% increase in signal to noise ratio and a mean 105% increase in contrast to noise ratio in the 70 and 140 kVp images we analysed. The method was designed to facilitate the analysis of radiographs by relying on two levels of visual inspection. The contrast of the full image is first enhanced by division with the signal modelling the soft-tissue distribution. Based on the result, a radiologist might decide to zoom in on a given image section. The full restoration method is then applied to that region of interest. Indeed, full image deconvolution is often unnecessary since enhanced small-scale details are not visible at large scale; only the section of interest is processed which is more efficient.

Bone enhancement in digital dual energy radiographs from normalization with a synthetic background image

We have developed a method to enhance bone contrast in dual x-ray images. The method relies on the creation of a synthetic background image representing soft tissues and air in the image. The original image is normalized with this background image, thus enhancing bone contrast and eliminating soft tissues and air holes. The central idea of the method resides in the proper combination of information contained in equivalent aluminum and plastic thicknesses, calculated from well-known dual-energy algorithms.

Resolution enhancement in dual-energy x-ray imaging

We have developed a method that uses a large amount of a priori information to generate super resolution radiographs. We measured and modeled analytically the point spread function of a low-dose gas microstrip x-ray detector at several beam energies. We measured the relationship between the local image intensity and the noise variance in the radiographs. The soft-tissue signal in the images was modeled using a minimum-curvature filtering technique. These results were then combined into an image deconvolution procedure using wavelet filtering to reduce restoration noise while keeping the enhanced small-scale features. The method was applied to a resolution grid image to measure its effects on the detector’s modulation transfer function. The restored images of a radiological human-torso phantom revealed small-scale details on the bones that were not seen before, and this, with improved SNR and image contrast. Dual-energy imaging was integrated to the restoration process in order to generate separate high-resolution images of the bones, the soft tissues, and the mean atomic number. This information could be used to detect bone micro-fractures in athletes and to assess bone demineralization in seniors due to osteoporosis. Super resolution radiographs are easier to segment due to their enhanced contrasts and uniform backgrounds; the boundaries of the features of interest can be delimited with a sub-pixel accuracy. This is highly relevant to the morphometric analysis of complex bone structures like individual vertebrae. The restoration method can be automated for a clinical environment use.