Jean-Marc Gregoire

3D Compressed sensing ultrasound imaging

This paper proposes a compressed sensing method adapted to 3D ultrasound (US) imaging. Three undersampling patterns suited for 3D US imaging, together with a nonlinear conjugate gradient reconstruction algorithm of the US image k-spaces, are investigated in vivo radio-frequency 3D US volumes. Reconstructions from 50% of the samples of the original 3D volume show little information loss in terms of normalized root mean squared errors.

Lead-free high-frequency linear-array transducer (30 MHz) for in vivo skin imaging

This work presents the fabrication of a 30 MHz, linear-array transducer based on a KN, 1-3 piezocomposite. Performances of the transducer were characterized and compared to a PZT-based linear array with similar structure. The composites were designed to minimize lateral modes of vibration which can severely degrade imaging performances. Fabrication steps were optimized to achieve a 40 MHz resonant frequency in air with a composite thickness of 69 μm. The measured thickness coupling factor was around 50%. A 128-element, linear array was then fabricated with 100 μm pitch and 1.5 mm elevation aperture. The structure of the transducer (backing, matching layers, and electric components) was optimized to deliver good fractional bandwidth and sensitivity. The final probe was integrated in a prototype, real-time, 128-channel scanner to acquire high-resolution images of the human skin in vivo. Results showed that, compared to PZT ceramics, KN single crystals provide low density and high acoustic velocity, both highly desirable for the manufacturing of HF transducers. The central frequency of the linear-array transducer was 30 MHz despite the KN composite being 20% thicker than equivalent PZT-based composites, and the relative bandwidth was about 50%. High-resolution images of the human skin were acquired and detailed features could be visualized.

Human Understandable Features for Segmentation of Solid Texture

The purpose of this paper is to present new texture descriptors dedicated to segmentation of solid textures. The proposed texture attributes are inspired by the human description of texture and allows a general description of texture. Moreover it is more convenient for a user to understand features signification particularly in a man-aided application. In comparison with psychological measurements for human subjects, our characteristics gave good correspondences in rank correlation of 12 different solid textures. Using these texture features, segmentation results obtained with the classical K-means method on solid textures and real three-dimensional ultrasound images of the skin are presented and discussed.

Multidimensional Complex Number Parametric Model Order and Parameters Estimation

This paper presents a technique for accessing multidimensional complex number AR model order and parameters through matrix factorization. The principle of this technique consists of the transformation of the multidimensional model to a pseudo SISO (simple input simple output) AR model then performing factorization of the covariance matrix of the data. This factorization then leads to a recursive form of the parameter and order estimation. The methodology developed here may be applied to an AR model of any dimension. Computer simulation results are provided to illustrate the behavior of this method

3D printed phantom for high frequency ultrasound imaging

In the field of high frequency ultrasound imaging (i.e. ≥ 20 MHz) there are few tools for characterizing imaging systems performances. Indeed, for this frequency range, commercial phantoms are often inadequate. We propose to investigate the use of 3D photo-polymer printing in order to make, resolution and tissue mimicking phantoms by using an original process which consist to use the normally removed support material as a surrounding medium. First, density, acoustic impedance, velocity, dispersion of wave phase velocity and attenuation of the two photo-polymer materials were measured by a reflection experimental set-up using a focused single-element transducer in frequency range of 15 MHz to 30 MHz. Then, a 3D phantom model was designed and printed. Finally, 2D and 3D acoustic images are shown by using a mechanically driven single element transducer. In conclusion, 3D printing allowed us to make phantoms adapted for high frequency including homogeneities of arbitrary geometries and surrounding medium with acoustics properties close to the human tissues like skin.

A framework of perceptual features for the characterisation of 3D textured images

This paper presents a multiresolution system for volumetric texture analysis. The originality of this system partially originates from its use of combinations of perceptual texture features that correspond to adjectives commonly used by humans to describe textures. To approximate these features, we use a combination of different families of texture analysis methods rather than a single texture analysis model. This choice is necessary to obtain a good perceptual feature approximation and allows our system to be robust and generic. Moreover, by using our human-understandable features (HUF), it is convenient for a user to manipulate and select the features that are, according to the user, relevant for a given application. Two experiments are presented: the first experiment demonstrates the strong correspondence between our features and a human’s description of textures, and the second demonstrates the performance of our proposed method. Finally, the proposed HUF are integrated into an interactive segmentation system and are compared to previously proposed descriptors through analysis of several segmentation results of 3D ultrasound images.

Ultrasound pre-clinical platform for diagnosis and targeted therapy

This study reports on the development and the first experimental assessment of a custom ultrasound pre-clinical platform developed for diagnosis and targeted therapy. Moreover, a specific dual-mode probe developed for our application is based on cMUT technology. The platform combines two operating modes: high frequency imaging in the 15-20 MHz frequency range and low frequency focused ultrasound at 1 MHz for local therapy.

Ultrasound imaging: signal acquisition, new advanced processing for biomedical and industrial applications

Use of ultrasound, namely in the biomedical diagnosis and industrial fields, pioneered in 1950s, is today particularly widespread. In the last decades, ultrasound imaging has benefited from advances in numerical technologies such as signal processing. On the other hand, the use of ultrasound imaging has increased the need for signal processing techniques. This paper presents a review and the up-to-date developments in ultrasound imaging techniques, including elementary principles, signal acquisition and processing, from one dimensional to multidimensional systems. This paper also deals with typical relevant applications.

Dual-frequency transducer for nonlinear contrast agent imaging

Detection of high-order nonlinear components issued from microbubbles has emerged as a sensitive method for contrast agent imaging. Nevertheless, the detection of these high-frequency components, including the third, fourth, and fifth harmonics, remains challenging because of the lack of transducer sensitivity and bandwidth. In this context, we propose a new design of imaging transducer based on a simple fabrication process for high-frequency nonlinear imaging. The transducer is composed of two elements: the outer low-frequency (LF) element was centered at 4 MHz and used in transmit mode, whereas the inner high-frequency (HF) element centered at 14 MHz was used in receive mode. The center element was pad-printed using a lead zirconate titanate (PZT) paste. The outer element was molded using a commercial PZT, and curved porous unpoled PZT was used as backing. Each piezoelectric element was characterized to determine the electromechanical performance with thickness coupling factor around 45%. After the assembly of the two transducer elements, hydrophone measurements (electroacoustic responses and radiation patterns) were carried out and demonstrated a large bandwidth (70% at -3 dB) of the HF transducer. Finally, the transducer was evaluated for contrast agent imaging using contrast agent microbubbles. The results showed that harmonic components (up to the sixth harmonic) of the microbubbles were successfully detected. Moreover, images from a flow phantom were acquired and demonstrated the potential of the transducer for high-frequency nonlinear contrast imaging.

Evaluation of high resolution ultrasound as a tool for assessing the 3D volume of blood clots during in vitro thrombolysis

Thrombosis is a major cause of several diseases, i.e. myocardial infarction, cerebral stroke and pulmonary embolism. Thrombolytic therapies are required to induce fast and efficient recanalization of occluded vessels. To evaluate the in vitro efficacy of these thrombolytic strategies, measuring clot dissolution is essential. This study aimed to evaluate and validate high resolution ultrasound as a tool to assess the exact volume of clots in 3D and in real time during in vitro thrombolytic drug testing. This new method was validated by measuring the effects of concentration range of recombinant tissue type plasminogen activator on a blood clot during complete occlusion or 70% stenosis of a vessel. This study shows that high resolution ultrasound imaging allows for a real-time assessment of the 3D volume of a blood clot with negligible inter- and intra-operator variabilities. The conclusions drawn from this study demonstrate the promising potential of high resolution ultrasound imaging for the in vitro assessment of new thrombolytic drugs.