Research into the development and application of ultrasound elastography, a technique for imaging stress-induced tissue strain, has so far taken insufficient account of stiffness anisotropy, either as a source or error or artefact in strain imaging, or as a means of extracting additional useful information. Skin was chosen as a model to investigate anisotropy using elastography, because it is anisotropic and easily accessible. Throughout the project, skin samples were stretched to evaluate stiffness, whilst two different imaging techniques were used to produce the tracking signal for generating strain images. Surface optical elastography was developed, using surface topography as the tracking signal. This novel technique proved feasible, so an optimisation technique was developed to improve the elastograms. High frequency ultrasound elasticity imaging was employed for the first time in combination with uniaxial tensile loading of the skin in the ultrasound scan plane, allowing depth and directional dependence of skin and subcutaneous tissue elasticity to be visualised. A novel ultrasound beam steering method for improving quality of tensile strain images was developed to overcome the problem that tracking across an ultrasound beam produces poor quality strain images. A feasibility study showed that strain image quality improves with the new technique. The techniques were used in two very preliminary clinical trials, for Breast Cancer Related Lymphedema and pigmented skin lesions, establishing that the new techniques were clinically feasible, testing 19 patients and 5 healthy volunteers. Differences in directional mechanical behaviour compared to healthy skin were demonstrated, further studies with larger numbers of cases would be valuable. Finite element modelling was used to provide further insight into the findings. It is concluded that anisotropy is important, as a source of variability in strain images and as a source of diagnostic information, and that the new techniques are clinically feasible, useful and worthy of improvement.

Medical ultrasound imaging with the nonlinear second harmonic component (SHC) provides improved spatial resolution and reduced reverberation artifacts. However, the signal-to-noise ratio (SNR) of the SHC is low compared to the fundamental frequency component. In this study, chirp coded signals are explored as an exci- tation technique in ultrasound harmonic imaging. The objective is to increase the SNR of the SHC and improve the axial resolution of ultrasound harmonic imaging using a single transmission event. In ultrasound harmonic imaging with chirp coded excitation, a harmonic matched filter (HMF) is typically used on the received signal to perform pulse compression of the SHC to restore axial resolution. Designing the HMF is a problematic issue as it requires optimal window selection. In the compressed signal, the sidelobe level may increase and the mainlobe width widen under a mismatched condition, resulting in loss of axial resolution. An alternate method of pulse compression using the fractional Fourier transform (FrFT) is presented. It is demonstrated that the FrFT can perform pulse compression of the SHC similar to the HMF with improved axial resolution and comparable peak sidelobe level. The use of nonlinear frequency modulated (NLFM) signals are explored in ultra- sound harmonic imaging. The objective is to optimise the SNR gain and to reduce the peak sidelobe level in the compressed second harmonic chirp signal. Results are presented which show a reduction of peak sidelobe level in the compressed NLFM signal with a comparable axial mainlobe width when compared with the compressed linear frequency modulated (LFM) signal. Chirp coded excitation is also proposed in the area of superharmonic imaging (SHI). It is demonstrated that the chirp coded signals can potentially enhance the SNR and axial resolution of the SHI when compared with conventional tone-burst excitation. Finally, the subharmonic response from contrast microbubbles is measured using LFM, NLFM and tone-burst excitations. Results are presented which indicate that the subharmonic response for wide bandwidth NLFM excitation is higher than the LFM excitation. Both LFM and NLFM excitations provide better subharmonic power than the tone-burst excitation.

No description available.

Recently the use of ultrasonic arrays for imaging defects in metal components has become economically attractive in Non-Destructive Testing. Given a certain array, the image quality strongly depends on how the measurements are process into an image. The current state-of-the-art imaging algorithm in actual use is delay-and-sum beamforming, which has a resolution capability that is fundamentally limited by the physical approximation used to describe how waves interact with matter. This thesis explores the practical use of alternative non-linear “super-resolution” imaging algorithms that use more accurate physical models, and can theoretically achieve unlimited resolution. This is made possible by utilising additional sources of information contained within the measurements, in particular the small amplitude multiply scattered signals. The distribution of information contained in the measurements, and utilised by the imaging algorithms is studied in the context of information capacity of signals. We discover some insights into the limits of imaging which depend on the signal-to-noise ratio. The accuracy of non-linear imaging algorithms can be strongly dependent on the accuracy of the measurements. Therefore several experiments are performed to assess their performance in practice. The experimental implementation of these methods poses a number of challenges, including removal of the incident field, and compensating for array element directivity. Super-resolution capability is demonstrated in a highly attenuative medium for the first time. To further improve the image quality we explore the possibility of using mirror reflections. This gives an increase in the effective aperture. We perform simulated and experimental reconstructions of a complex scatterer and find that the completeness of the image is improved. The mirror interface also allows quantitative speed-of-sound imaging of penetrable scatterers using the HARBUT algorithm. This is tested experimentally for the first time.

Η παρούσα διδακτορική διατριβή έχει στόχο την υπολογιστική υποβοήθηση της διάγνωσης της αθηρωμάτωσης μέσω της ανάπτυξης και εφαρμογής προηγμένων μεθοδολογιών επεξεργασίας και ανάλυσης ακολουθιών εικόνων υπερήχων β-σάρωσης της καρωτίδας. Οι μέθοδοι που παρουσιάζονται στη διατριβή αξιολογούνται τόσο σε συνθετικά όσο και σε πραγματικά απεικονιστικά δεδομένα. Μετά από διεξοδική μελέτη προτείνονται οι κατάλληλες ρυθμίσεις της υπερηχοτομογραφικής διάταξης καθώς και η εφαρμογή διαδικασίας κανονικοποίησης των λαμβανομένων εικόνων, με στόχο τη βελτιστοποίηση του αποτελέσματος των διαδικασιών επεξεργασίας εικόνας. Για την ανάλυση της υφής της αθηρωματικής πλάκας στην καρωτίδα, προτείνονται και αξιολογούνται συγκριτικά τρεις τεχνικές μετασχηματισμού: μετασχηματισμός Fourier, μετασχηματισμός κυματιδίου και φίλτρα Gabor. Αποδεικνύεται ότι χαρακτηριστικά υφής που υπολογίζονται τόσο με βάση το μετασχηματισμό κυματιδίου όσο και με τα φίλτρα Gabor μπορούν να διαχωρίσουν τους δυο τύπους αθηρωματικών πλακών (συμπτωματικές και ασυμπτωματικές). Για την εκτίμηση της κίνησης του τοιχώματος της καρωτίδας προτείνονται και αξιολογούνται τέσσερις μέθοδοι, από τις οποίες η μέθοδος της οπτικής ροής ελαχίστων τετραγώνων με βάρη βρέθηκε να έχει τη βέλτιστη απόδοση. Η μέθοδος αυτή χρησιμοποιείται για την εκτίμηση της μετατόπισης τόσο του υγιούς τοιχώματος της καρωτίδας, όσο και τοιχώματος με ασυμπτωματική ή συμπτωματική αθηρωματική πλάκα. Επιπλέον, προτείνεται μια αυτόματη μέθοδος κατάτμησης του τοιχώματος της καρωτίδας από διαμήκεις και εγκάρσιες εικόνες υπερήχων, η οποία βασίζεται στο μετασχηματισμό Hough. Η προτεινόμενη μεθοδολογία χρησιμοποιείται για τον υπολογισμό χρήσιμων για τη διάγνωση ποσοτικών δεικτών, όπως το πάχος του έσω-μέσου χιτώνα και οι κυματομορφές μεταβολής της αρτηριακής διαμέτρου. Σύμφωνα με τα αποτελέσματα, η εν γένει υψηλή ακρίβεια της μεθόδου περιορίζεται για την περίπτωση εγκάρσιων τομών και αθηρωμάτωσης. Για τη βελτίωση του αποτελέσματος της κατάτμησης σε αυτές τις περιπτώσεις, προτείνεται ο συνδυασμός του μετασχηματισμού Hough με ενεργές καμπύλες. Η μελέτη της κίνησης του αρτηριακού τοιχώματος συμπληρώνεται με ένα μαθηματικό μοντέλο για την παραμόρφωση του αρτηριακού τοιχώματος κατά την ακτινική και αξονική διεύθυνση. Για την εξατομίκευση του μοντέλου και τον προσδιορισμό των παραμέτρων του, πραγματοποιείται προσαρμογή σε πραγματικές μετρήσεις της παραμόρφωσης, που υπολογίζονται με εφαρμογή της μεθόδου ανάλυσης κίνησης σε ακολουθίες εικόνων συγκεκριμένων ασθενών. Ορισμένες παράμετροι του μαθηματικού μοντέλου βρέθηκαν να διαφοροποιούνται σημαντικά μεταξύ των στρωμάτων του αρτηριακού τοιχώματος αλλά και μεταξύ υγιούς τοιχώματος και τοιχώματος με αθηρωμάτωση. Η υλοποίηση των παραπάνω μεθοδολογιών σε συνδυασμό με κατάλληλη διεπιφάνεια χρήσης οδήγησε στην ανάπτυξη ενός ολοκληρωμένου σύστηματος λογισμικού (ANALYSIS), το οποίο στοχεύει στην υποβοήθηση της διάγνωσης της αθηρωμάτωσης της καρωτίδας. Το σύνολο των μεθόδων που παρουσιάζονται στη διδακτορική διατριβή αναμένεται να συμβάλουν αφενός στη μελέτη της μηχανικής συμπεριφοράς του φυσιολογικού και αθηρωματικού αρτηριακού τοιχώματος και αφετέρου στην καθιέρωση μιας πιο αντικειμενικής και αξιόπιστης προσέγγισης για τη διάγνωση της αθηρωμάτωσης και την επιλογή ασθενών υποψήφιων για ενδαρτηρεκτομή. / The purpose of this Ph.D. thesis is to develop and apply advanced image processing and analysis methods to sequences of B-mode ultrasound images of the carotid artery aiming to support the diagnosis of carotid atherosclerosis. The computational methods, presented in the thesis, are applied to both synthetic and real ultrasound data. Based on the findings of a carefully designed study, optimal ultrasound device settings are proposed for reliable motion estimation. Standardized techniques, including image normalization, are also recommended for image processing tasks. Texture analysis of the carotid atheromatous plaque was performed using three transform-based methods (Fourier transform, Wavelet transform and Gabor filters). Texture features estimated using the discrete 2D Wavelet transform and the Gabor filters are found significantly different between symptomatic and asymptomatic subjects. Four different approaches are proposed for the analysis of motion of the carotid artery wall and a validation study is performed using simulated data. The weighted least-squares optical flow method is found to have the best performance. This method is subsequently used to analyze the motion of the healthy carotid artery wall, as well as of the arterial wall with asymptomatic and symptomatic atheromatous plaque. Moreover, an automatic segmentation method based on Hough transform is proposed for the segmentation of the arterial wall from B-mode ultrasound images of longitudinal and transverse sections of the carotid artery. The method can be used for the estimation of widely used diagnostic measures, such as the intima-media thickness and the arterial distension waveforms. The accuracy of the method was reasonably high for longitudinal sections and somewhat lower for transverse sections and diseased arteries. A combination of Hough transform and active contours is proposed to improve the segmentation results in those cases. A mathematical model of the mechanical deformation of the carotid artery wall is also proposed. In an attempt to determine a patient-specific approach, the model is fitted to actual displacement waveforms estimated using the leastsquares optical flow method to B-mode ultrasound image sequences of the carotid artery. A number of model parameters are found significantly different between different layers of the arterial wall and between healthy and diseased wall. The previous methods for the processing and analysis of B-mode ultrasound images of the carotid artery are integrated to a modular software system (ANALYSIS). ANALYSIS can be a useful and powerful tool for the diagnosis of carotid atherosclerosis. The computational methods presented in this thesis are expected to contribute not only to the study of the mechanical behavior of the healthy and diseased carotid artery wall but also to the definition of an objective and reliable approach for the diagnosis of carotid atherosclerosis and the optimal selection of patients for carotid endarterectomy.

In this thesis, two algorithms are introduced for use in ultrasound elastography. Ultrasound elastography is a technique developed in the last 20 years by which anomalous regions in soft tissue are located and diagnosed without the need for biopsy. Due to this, the relativity cheap cost of ultrasound imaging and the high level of accuracy in the methods, ultrasound elastography methods have shown great potential for the diagnosis of cancer in soft tissues. The algorithms introduced in this thesis represent an advance in this field. The first algorithm is a two-step iteration procedure consisting of two minimization problems - displacement estimation and elastic parameter calculation that allow for diagnosis of any anomalous regions within soft tissue. The algorithm represents an improvement on existing methods in several ways. A weighting factor is introduced for each different point in the tissue dependent on the confidence in the accuracy of the data at that point, an exponential substitution is made for the elasticity modulus, an adjoint method is used for efficient calculation of the gradient vector and a total variation regularization technique is used. Most importantly, an adaptive mesh refinement strategy is introduced that allows highly efficient calculation of the elasticity distribution of the tissue though using a number of degrees of freedom several orders lower than methods that use a uniform mesh refinement strategy. Results are presented that show the algorithm is robust even in the presence of significant noise and that it can locate a tumour of 4mm in diameter within a 5cm square region of tissue. Also, the algorithm is extended into 3 dimensions and results are presented that show that it can calculate a 3 dimensional elasticity distribution efficiently. This extension into 3-d is a significant advance in the field. The second algorithm is a one-step algorithm that seeks to combine the two problems of elasticity distribution and displacement calculation into one. As in the two-step algorithm, a weighting factor, exponential substitution for the elasticity parameter, adjoint method for calculation of the gradient vector, total variation regularization and adaptive mesh refinement strategy are incorporated. Results are presented that show that this original approach can locate tumours of varying sizes and shapes in the presence of varying levels of added artificial noise and that it can determine the presence of a tumour in images taken from breast tissue in vivo.