Bone morphology of the femur and tibia captured by statistical shape modelling predicts rapid bone loss in acute spinal cord injury patients

We would like to thank the study participants and the clinical and research staff at the Queen Elizabeth National Spinal Injury Unit, as without them this study would not have been possible. We are grateful for the funding received from Glasgow Research Partnership in Engineering for the employment of SC during data collection for this study. We would like to thank the Royal Society of Edinburgh's Scottish Crucible scheme for providing the opportunity for this collaboration to occur. We are also indebted to Maria Dumitrascuta for her time and effort in producing inter-repeatability results for the shape models.

Avaliação do AAM - Active Appearance Model para detecção de regiões de interesse em radiografias carpais na estimativa da idade óssea através do método TW

In several areas of health professionals (pediatricians, nutritionists, orthopedists, endocrinologists, dentists, etc.) are used in the assessment of bone age to diagnose growth disorders in children. Through interviews with specialists in diagnostic imaging and research done in the literature, we identified the TW method - Tanner and Whitehouse as the most efficient. Even achieving better results than other methods, it is still not the most used, due to the complexity of their use. This work presents the possibility of automation of this method and therefore that its use more widespread. Also in this work, they are met two important steps in the evaluation of bone age, identification and classification of regions of interest. Even in the radiography in which the positioning of the hands were not suitable for TW method, the identification algorithm of the fingers showed good results. As the use AAM - Active Appearance Models showed good results in the identification of regions of interest even in radiographs with high contrast and brightness variation. It has been shown through appearance, good results in the classification of the epiphysis in their stages of development, being chosen the average epiphysis finger III (middle) to show the performance. The final results show an average percentage of 90% hit and misclassified, it was found that the error went away just one stage of the correct stage.

Automatically detecting pain in video through facial action units

In a clinical setting, pain is reported either through patient self-report or via an observer. Such measures are problematic as they are: 1) subjective, and 2) give no specific timing information. Coding pain as a series of facial action units (AUs) can avoid these issues as it can be used to gain an objective measure of pain on a frame-by-frame basis. Using video data from patients with shoulder injuries, in this paper, we describe an active appearance model (AAM)-based system that can automatically detect the frames in video in which a patient is in pain. This pain data set highlights the many challenges associated with spontaneous emotion detection, particularly that of expression and head movement due to the patient's reaction to pain. In this paper, we show that the AAM can deal with these movements and can achieve significant improvements in both the AU and pain detection performance compared to the current-state-of-the-art approaches which utilize similarity-normalized appearance features only.

In the pursuit of effective affective computing : the relationship between features and registration

For facial expression recognition systems to be applicable in the real world, they need to be able to detect and track a previously unseen person's face and its facial movements accurately in realistic environments. A highly plausible solution involves performing a "dense" form of alignment, where 60-70 fiducial facial points are tracked with high accuracy. The problem is that, in practice, this type of dense alignment had so far been impossible to achieve in a generic sense, mainly due to poor reliability and robustness. Instead, many expression detection methods have opted for a "coarse" form of face alignment, followed by an application of a biologically inspired appearance descriptor such as the histogram of oriented gradients or Gabor magnitudes. Encouragingly, recent advances to a number of dense alignment algorithms have demonstrated both high reliability and accuracy for unseen subjects [e.g., constrained local models (CLMs)]. This begs the question: Aside from countering against illumination variation, what do these appearance descriptors do that standard pixel representations do not? In this paper, we show that, when close to perfect alignment is obtained, there is no real benefit in employing these different appearance-based representations (under consistent illumination conditions). In fact, when misalignment does occur, we show that these appearance descriptors do work well by encoding robustness to alignment error. For this work, we compared two popular methods for dense alignment-subject-dependent active appearance models versus subject-independent CLMs-on the task of action-unit detection. These comparisons were conducted through a battery of experiments across various publicly available data sets (i.e., CK+, Pain, M3, and GEMEP-FERA). We also report our performance in the recent 2011 Facial Expression Recognition and Analysis Challenge for the subject-independent task.

Towards spontaneous facial expression recognition in real-world video

Facial expression is an important channel of human social communication. Facial expression recognition (FER) aims to perceive and understand emotional states of humans based on information in the face. Building robust and high performance FER systems that can work in real-world video is still a challenging task, due to the various unpredictable facial variations and complicated exterior environmental conditions, as well as the difficulty of choosing a suitable type of feature descriptor for extracting discriminative facial information. Facial variations caused by factors such as pose, age, gender, race and occlusion, can exert profound influence on the robustness, while a suitable feature descriptor largely determines the performance. Most present attention on FER has been paid to addressing variations in pose and illumination. No approach has been reported on handling face localization errors and relatively few on overcoming facial occlusions, although the significant impact of these two variations on the performance has been proved and highlighted in many previous studies. Many texture and geometric features have been previously proposed for FER. However, few comparison studies have been conducted to explore the performance differences between different features and examine the performance improvement arisen from fusion of texture and geometry, especially on data with spontaneous emotions. The majority of existing approaches are evaluated on databases with posed or induced facial expressions collected in laboratory environments, whereas little attention has been paid on recognizing naturalistic facial expressions on real-world data. This thesis investigates techniques for building robust and high performance FER systems based on a number of established feature sets. It comprises of contributions towards three main objectives: (1) Robustness to face localization errors and facial occlusions. An approach is proposed to handle face localization errors and facial occlusions using Gabor based templates. Template extraction algorithms are designed to collect a pool of local template features and template matching is then performed to covert these templates into distances, which are robust to localization errors and occlusions. (2) Improvement of performance through feature comparison, selection and fusion. A comparative framework is presented to compare the performance between different features and different feature selection algorithms, and examine the performance improvement arising from fusion of texture and geometry. The framework is evaluated for both discrete and dimensional expression recognition on spontaneous data. (3) Evaluation of performance in the context of real-world applications. A system is selected and applied into discriminating posed versus spontaneous expressions and recognizing naturalistic facial expressions. A database is collected from real-world recordings and is used to explore feature differences between standard database images and real-world images, as well as between real-world images and real-world video frames. The performance evaluations are based on the JAFFE, CK, Feedtum, NVIE, Semaine and self-collected QUT databases. The results demonstrate high robustness of the proposed approach to the simulated localization errors and occlusions. Texture and geometry have different contributions to the performance of discrete and dimensional expression recognition, as well as posed versus spontaneous emotion discrimination. These investigations provide useful insights into enhancing robustness and achieving high performance of FER systems, and putting them into real-world applications.

Fourier Lucas-Kanade Algorithm

In this paper we propose a framework for both gradient descent image and object alignment in the Fourier domain. Our method centers upon the classical Lucas & Kanade (LK) algorithm where we represent the source and template/model in the complex 2D Fourier domain rather than in the spatial 2D domain. We refer to our approach as the Fourier LK (FLK) algorithm. The FLK formulation is advantageous when one pre-processes the source image and template/model with a bank of filters (e.g. oriented edges, Gabor, etc.) as: (i) it can handle substantial illumination variations, (ii) the inefficient pre-processing filter bank step can be subsumed within the FLK algorithm as a sparse diagonal weighting matrix, (iii) unlike traditional LK the computational cost is invariant to the number of filters and as a result far more efficient, and (iv) this approach can be extended to the inverse compositional form of the LK algorithm where nearly all steps (including Fourier transform and filter bank pre-processing) can be pre-computed leading to an extremely efficient and robust approach to gradient descent image matching. Further, these computational savings translate to non-rigid object alignment tasks that are considered extensions of the LK algorithm such as those found in Active Appearance Models (AAMs).

Robust recognition of human behaviour in challenging environments

Novel techniques have been developed for the automatic recognition of human behaviour in challenging environments using information from visual and infra-red camera feeds. The techniques have been applied to two interesting scenarios: Recognise drivers' speech using lip movements and recognising audience behaviour, while watching a movie, using facial features and body movements. Outcome of the research in these two areas will be useful in the improving the performance of voice recognition in automobiles for voice based control and for obtaining accurate movie interest ratings based on live audience response analysis.

Lightcurve inversion for asteroid spins and shapes

Knowledge of the physical properties of asteroids is crucial in many branches of solar-system research. Knowledge of the spin states and shapes is needed, e.g., for accurate orbit determination and to study the history and evolution of the asteroids. In my thesis, I present new methods for using photometric lightcurves of asteroids in the determination of their spin states and shapes. The convex inversion method makes use of a general polyhedron shape model and provides us at best with an unambiguous spin solution and a convex shape solution that reproduces the main features of the original shape. Deriving information about the non-convex shape features is, in principle, also possible, but usually requires a priori information about the object. Alternatively, a distribution of non-convex solutions, describing the scale of the non-convexities, is also possible to be obtained. Due to insufficient number of absolute observations and inaccurately defined asteroid phase curves, the $c/b$-ratio, i.e., the flatness of the shape model is often somewhat ill-defined. However, especially in the case of elongated objects, the flatness seems to be quite well constrained, even in the case when only relative lightcurves are available. The results prove that it is, contrary to the earlier misbelief, possible to derive shape information from the lightcurve data if a sufficiently wide range of observing geometries is covered by the observations. Along with the more accurate shape models, also the rotational states, i.e., spin vectors and rotation periods, are defined with improved accuracy. The shape solutions obtained so far reveal a population of irregular objects whose most descriptive shape characteristics, however, can be expressed with only a few parameters. Preliminary statistical analyses for the shapes suggests that there are correlations between shape and other physical properties, such as the size, rotation period and taxonomic type of the asteroids. More shape data of, especially, the smallest and largest asteroids, as well as the fast and slow rotators is called for in order to be able to study the statistics more thoroughly.

A hybrid ASM approach for sparse volumetric data segmentation

Y. Zhu, S. Williams and R. Zwiggelaar, 'A hybrid ASM approach for sparse volumetric data segmentation', Pattern Recognition and Image Analysis 17 (2), 252-258 (2007)

Segmentation of Brain Tissue from Magnetic Resonance Images

Segmentation of medical imagery is a challenging problem due to the complexity of the images, as well as to the absence of models of the anatomy that fully capture the possible deformations in each structure. Brain tissue is a particularly complex structure, and its segmentation is an important step for studies in temporal change detection of morphology, as well as for 3D visualization in surgical planning. In this paper, we present a method for segmentation of brain tissue from magnetic resonance images that is a combination of three existing techniques from the Computer Vision literature: EM segmentation, binary morphology, and active contour models. Each of these techniques has been customized for the problem of brain tissue segmentation in a way that the resultant method is more robust than its components. Finally, we present the results of a parallel implementation of this method on IBM's supercomputer Power Visualization System for a database of 20 brain scans each with 256x256x124 voxels and validate those against segmentations generated by neuroanatomy experts.

Fractal geometry of aggregate snowflakes revealed by triple wavelength radar measurements

Radar reflectivity measurements from three different wavelengths are used to retrieve information about the shape of aggregate snowflakes in deep stratiform ice clouds. Dual-wavelength ratios are calculated for different shape models and compared to observations at 3, 35 and 94 GHz. It is demonstrated that many scattering models, including spherical and spheroidal models, do not adequately describe the aggregate snowflakes that are observed. The observations are consistent with fractal aggregate geometries generated by a physically-based aggregation model. It is demonstrated that the fractal dimension of large aggregates can be inferred directly from the radar data. Fractal dimensions close to 2 are retrieved, consistent with previous theoretical models and in-situ observations.

Analysis of stamping production data with view towards quality management

A quality analysis trial was undertaken at Ford Geelong Stamping Plant on a press line that was fitted with standard press sensors to measure press and binder force over the stamping cycle for each panel. The quality of randomly sampled panels was measured by obtaining the panel thicknesses at five points, for 135 panels in total. These points were chosen such that they exhibited different forming modes. This paper analyses the input force data and the output quality data from the trial to determine any potential relationships. The analysis of the production data was performed using statistical correlation techniques to determine initial potential relationships between input and output variables. An Active Shape Model was used to extract features when identifying the major sources of variation within the input data. However, the initial analysis of the data elicited no direct relationship between the input variables measured and the panel thicknesses. This result is significant as the data collected is from a standard sensor configuration found in many press lines through-out the world. The reason for the lack of a direct relationship is believed to come from the lack of sensitivity in the force measurements which are not able to identify small changes in the process, whereas gross geometric variations have in previous studies shown an obvious relationship with changes in the force press profile. This means that existing force sensors require augmentation by additional sensors if a detailed automatic quality control system for the press lines based on input sensors alone.

Does the presence and mix of destinations influence walking and physical activity?

Background: Local destinations have previously been shown to be associated with higher levels of both physical activity and walking, but little is known about how specific destinations are related to activity. This study examined associations between types and mix of destinations and both walking frequency and physical activity. Method: The sample consisted of 2349 residents of 50 urban areas in metropolitan Melbourne, Australia. Using geographic information systems, seven types of destinations were examined within three network buffers (400 meters (m), 800 m and 1200 m) of respondents' homes. Multilevel logistic regression was used to estimate effects of each destination type separately, as well as destination mix (variety) on: 1) likelihood of walking for at least 10 min ≥ 4/week; 2) likelihood of being sufficiently physically active. All models were adjusted for potential confounders. Results: All destination types were positively associated with walking frequency, and physical activity sufficiency at 1200 m. For the 800 m buffer: all destinations except transport stops and sports facilities were significantly associated with physical activity, while all except sports facilities were associated with walking frequency; at 400 m, café/takeaway food stores and transport stops were associated with walking frequency and physical activity sufficiency, and sports facilities were also associated with walking frequency. Strongest associations for both outcomes were observed for community resources and small food stores at both 800 m and 1200 m. For all buffer distances: greater mix was associated with greater walking frequency. Inclusion of walking in physical activity models led to attenuation of associations. Conclusions: The results of this analysis indicate that there is an association between destinations and both walking frequency and physical activity sufficiency, and that this relationship varies by destination type. It is also clear that greater mix of destinations positively predicts walking frequency and physical activity sufficiency.

Determination of Pelvic Orientation from Ultrasound Images Using Patch-SSMs and a Hierarchical Speed of Sound Compensation Strategy

In the field of computer assisted orthopedic surgery (CAOS) the anterior pelvic plane (APP) is a common concept to determine the pelvic orientation by digitizing distinct pelvic landmarks. As percutaneous palpation is - especially for obese patients - known to be error-prone, B-mode ultrasound (US) imaging could provide an alternative means. Several concepts of using ultrasound imaging to determine the APP landmarks have been introduced. In this paper we present a novel technique, which uses local patch statistical shape models (SSMs) and a hierarchical speed of sound compensation strategy for an accurate determination of the APP. These patches are independently matched and instantiated with respect to associated point clouds derived from the acquired ultrasound images. Potential inaccuracies due to the assumption of a constant speed of sound are compensated by an extended reconstruction scheme. We validated our method with in-vitro studies using a plastic bone covered with a soft-tissue simulation phantom and with a preliminary cadaver trial.