Supervised latent dirichlet allocation models for efficient activity representation

Local spatio-temporal features with a Bag-of-visual words model is a popular approach used in human action recognition. Bag-of-features methods suffer from several challenges such as extracting appropriate appearance and motion features from videos, converting extracted features appropriate for classification and designing a suitable classification framework. In this paper we address the problem of efficiently representing the extracted features for classification to improve the overall performance. We introduce two generative supervised topic models, maximum entropy discrimination LDA (MedLDA) and class- specific simplex LDA (css-LDA), to encode the raw features suitable for discriminative SVM based classification. Unsupervised LDA models disconnect topic discovery from the classification task, hence yield poor results compared to the baseline Bag-of-words framework. On the other hand supervised LDA techniques learn the topic structure by considering the class labels and improve the recognition accuracy significantly. MedLDA maximizes likelihood and within class margins using max-margin techniques and yields a sparse highly discriminative topic structure; while in css-LDA separate class specific topics are learned instead of common set of topics across the entire dataset. In our representation first topics are learned and then each video is represented as a topic proportion vector, i.e. it can be comparable to a histogram of topics. Finally SVM classification is done on the learned topic proportion vector. We demonstrate the efficiency of the above two representation techniques through the experiments carried out in two popular datasets. Experimental results demonstrate significantly improved performance compared to the baseline Bag-of-features framework which uses kmeans to construct histogram of words from the feature vectors.

Whole-brain imaging with single-cell resolution using chemical cocktails and computational analysis

Systems-level identification and analysis of cellular circuits in the brain will require the development of whole-brain imaging with single-cell resolution. To this end, we performed comprehensive chemical screening to develop a whole-brain clearing and imaging method, termed CUBIC (clear, unobstructed brain imaging cocktails and computational analysis). CUBIC is a simple and efficient method involving the immersion of brain samples in chemical mixtures containing aminoalcohols, which enables rapid whole-brain imaging with single-photon excitation microscopy. CUBIC is applicable to multicolor imaging of fluorescent proteins or immunostained samples in adult brains and is scalable from a primate brain to subcellular structures. We also developed a whole-brain cell-nuclear counterstaining protocol and a computational image analysis pipeline that, together with CUBIC reagents, enable the visualization and quantification of neural activities induced by environmental stimulation. CUBIC enables time-course expression profiling of whole adult brains with single-cell resolution.

Analysis and evaluation of use of the V1 and South East Freeway Bicycleway post completion of Stage C of the V1

Executive Summary: Completion of the Veloway 1 (V1) will provide a dedicated and safe route for cyclists between the Brisbane CBD and the Gateway Motorway off-ramp at Eight Mile Plains alongside the South East Motorway. The V1 is being delivered in stages and when completed will provide a dedicated 3m wide cycleway 17km in length. Two stages (D and E) remain to be constructed to complete the V1. Major trip attractors along the V1 include the Mater, Princes Alexandra and Greenslopes Hospitals, two campuses of Griffith University, Garden City shopping centre and the Australian Tax Office. This report assesses the available evidence on the impacts on cycling behaviour of the recently completed V1 Stage C. The data sources informing this review include three intercept surveys, motion activated traffic cameras and travel time surveys on the V1 and adjoining South East Freeway Bikeway (SEFB), Strava app data, and cyclist crash data along Logan Road. The key findings from the evidence are that the completed V1 Stage C has: a Attracted cyclists from Holland Park, Holland Park West, Mt Gravatt and southern parts of Tarragindi onto the V1 Stage C. b Reduced the crash exposure of pedestrians to cyclists by attracting higher speed cyclists off the adjoining SEFB onto the cycling dedicated V1 Stage C. c Reduced the potential crash exposure of cyclists to motor vehicles by attracting cyclists off Logan Road on to the V1. d Provided travel time benefits to cyclists and reduced road crossings (eight down to two). e Predominantly attracted adults commuting alone to and from work and university. The evidence shows that the two traffic crossings across Birdwood Road (required as a temporary measure until the V1 is completed) negate much of the travel time gains of the V1 Stage C compared to the adjoining SEFB for southbound cyclists. Many cyclists accessing the V1 Stage C from the south are cycling in high-volume vehicular traffic lanes to reduce their travel time along Birdwood Road, but in the process are increasing their exposure to crashes with motor vehicles. Based on these findings this report recommends that TMR: a. Continue with plans to complete the V1 Veloway b. Undertake an engineering feasibility assessment to determine the viability of constructing a section of the V1 Stage E from the intersection Weller and Birdwood Roads over Marshall Road and along Bapaume Road on the western side of the Motorway to the intersection of Bapaume and Sterculia Roads. c. In the interim, improve signage and Birdwood Road crossing points for cyclists accessing and egressing the southern end of the V1 Stage C. d. Work with Brisbane City Council to identify the safest and most practical bicycle facilities to facilitate cycle travel between Logan Road and the V1 south of Birdwood Road. e. Improve the awareness of the V1 Stage C through signage for cyclists approaching from the north with the aim of providing a better understanding of the route of the V1 to the south. f. Refine the use of motion activated traffic cameras to improve the capture rate of useable images and obtain an ongoing collection over time of V1 usage data. g. Undertake discussions with Strava, Inc. to refine the presentation of Strava data to improve visual understanding of maps showing before and after cycle route volumes along and on roads leading to the V1.

Bayesian computational methods for spatial analysis of images

This thesis introduces a new way of using prior information in a spatial model and develops scalable algorithms for fitting this model to large imaging datasets. These methods are employed for image-guided radiation therapy and satellite based classification of land use and water quality. This study has utilized a pre-computation step to achieve a hundredfold improvement in the elapsed runtime for model fitting. This makes it much more feasible to apply these models to real-world problems, and enables full Bayesian inference for images with a million or more pixels.

Made by motion: A conceptual framework for abstracted animation derived from motion-captured movements

Despite an increasing number of acclaimed abstract animations being created through the application of motion capture technologies there has been little detailed documentation and analysis of this approach for abstract animation production. More specifically, it is unclear what the key considerations are, and what issues practitioners might face, when integrating motion capture movement data into their practice. In response to this issue this study explored and documented the practice of generating abstract visual and temporal artefacts from motion captured dance movements that compose abstract animated short films. The study has resulted in a possible framework for this form of practice and outlines five key considerations which should be taken into account by practitioners who use motion capture in the production of abstract animated short films.

Left ventricular volume and valve gradient analysis using an Apple Macintosh computer

A description of a computer program to analyse cine angiograms of the heart and pressure waveforms to calculate valve gradients.

Analysis of arm elevation muscle activity through different movement planes and speeds during in-water and dry-land exercise

Objective The objectives of this cross-sectional, analytical inference analysis were to compare shoulder muscle activation at arm elevations of 0° to 90° through different movement planes and speeds during in-water and dry-land exercise and to extrapolate this information to a clinical rehabilitation model. Methods Six muscles of right-handed adult subjects (n = 16; males/females: 50%; age: 26.1 ± 4.5 years) were examined with surface electromyography during arm elevation in water and on dry land. Participants randomly performed 3 elevation movements (flexion, abduction, and scaption) through 0° to 90°. Three movement speeds were used for each movement as determined by a metronome (30°/sec, 45°/sec, and 90°/sec). Dry-land maximal voluntary contraction tests were used to determine movement normalization. Results Muscle activity levels were significantly lower in water compared with dry land at 30°/sec and 45°/sec but significantly higher at 90°/sec. This sequential progressive activation with increased movement speed was proportionally higher on transition from gravity-based on-land activity to water-based isokinetic resistance. The pectoralis major and latissimus dorsi muscles showed higher activity during abduction and scaption. Conclusions These findings on muscle activation suggest protocols in which active flexion is introduced first at low speeds (30°/sec) in water, then at medium speeds (45°/sec) in water or on dry land, and finally at high speeds (90°/sec) on dry land before in water. Abduction requires higher stabilization, necessitating its introduction after flexion, with scaption introduced last. This model of progressive sequential movement ensures that early active motion and then stabilization are appropriately introduced. This should reduce rehabilitation time and improve therapeutic goals without compromising patient safety or introducing inappropriate muscle recruitment or movement speed.

Development of a new ultrasound-based system for tracking motion of the human lumbar spine: Reliability, stability and repeatability during forward bending movement trials

The aim of this study was to develop a new method for quantifying intersegmental motion of the spine in an instrumented motion segment L4–L5 model using ultrasound image post-processing combined with an electromagnetic device. A prospective test–retest design was employed, combined with an evaluation of stability and within- and between-day intra-tester reliability during forward bending by 15 healthy male patients. The accuracy of the measurement system using the model was calculated to be ± 0.9° (standard deviation = 0.43) over a 40° range and ± 0.4 cm (standard deviation = 0.28) over 1.5 cm. The mean composite range of forward bending was 15.5 ± 2.04° during a single trial (standard error of the mean = 0.54, coefficient of variation = 4.18). Reliability (intra-class correlation coefficient = 2.1) was found to be excellent for both within-day measures (0.995–0.999) and between-day measures (0.996–0.999). Further work is necessary to explore the use of this approach in the evaluation of biomechanics, clinical assessments and interventions.

Extending genetic linkage analysis to diffusion tensor images to map single gene effects on brain fiber architecture

We extended genetic linkage analysis - an analysis widely used in quantitative genetics - to 3D images to analyze single gene effects on brain fiber architecture. We collected 4 Tesla diffusion tensor images (DTI) and genotype data from 258 healthy adult twins and their non-twin siblings. After high-dimensional fluid registration, at each voxel we estimated the genetic linkage between the single nucleotide polymorphism (SNP), Val66Met (dbSNP number rs6265), of the BDNF gene (brain-derived neurotrophic factor) with fractional anisotropy (FA) derived from each subject's DTI scan, by fitting structural equation models (SEM) from quantitative genetics. We also examined how image filtering affects the effect sizes for genetic linkage by examining how the overall significance of voxelwise effects varied with respect to full width at half maximum (FWHM) of the Gaussian smoothing applied to the FA images. Raw FA maps with no smoothing yielded the greatest sensitivity to detect gene effects, when corrected for multiple comparisons using the false discovery rate (FDR) procedure. The BDNF polymorphism significantly contributed to the variation in FA in the posterior cingulate gyrus, where it accounted for around 90-95% of the total variance in FA. Our study generated the first maps to visualize the effect of the BDNF gene on brain fiber integrity, suggesting that common genetic variants may strongly determine white matter integrity.

Information-theoretic analysis of brain white matter fiber orientation distribution functions

We propose a new information-theoretic metric, the symmetric Kullback-Leibler divergence (sKL-divergence), to measure the difference between two water diffusivity profiles in high angular resolution diffusion imaging (HARDI). Water diffusivity profiles are modeled as probability density functions on the unit sphere, and the sKL-divergence is computed from a spherical harmonic series, which greatly reduces computational complexity. Adjustment of the orientation of diffusivity functions is essential when the image is being warped, so we propose a fast algorithm to determine the principal direction of diffusivity functions using principal component analysis (PCA). We compare sKL-divergence with other inner-product based cost functions using synthetic samples and real HARDI data, and show that the sKL-divergence is highly sensitive in detecting small differences between two diffusivity profiles and therefore shows promise for applications in the nonlinear registration and multisubject statistical analysis of HARDI data.

Multi-site genetic analysis of diffusion images and voxelwise heritability analysis: A pilot project of the ENIGMA-DTI working group

The ENIGMA (Enhancing NeuroImaging Genetics through Meta-Analysis) Consortium was set up to analyze brain measures and genotypes from multiple sites across the world to improve the power to detect genetic variants that influence the brain. Diffusion tensor imaging (DTI) yields quantitative measures sensitive to brain development and degeneration, and some common genetic variants may be associated with white matter integrity or connectivity. DTI measures, such as the fractional anisotropy (FA) of water diffusion, may be useful for identifying genetic variants that influence brain microstructure. However, genome-wide association studies (GWAS) require large populations to obtain sufficient power to detect and replicate significant effects, motivating a multi-site consortium effort. As part of an ENIGMA-DTI working group, we analyzed high-resolution FA images from multiple imaging sites across North America, Australia, and Europe, to address the challenge of harmonizing imaging data collected at multiple sites. Four hundred images of healthy adults aged 18-85 from four sites were used to create a template and corresponding skeletonized FA image as a common reference space. Using twin and pedigree samples of different ethnicities, we used our common template to evaluate the heritability of tract-derived FA measures. We show that our template is reliable for integrating multiple datasets by combining results through meta-analysis and unifying the data through exploratory mega-analyses. Our results may help prioritize regions of the FA map that are consistently influenced by additive genetic factors for future genetic discovery studies. Protocols and templates are publicly available at (http://enigma.loni.ucla.edu/ongoing/dti-working-group/).

A genetic analysis of cortical thickness in 372 twins

Imaging genetics is a new field of neuroscience that blends methods from computational anatomy and quantitative genetics to identify genetic influences on brain structure and function. Here we analyzed brain MRI data from 372 young adult twins to identify cortical regions in which gray matter volume is influenced by genetic differences across subjects. Thickness maps, reconstructed from surface models of the cortical gray/white and gray/CSF interfaces, were smoothed with a 25 mm FWHM kernel and automatically parcellated into 34 regions of interest per hemisphere. In structural equation models fitted to volume values at each surface vertex, we computed components of variance due to additive genetic (A), shared (C) and unique (E) environmental factors, and tested their significance. Cortical regions in the vicinity of the perisylvian language cortex, and at the frontal and temporal poles, showed significant additive genetic variance, suggesting that volume measures from these regions may provide quantitative phenotypes to narrow the search for quantitative trait loci that influence brain structure.

Tensor-based analysis of genetic influences on brain integrity using DTI in 100 twins

Information from the full diffusion tensor (DT) was used to compute voxel-wise genetic contributions to brain fiber microstructure. First, we designed a new multivariate intraclass correlation formula in the log-Euclidean framework. We then analyzed used the full multivariate structure of the tensor in a multivariate version of a voxel-wise maximum-likelihood structural equation model (SEM) that computes the variance contributions in the DTs from genetic (A), common environmental (C) and unique environmental (E) factors. Our algorithm was tested on DT images from 25 identical and 25 fraternal twin pairs. After linear and fluid registration to a mean template, we computed the intraclass correlation and Falconer's heritability statistic for several scalar DT-derived measures and for the full multivariate tensors. Covariance matrices were found from the DTs, and inputted into SEM. Analyzing the full DT enhanced the detection of A and C effects. This approach should empower imaging genetics studies that use DTI.

MR image-based measurement of rates of change in volumes of brain structures. Part II: Application to a study of Alzheimer's disease and normal aging

We present global and regional rates of brain atrophy measured on serially acquired Tl-weighted brain MR images for a group of Alzheimer's disease (AD) patients and age-matched normal control (NC) subjects using the analysis procedure described in Part I. Three rates of brain atrophy: the rate of atrophy in the cerebrum, the rate of lateral ventricular enlargement and the rate of atrophy in the region of temporal lobes, were evaluated for 14 AD patients and 14 age-matched NC subjects. All three rates showed significant differences between the two groups. However, the greatest separation of the two groups was obtained when the regional rates were combined. This application has demonstrated that rates of brain atrophy, especially in specific regions of the brain, based on MR images can provide sensitive measures for evaluating the progression of AD. These measures will be useful for the evaluation of therapeutic effects of novel therapies for AD.

Investigation of the relationship between BMI and prostate motion in radiation therapy treatment for prostate cancer

This thesis studied the influence of patient obesity on prostate motion during radiation therapy treatment delivery, an important consideration in the accurate treatment of prostate cancer. The study highlighted the importance of daily image guidance to correct for prostate motion, increasing radiation dose to the prostate while decreasing radiation dose to surrounding healthy tissues, thereby increasing patient quality of life.