Pixels2Pose: Super-resolution time-of-flight imaging for 3D pose estimation.

Single-photon-sensitive depth sensors are being increasingly used in next-generation electronics for human pose and gesture recognition. However, cost-effective sensors typically have a low spatial resolution, restricting their use to basic motion identification and simple object detection. Here, we perform a temporal to spatial mapping that drastically increases the resolution of a simple time-of-flight sensor, i.e., an initial resolution of 4 × 4 pixels to depth images of resolution 32 × 32 pixels. The output depth maps can then be used for accurate three-dimensional human pose estimation of multiple people. We develop a new explainable framework that provides intuition to how our network uses its input data and provides key information about the relevant parameters. Our work greatly expands the use cases of simple single-photon avalanche detector time-of-flight sensors and opens up promising possibilities for future super-resolution techniques applied to other types of sensors with similar data types, i.e., radar and sonar.

Observation of laser pulse propagation in optical fibers with a SPAD camera.

Recording processes and events that occur on sub-nanosecond timescales poses a difficult challenge. Conventional ultrafast imaging techniques often rely on long data collection times, which can be due to limited device sensitivity and/or the requirement of scanning the detection system to form an image. In this work, we use a single-photon avalanche detector array camera with pico-second timing accuracy to detect photons scattered by the cladding in optical fibers. We use this method to film supercontinuum generation and track a GHz pulse train in optical fibers. We also show how the limited spatial resolution of the array can be improved with computational imaging. The single-photon sensitivity of the camera and the absence of scanning the detection system results in short total acquisition times, as low as a few seconds depending on light levels. Our results allow us to calculate the group index of different wavelength bands within the supercontinuum generation process. This technology can be applied to a range of applications, e.g., the characterization of ultrafast processes, time-resolved fluorescence imaging, three-dimensional depth imaging, and tracking hidden objects around a corner.

Fast Unsupervised Bayesian Image Segmentation With Adaptive Spatial Regularisation.

This paper presents a new Bayesian estimation technique for hidden Potts-Markov random fields with unknown regularisation parameters, with application to fast unsupervised K -class image segmentation. The technique is derived by first removing the regularisation parameter from the Bayesian model by marginalisation, followed by a small-variance-asymptotic (SVA) analysis in which the spatial regularisation and the integer-constrained terms of the Potts model are decoupled. The evaluation of this SVA Bayesian estimator is then relaxed into a problem that can be computed efficiently by iteratively solving a convex total-variation denoising problem and a least-squares clustering ( K -means) problem, both of which can be solved straightforwardly, even in high-dimensions, and with parallel computing techniques. This leads to a fast fully unsupervised Bayesian image segmentation methodology in which the strength of the spatial regularisation is adapted automatically to the observed image during the inference procedure, and that can be easily applied in large 2D and 3D scenarios or in applications requiring low computing times. Experimental results on synthetic and real images, as well as extensive comparisons with state-of-the-art algorithms, confirm that the proposed methodology offer extremely fast convergence and produces accurate segmentation results, with the important additional advantage of self-adjusting regularisation parameters.

Lidar Waveform-Based Analysis of Depth Images Constructed Using Sparse Single-Photon Data.

This paper presents a new Bayesian model and algorithm used for depth and reflectivity profiling using full waveforms from the time-correlated single-photon counting measurement in the limit of very low photon counts. The proposed model represents each Lidar waveform as a combination of a known impulse response, weighted by the target reflectivity, and an unknown constant background, corrupted by Poisson noise. Prior knowledge about the problem is embedded through prior distributions that account for the different parameter constraints and their spatial correlation among the image pixels. In particular, a gamma Markov random field (MRF) is used to model the joint distribution of the target reflectivity, and a second MRF is used to model the distribution of the target depth, which are both expected to exhibit significant spatial correlations. An adaptive Markov chain Monte Carlo algorithm is then proposed to perform Bayesian inference. This algorithm is equipped with a stochastic optimization adaptation mechanism that automatically adjusts the parameters of the MRFs by maximum marginal likelihood estimation. Finally, the benefits of the proposed methodology are demonstrated through a series of experiments using real data.

Exploiting information geometry to improve the convergence of nonparametric active contours.

This paper presents a fast converging Riemannian steepest descent method for nonparametric statistical active contour models, with application to image segmentation. Unlike other fast algorithms, the proposed method is general and can be applied to any statistical active contour model from the exponential family, which comprises most of the models considered in the literature. This is achieved by first identifying the intrinsic statistical manifold associated with this class of active contours, and then constructing a steepest descent on that manifold. A key contribution of this paper is to derive a general and tractable closed-form analytic expression for the manifold's Riemannian metric tensor, which allows computing discrete gradient flows efficiently. The proposed methodology is demonstrated empirically and compared with other state of the art approaches on several standard test images, a phantom positron-emission-tomography scan and a B-mode echography of in-vivo human dermis.

Residual component analysis of hyperspectral images--application to joint nonlinear unmixing and nonlinearity detection.

This paper presents a nonlinear mixing model for joint hyperspectral image unmixing and nonlinearity detection. The proposed model assumes that the pixel reflectances are linear combinations of known pure spectral components corrupted by an additional nonlinear term, affecting the end members and contaminated by an additive Gaussian noise. A Markov random field is considered for nonlinearity detection based on the spatial structure of the nonlinear terms. The observed image is segmented into regions where nonlinear terms, if present, share similar statistical properties. A Bayesian algorithm is proposed to estimate the parameters involved in the model yielding a joint nonlinear unmixing and nonlinearity detection algorithm. The performance of the proposed strategy is first evaluated on synthetic data. Simulations conducted with real data show the accuracy of the proposed unmixing and nonlinearity detection strategy for the analysis of hyperspectral images.

Rural clinical experiences for emergency medicine residents: a curriculum template.

Rural emergency departments (EDs) in the United States are less likely to be staffed with emergency medicine (EM) residency-trained and American Board of Emergency Medicine (ABEM)-certified physicians than urban EDs. Rural EM clinical experiences during residency training have been suggested as a strategy to encourage future rural practice, but past Accreditation Council for Graduate Medical Education (ACGME) Residency Review Committee for Emergency Medicine program requirements and a lack of familiarity with rural rotations in the EM graduate medical education (GME) community have limited their availability. To provide a template for the development and implementation of a rural EM clinical experience, Kern's six-step approach was followed.

Simulation center accreditation and programmatic benchmarks: a review for emergency medicine.

Simulation-based education has grown significantly over the past 10 years. As a result, more professional organizations are developing or implementing accreditation processes to help define minimum standards and best practices in simulation-based training. However, the benefits and potential pitfalls of sponsoring and implementing such programs have yet to be fully evaluated across specialties. The board of directors of the Society for Academic Emergency Medicine (SAEM) requested an evaluation of the potential to create an emergency medicine (EM)-based Simulation Consultation and Accreditation Service. In response to this request, the Simulation Accreditation and Consultation Work Group, a subgroup of the Committee on Technology in Medical Education (now Simulation Academy), was created. The work group was charged with: 1) reviewing current benchmarks and standards set by existing simulation accreditation programs; 2) analyzing current EM simulation program structures, including leadership, administrative, and financial components; and 3) proposing a potential model for EM-based simulation accreditation. This article outlines currently existing and proposed accreditation models and identifies components that support best practices. It then goes on to describe three general programmatic models to better understand how simulation training can be operationalized in EM. Finally, the work group uses this collective information to propose how an accreditation process, in concert with the SAEM Simulation Consultation Service, can enhance and advance EM simulation training.

Segmented motion compensation for complementary coded ultrasonic imaging.

Ultrasonic imaging using complementary coded pulses offers the SNR improvements of signal coding without the filter side-lobes introduced by single-transmit codes. Tissue motion between coded pulse emissions, however, can introduce high side-lobes caused by misalignment of complementary filter outputs. This paper presents a method for filtering and motion compensation of complementary coded signals appropriate for use in medical imaging. The method is robust to the effects of non-ideal transducers on the imaging signals, includes mirrored compensation stages to reduce the impact of motion estimation error, and has been shown to reduce side-lobes to levels that compare favorably to systems using FM-coded signals of similar length and bandwidth while providing increased coding gain and range resolution. In addition, motion compensation allows the received data to be used without the frame-rate penalty usually incurred by complementary-coded imaging. The method has been verified using simulated point and speckle targets with both homogeneous and inhomogeneous motion profiles. Selected results have been verified experimentally.

Time-frequency and advanced frequency estimation techniques for the investigation of bat echolocation calls.

In this paper, techniques for time-frequency analysis and investigation of bat echolocation calls are studied. Particularly, enhanced resolution techniques are developed and/or used in this specific context for the first time. When compared to traditional time-frequency representation methods, the proposed techniques are more capable of showing previously unseen features in the structure of bat echolocation calls. It should be emphasized that although the study is focused on bat echolocation recordings, the results are more general and applicable to many other types of signal.

Towards establishing a national colorectal cancer database: lessons learnt from Bio21 molecular medicine informatics model.

BACKGROUND: Collecting data regarding treatment and outcomes of patients with cancer, for both audit and research purposes, is a common but challenging goal. Modern technology promises greater ease and sophistication for data collection, linkage and analysis, but many traditional challenges remain. METHOD: Here we relate our experience of an initiative aimed at multicentre colorectal cancer data collection, that is, in collaboration with the Colorectal Surgical Society Australia and New Zealand, moving towards a national initiative. RESULTS: Initiated from a single site in Melbourne, using a range of data collection and linkage processes, and optimizing the use of modern technology, we have now implemented and sustained comprehensive and multidisciplinary data collection across multiple Victorian and interstate institutions. Specific challenges related to ethics and privacy, data accuracy and completeness and data ownership have been addressed and many lessons have been learnt. Multicentre audit and research queries can now be conducted with confidence that privacy, security and intellectual property issues are addressed. Research output, including many studies that were not previously possible, has informed a broad range of topics relevant to colorectal cancer. CONCLUSION: Multicentre and comprehensive data collection for colorectal cancer is achievable and sustainable and promises great benefit as an audit and research tool. Similar initiatives could be established for other tumour types.

National growth in simulation training within emergency medicine residency programs, 2003-2008.

OBJECTIVES: The use of medical simulation has grown dramatically over the past decade, yet national data on the prevalence and growth of use among individual specialty training programs are lacking. The objectives of this study were to describe the current role of simulation training in emergency medicine (EM) residency programs and to quantify growth in use of the technology over the past 5 years. METHODS: In follow-up of a 2006 study (2003 data), the authors distributed an updated survey to program directors (PDs) of all 179 EM residency programs operating in early 2008 (140 Accreditation Council on Graduate Medical Education [ACGME]-approved allopathic programs and 39 American Osteopathic Association [AOA]-accredited osteopathic programs). The brief survey borrowed from the prior instrument, was edited and revised, and then distributed at a national PDs meeting. Subsequent follow-up was conducted by e-mail and telephone. The survey concentrated on technology-enhanced simulation modalities beyond routine static trainers or standardized patient-actors (high-fidelity mannequin simulation, part-task/procedural simulation, and dynamic screen-based simulation). RESULTS: A total of 134 EM residency programs completed the updated survey, yielding an overall response rate of 75%. A total of 122 (91%) use some form of simulation in their residency training. One-hundred fourteen (85%) specifically use mannequin-simulators, compared to 33 (29%) in 2003 (p < 0.001). Mannequin-simulators are now owned by 58 (43%) of the programs, whereas only 9 (8%) had primary responsibility for such equipment in 2003 (p < 0.001). Fifty-eight (43%) of the programs reported that annual resident simulation use now averages more than 10 hours per year. CONCLUSIONS: Use of medical simulation has grown significantly in EM residency programs in the past 5 years and is now widespread among training programs across the country.

Simulation in graduate medical education 2008: a review for emergency medicine.

Health care simulation includes a variety of educational techniques used to complement actual patient experiences with realistic yet artificial exercises. This field is rapidly growing and is widely used in emergency medicine (EM) graduate medical education (GME) programs. We describe the state of simulation in EM resident education, including its role in learning and assessment. The use of medical simulation in GME is increasing for a number of reasons, including the limitations of the 80-hour resident work week, patient dissatisfaction with being "practiced on," a greater emphasis on patient safety, and the importance of early acquisition of complex clinical skills. Simulation-based assessment (SBA) is advancing to the point where it can revolutionize the way clinical competence is assessed in residency training programs. This article also discusses the design of simulation centers and the resources available for developing simulation programs in graduate EM education. The level of interest in these resources is evident by the numerous national EM organizations with internal working groups focusing on simulation. In the future, the health care system will likely follow the example of the airline industry, nuclear power plants, and the military, making rigorous simulation-based training and evaluation a routine part of education and practice.

Error analysis of ultrasonic tissue doppler velocity estimation techniques for quantification of velocity and strain.

Recent work in the field of Doppler tissue imaging has focused mainly on the quantification of results involving the use of techniques of strain and strain-rate imaging. These results are based on measuring a velocity gradient between two points, a known distance apart, in the region-of-interest. Although many recent publications have demonstrated the potential of this technique in clinical terms, the method still suffers from low repeatability. The work presented here demonstrates, through the use of a rotating phantom arrangement and a custom developed single element ultrasound system, that this is a consequence of the fundamental accuracy of the technique used to estimate the original velocities. Results are presented comparing the performance of the conventional Kasai autocorrelation velocity estimator with those obtained using time domain cross-correlation and the complex cross-correlation model based estimator. The results demonstrate that the complex cross-correlation model based technique is able to offer lower standard deviations of the velocity gradient estimations compared with the Kasai algorithm.

Serologic response to inactivated poliovirus vaccine: a randomized clinical trial comparing 2 vaccination schedules in Puerto Rico.

BACKGROUND: The World Health Organization (WHO) recommends the discontinuation of oral poliovirus vaccine after eradication of wild poliovirus. Studies assessing inactivated poliovirus vaccine (IPV) immunogenicity in tropical countries, using the WHO Expanded Programme on Immunization (EPI) schedule, have been limited. METHODS: We conducted a randomized clinical trial in Ponce, Puerto Rico. Infants were assigned to 1 of 2 study arms: those in the EPI arm received IPV at 6, 10, and 14 weeks of age, and those in the US arm received IPV at 2, 4, and 6 months of age. Neutralizing antibody titers against poliovirus types 1, 2, and 3 were tested on serum specimens obtained before administration of the first dose of IPV and 28-45 days after administration of the last dose of IPV. RESULTS: Seroconversion rates for the EPI (n=225) and US (n=230) arms, respectively, were 85.8% and 99.6% for poliovirus type 1 (P<.001), 86.2% and 100% for poliovirus type 2 (P<.001), and 96.9% and 99.1% for poliovirus type 3 (P=.08). Seroconversion rates were lower among infants in the EPI arm who had high maternal antibody levels for all 3 poliovirus types (P<.001). CONCLUSIONS: The EPI schedule resulted in lower seroconversion rates for poliovirus types 1 and 2. These results are relevant for tropical countries planning to use IPV in a posteradication environment.

The use of the fractional Fourier transform with coded excitation in ultrasound imaging.

Medical ultrasound systems are limited by a tradeoff between axial resolution and the maximum imaging depth which may be achieved. The technique of coded excitation has been used extensively in the field of RADAR and SONAR for some time, but has only relatively recently been exploited in the area of medical ultrasound. This technique is attractive because allows the relationship between the pulse length and the maximum achievable spatial resolution to be changed. The work presented here explores the possibility of using the fractional Fourier transform as an effective means for the processing of signals received after the transmission of linear frequency modulated chirps. Results are presented which demonstrate that this technique is able to offer spatial resolutions similar to those obtained with a single cycle duration signal.

Empirical mode decomposition and tissue harmonic imaging.

Empirical mode decomposition (EMD) is a relatively new technique used in the analysis of nonlinear and nonstationary time series. Previous signal-processing methods used for medical ultrasound have been based on the assumption of a linear time-invariant system. More recently, the technique of tissue harmonic imaging (THI) has become prevalent. This technique relies on the nonlinear propagation of the sound wave through the medium to disperse the signal energy into the harmonic frequencies of the transmitted signal. In this paper, results are presented from using EMD to process received ultrasound echo signals that have passed through nonlinear media. The Hilbert spectrum is used to demonstrate an interpretation of the physical process underlying THI that is based on the concept of intrawave frequency modulation, rather then the spreading of signal energy into harmonic frequencies. The technique of EMD is shown to be able to produce superior results to the bandpass filtering method of THI, even when the band width of the transducer was such that the second harmonic would be suppressed.

Carcinoma of the anal canal: a local experience and review of the literature.

BACKGROUND: Through the 1970s patients presenting with anal canal carcinoma were managed with a surgical approach--abdomino-perineal resection. Since then, the pioneering work of Nigro et al. and a series of large clinical trials have clearly demonstrated that combined chemotherapy and radiotherapy result in greater local control, colostomy-free survival and increase in overall patient survival. The aim of the present study is to determine how widely the combined modality approach has been adopted in routine clinical practice and what outcomes are achieved in this setting. METHODS: All patients with anal cancer treated at three tertiary referral centres over an 11-year period (1991-2001) were identified. Data were collected by a retrospective record review. RESULTS: Our search identified a total of 50 patients: 22 men and 28 women, with a median age of 62 years. Four patients had metastatic disease diagnosed at presentation. Nine patients (18%) were at least 75 years of age and three were known to be HIV positive. Median potential follow up is 52 months. Of the 46 patients treated for cure, 38 received a combination of chemotherapy and radiation, with 79% achieving a complete response. Efficacy was maintained in treated elderly patients (> or =75 years). The 5-year survival of the 38 patients with local or locoregional disease who received combined chemoradiation modality was 63%. CONCLUSIONS: Overall this series demonstrates that combined chemotherapy and radiotherapy has been adopted as standard treatment with outcome data similar to those reported in the randomized clinical trials. Where possible elderly patients should receive combined modality therapy.