Structured Low-Rank Algorithms: Theory, Magnetic Resonance Applications, and Links to Machine Learning.

In this survey, we provide a detailed review of recent advances in the recovery of continuous domain multidimensional signals from their few non-uniform (multichannel) measurements using structured low-rank matrix completion formulation. This framework is centered on the fundamental duality between the compactness (e.g., sparsity) of the continuous signal and the rank of a structured matrix, whose entries are functions of the signal. This property enables the reformulation of the signal recovery as a low-rank structured matrix completion, which comes with performance guarantees. We will also review fast algorithms that are comparable in complexity to current compressed sensing methods, which enables the application of the framework to large-scale magnetic resonance (MR) recovery problems. The remarkable flexibility of the formulation can be used to exploit signal properties that are difficult to capture by current sparse and low-rank optimization strategies. We demonstrate the utility of the framework in a wide range of MR imaging (MRI) applications, including highly accelerated imaging, calibration-free acquisition, MR artifact correction, and ungated dynamic MRI.

Model-Based Deep Learning for Reconstruction of Joint k-q Under-sampled High Resolution Diffusion MRI.

We propose a model-based deep learning architecture for the reconstruction of highly accelerated diffusion magnetic resonance imaging (MRI) that enables high resolution imaging. The proposed reconstruction jointly recovers all the diffusion weighted images in a single step from a joint k-q under-sampled acquisition in a parallel MRI setting. We propose the novel use of a pre-trained denoiser as a regularizer in a model-based reconstruction for the recovery of highly under-sampled data. Specifically, we designed the denoiser based on a general diffusion MRI tissue microstructure model for multi-compartmental modeling. By using a wide range of biologically plausible parameter values for the multi-compartmental microstructure model, we simulated diffusion signal that spans the entire microstructure parameter space. A neural network was trained in an unsupervised manner using an autoencoder to learn the diffusion MRI signal subspace. We employed the autoencoder in a model-based reconstruction and show that the autoencoder provides a strong denoising prior to recover the q-space signal. We show reconstruction results on a simulated brain dataset that shows high acceleration capabilities of the proposed method.

MoDL-MUSSELS: Model-Based Deep Learning for Multishot Sensitivity-Encoded Diffusion MRI.

We introduce a model-based deep learning architecture termed MoDL-MUSSELS for the correction of phase errors in multishot diffusion-weighted echo-planar MR images. The proposed algorithm is a generalization of the existing MUSSELS algorithm with similar performance but significantly reduced computational complexity. In this work, we show that an iterative re-weighted least-squares implementation of MUSSELS alternates between a multichannel filter bank and the enforcement of data consistency. The multichannel filter bank projects the data to the signal subspace, thus exploiting the annihilation relations between shots. Due to the high computational complexity of the self-learned filter bank, we propose replacing it with a convolutional neural network (CNN) whose parameters are learned from exemplary data. The proposed CNN is a hybrid model involving a multichannel CNN in the k-space and another CNN in the image space. The k-space CNN exploits the annihilation relations between the shot images, while the image domain network is used to project the data to an image manifold. The experiments show that the proposed scheme can yield reconstructions that are comparable to state-of-the-art methods while offering several orders of magnitude reduction in run-time.

MULTI-SHOT SENSITIVITY-ENCODED DIFFUSION MRI USING MODEL-BASED DEEP LEARNING (MODL-MUSSELS).

We propose a model-based deep learning architecture for the correction of phase errors in multishot diffusion-weighted echo-planar MRI images. This work is a generalization of MUSSELS, which is a structured low-rank algorithm. We show that an iterative reweighted least-squares implementation of MUSSELS resembles the model-based deep learning (MoDL) framework. We propose to replace the self-learned linear filter bank in MUSSELS with a convolutional neural network, whose parameters are learned from exemplary data. The proposed algorithm reduces the computational complexity of MUSSELS by several orders of magnitude, while providing comparable image quality.

MoDL: Model-Based Deep Learning Architecture for Inverse Problems.

We introduce a model-based image reconstruction framework with a convolution neural network (CNN)-based regularization prior. The proposed formulation provides a systematic approach for deriving deep architectures for inverse problems with the arbitrary structure. Since the forward model is explicitly accounted for, a smaller network with fewer parameters is sufficient to capture the image information compared to direct inversion approaches. Thus, reducing the demand for training data and training time. Since we rely on end-to-end training with weight sharing across iterations, the CNN weights are customized to the forward model, thus offering improved performance over approaches that rely on pre-trained denoisers. Our experiments show that the decoupling of the number of iterations from the network complexity offered by this approach provides benefits, including lower demand for training data, reduced risk of overfitting, and implementations with significantly reduced memory footprint. We propose to enforce data-consistency by using numerical optimization blocks, such as conjugate gradients algorithm within the network. This approach offers faster convergence per iteration, compared to methods that rely on proximal gradients steps to enforce data consistency. Our experiments show that the faster convergence translates to improved performance, primarily when the available GPU memory restricts the number of iterations.

MODEL BASED IMAGE RECONSTRUCTION USING DEEP LEARNED PRIORS (MODL).

We introduce a model-based image reconstruction framework, where we use a deep convolution neural network (CNN) based regularization prior. We rely on a recursive algorithm, which alternates between a CNN based denoising step and enforcement of data consistency. Unrolling the recursive algorithm yields a deep network that is trained using backpropagation. The unique aspect of this method is the use of the same CNN weights at each iteration, which makes the resulting structure consistent with the model-based formulation. Also, this approach reduces the number of trainable parameters, which hence lower the amount of training data needed. The use of a forward model also reduces the size of the network and enables the exploitation additional prior information available from calibration data. The use of the framework for multichannel MRI reconstruction provides improved reconstructions, compared to other state-of-the-art methods.

Reducing waiting times for sleep apnoea hypopnoea syndrome and snoring using a questionnaire and home oximetry: results of a second audit cycle.

As a result of a previous audit on the management of sleep apnoea hypopnoea syndrome (SAHS) which showed long waiting times that were primarily due to unnecessary interspecialty referrals, a change in practice was adopted. All referrals are now sent a questionnaire about symptoms suggestive of SAHS, the Epworth Sleepiness Scale score and their body mass index (BMI) which when returned are categorized into having a high, intermediate or low risk of SAHS. Those patients with a high probability have home overnight oximetry and those with intermediate probability have video oximetry. Those with a low probability are referred directly to ENT. We audited the first 100 patients referred. All were General Practitioner referrals to either ENT or respiratory medicine. Only two patients had a low probability score and were seen directly in ENT. Following sleep study analysis, 10 patients were referred directly to ENT with no respiratory medicine follow-up and nine were discharged back to the General Practitioner with no apnoea or snoring. Eighty-one patients were followed up by respiratory medicine. Of these, 49 received a trial of nasal continuous positive airway pressure (nCPAP) and six were referred to ENT. Therefore the majority justified an investigation to exclude SAHS in the first instance and an unnecessary initial ENT appointment was avoided. We have reduced the average waiting times to sleep study by approximately 90 days and to nCPAP trial by 32 days, mostly due to decreased delays in interspeciality referrrals. We have also demonstrated a greater than 50 per cent reduction in ENT clinic visits, a small increase in the number of sleep studies but no increase in respiratory clinic workload.

Benchmarking: the five year outcome of rheumatoid arthritis assessed using a pain score, the Health Assessment Questionnaire, and the Short Form-36 (SF-36) in a community and a clinic based sample.

BACKGROUND: Treatment, and therefore outcome, of rheumatoid arthritis (RA) will improve in the next few years. However, improvement in outcome can only be judged against the probability of certain outcomes with current conventional treatment. AIM: To document the five year outcome of RA in the late 1990s. SETTING: Norfolk Arthritis Register (NOAR). DESIGN: Longitudinal observational cohort study. METHODS: 318 patients with recent onset inflammatory polyarthritis recruited by NOAR in 1990-91 completed five years of follow up. Four groups were assessed: the whole cohort, all those referred to hospital, those who satisfied criteria for RA at baseline, and those referred to hospital who satisfied criteria for RA at baseline. Outcome was assessed with a visual analogue scale for pain, the Health Assessment Questionnaire (HAQ), and the Short Form-36 (SF-36). RESULTS: Of the RA hospital attenders, 50% had a visual analogue scale pain score of 5 cm or less and an HAQ score of 1.125 or less. SF-36 scores were reduced in all domains. Results are presented as cumulative percentages. CONCLUSIONS: These results can be used for comparison and to set targets for improvement.

Corticosteroid injection for the treatment of carpal tunnel syndrome.

OBJECTIVE: To compare low and high dose, and short and long acting corticosteroids in the treatment of carpal tunnel syndrome. METHODS: A randomised, controlled, single blind trial with electromyographic and subjective outcome measures. RESULTS: 25 mg hydrocortisone is as effective as higher doses or long acting triamcinolone at a six week and six month follow up. CONCLUSION: As low dose steroid is as effective, and potentially less toxic, this should be the recommended dose for injection of carpal tunnel syndrome.

A diagnostic algorithm for carpal tunnel syndrome based on Bayes's theorem.

OBJECTIVE: To construct a diagnostic algorithm based on Bayes's theorem and using simple clinical tests to allow accurate diagnosis without resort to nerve conduction studies. METHOD: A cohort of patients being referred with possible carpal tunnel syndrome had clinical and electrophysiological testing, from which the simple calculations for sensitivity, specificity and prevalence were made and subsequently used in the formula of Bayes's theorem. The algorithm was then tested prospectively in a further cohort of similarly referred patients. RESULTS: The algorithm proved to be reliable when tested prospectively, and was similar to nerve conduction studies in diagnostic accuracy. CONCLUSION: A simple algorithm of clinical tests can identify patients without resort to nerve conduction studies, facilitating early treatment.

Systemic vasculitis and atypical infections: report of two cases.

Two cases of systemic vasculitis are described; one presenting with adult Henoch-Schonlein purpura secondary to a concomitant Chlamydia infection and the other with leucocytoclastic vasculitis and mesangioproliferative glomerulonephritis secondary to a recent parvovirus B19 infection. Association of chlamydial infection has not previously been described with Henoch-Schonlein purpura and this infection should, perhaps, be added to the list of aetiologies of this disease. Parvovirus B19 causing significant urinary sediment abnormalities associated with mesangioproliferative glomerulonephritis and leucocytoclastic vasculitis has also not been described previously.

Healthcare information. Slow to learn.

Access to information in the NHS is in need of huge improvements. In 1992 more than half of all staff had no access to a library service geared to their requirements. NHS libraries are now expected to serve patients and managers as well as clinical staff.