Accelerated motion correction with deep generative diffusion models.

PURPOSE: The aim of this work is to develop a method to solve the ill-posed inverse problem of accelerated image reconstruction while correcting forward model imperfections in the context of subject motion during MRI examinations. METHODS: The proposed solution uses a Bayesian framework based on deep generative diffusion models to jointly estimate a motion-free image and rigid motion estimates from subsampled and motion-corrupt two-dimensional (2D) k-space data. RESULTS: We demonstrate the ability to reconstruct motion-free images from accelerated two-dimensional (2D) Cartesian and non-Cartesian scans without any external reference signal. We show that our method improves over existing correction techniques on both simulated and prospectively accelerated data. CONCLUSION: We propose a flexible framework for retrospective motion correction of accelerated MRI based on deep generative diffusion models, with potential application to other forward model corruptions.

An Experimental Phototherapy Device for Studying the Effects of Blue Light on Patients with Raynaud's Phenomenon.

Raynaud's phenomenon (RP) is a condition that causes decreased blood flow to areas perfused by small blood vessels (e.g., fingers, toes). In severe cases, ulceration, gangrene, and loss of fingers may occur. Most treatments focus on inducing vasorelaxation in affected areas by the way of pharmaceuticals. Recently, animal studies have shown that vasorelaxation can be induced by non-coherent blue light (wavelength ~ 430-460 nm) through the actions of melanopsin, a photoreceptive opsin protein encoded by the OPN4 gene. To study this effect in humans, a reliable phototherapy device (PTD) is needed. We outline the construction of a PTD to be used in studying blue light effects on Raynaud's patients. Our design addresses user safety, calibration, electromagnetic compatibility/interference (EMC/EMI), and techniques for measuring physiological responses (temperature sensors, laser Doppler flow sensors, infrared thermal imaging of the hands). We tested our device to ensure (1) safe operating conditions, (2) predictable, user-controlled irradiance output levels, (3) an ability for measuring physiological responses, and (4) features necessary to enable a double-blinded crossover study for a clinical trial. We also include in the Methods an approved research protocol utilizing our device that may serve as a starting point for clinical study. We introduced a reliable PTD for studying the effects of blue light therapy for patients suffering from Raynaud's phenomenon and showed that our device is safe and reliable and includes the required measurement vectors for tracking treatment effects throughout the duration of a clinical study.

Federated End-to-End Unrolled Models for Magnetic Resonance Image Reconstruction.

Image reconstruction is the process of recovering an image from raw, under-sampled signal measurements, and is a critical step in diagnostic medical imaging, such as magnetic resonance imaging (MRI). Recently, data-driven methods have led to improved image quality in MRI reconstruction using a limited number of measurements, but these methods typically rely on the existence of a large, centralized database of fully sampled scans for training. In this work, we investigate federated learning for MRI reconstruction using end-to-end unrolled deep learning models as a means of training global models across multiple clients (data sites), while keeping individual scans local. We empirically identify a low-data regime across a large number of heterogeneous scans, where a small number of training samples per client are available and non-collaborative models lead to performance drops. In this regime, we investigate the performance of adaptive federated optimization algorithms as a function of client data distribution and communication budget. Experimental results show that adaptive optimization algorithms are well suited for the federated learning of unrolled models, even in a limited-data regime (50 slices per data site), and that client-sided personalization can improve reconstruction quality for clients that did not participate in training.

Chronic Low-Level Vagus Nerve Stimulation Improves Long-Term Survival in Salt-Sensitive Hypertensive Rats.

Chronic hypertension (HTN) affects more than 1 billion people worldwide, and is associated with an increased risk of cardiovascular disease. Despite decades of promising research, effective treatment of HTN remains challenging. This work investigates vagus nerve stimulation (VNS) as a novel, device-based therapy for HTN treatment, and specifically evaluates its effects on long-term survival and HTN-associated adverse effects. HTN was induced in Dahl salt-sensitive rats using a high-salt diet, and the rats were randomly divided into two groups: VNS (n = 9) and Sham (n = 8), which were implanted with functional or non-functional VNS stimulators, respectively. Acute and chronic effects of VNS therapy were evaluated through continuous monitoring of blood pressure (BP) and ECG via telemetry devices. Autonomic tone was quantified using heart rate (HR), HR variability (HRV) and baroreflex sensitivity (BRS) analysis. Structural cardiac changes were quantified through gross morphology and histology studies. VNS significantly improved the long-term survival of hypertensive rats, increasing median event-free survival by 78% in comparison to Sham rats. Acutely, VNS improved autonomic balance by significantly increasing HRV during stimulation, which may lead to beneficial chronic effects of VNS therapy. Chronic VNS therapy slowed the progression of HTN through an attenuation of SBP and by preserving HRV. Finally, VNS significantly altered cardiac structure, increasing heart weight, but did not alter the amount of fibrosis in the hypertensive hearts. These results suggest that VNS has the potential to improve outcomes in subjects with severe HTN.