A Review of Personalised Cardiac Computational Modelling Using Electroanatomical Mapping Data.

Computational models of cardiac electrophysiology have gradually matured during the past few decades and are now being personalised to provide patient-specific therapy guidance for improving suboptimal treatment outcomes. The predictive features of these personalised electrophysiology models hold the promise of providing optimal treatment planning, which is currently limited in the clinic owing to reliance on a population-based or average patient approach. The generation of a personalised electrophysiology model entails a sequence of steps for which a range of activation mapping, calibration methods and therapy simulation pipelines have been suggested. However, the optimal methods that can potentially constitute a clinically relevant in silico treatment are still being investigated and face limitations, such as uncertainty of electroanatomical data recordings, generation and calibration of models within clinical timelines and requirements to validate or benchmark the recovered tissue parameters. This paper is aimed at reporting techniques on the personalisation of cardiac computational models, with a focus on calibrating cardiac tissue conductivity based on electroanatomical mapping data.

Non-local degradation modeling for spatially adaptive single image super-resolution.

Existing methods for single image super-resolution (SISR) model the blur kernel as spatially invariant across the entire image, and are susceptible to the adverse effects of textureless patches. To achieve improved results, adaptive estimation of the degradation kernel is necessary. We explore the synergy of joint global and local degradation modeling for spatially adaptive blind SISR. Our model, named spatially adaptive network for blind super-resolution (SASR), employs a simple encoder to estimate global degradation representations and a decoder to extract local degradation. These two representations are fused with a cross-attention mechanism and applied using spatially adaptive filtering to enhance the local image detail. Specifically, SASR contains two novel features: (1) a non-local degradation modeling with contrastive learning to learn global and local degradation representations, and (2) a non-local spatially adaptive filtering module (SAFM) that incorporates the global degradation and spatial-detail factors to preserve and enhance local details. We demonstrate that SASR can efficiently estimate degradation representations and handle multiple types of degradation. The local representations avoid the detrimental effect of estimating the entire super-resolved image with only one kernel through locally adaptive adjustments. Extensive experiments are performed to quantitatively and qualitatively demonstrate that SASR not only performs favorably for degradation estimation but also leads to state-of-the-art blind SISR performance when compared to alternative approaches.

Phenotyping left ventricular systolic dysfunction in asymptomatic individuals for improved risk stratification.

AIMS: Left ventricular systolic dysfunction (LSVD) is a heterogeneous condition with several factors influencing prognosis. Better phenotyping of asymptomatic individuals can inform preventative strategies. This study aims to explore the clinical phenotypes of LVSD in initially asymptomatic subjects and their association with clinical outcomes and cardiovascular abnormalities through multi-dimensional data clustering. METHODS AND RESULTS: Clustering analysis was performed on 60 clinically available variables from 1563 UK Biobank participants without pre-existing heart failure (HF) and with left ventricular ejection fraction (LVEF) < 50% on cardiovascular magnetic resonance (CMR) assessment. Risks of developing HF, other cardiovascular events, death, and a composite of major adverse cardiovascular events (MACE) associated with clusters were investigated. Cardiovascular imaging characteristics, not included in the clustering analysis, were also evaluated. Three distinct clusters were identified, differing considerably in lifestyle habits, cardiovascular risk factors, electrocardiographic parameters, and cardiometabolic profiles. A stepwise increase in risk profile was observed from Cluster 1 to Cluster 3, independent of traditional risk factors and LVEF. Compared with Cluster 1, the lowest risk subset, the risk of MACE ranged from 1.42 [95% confidence interval (CI): 1.03-1.96; P < 0.05] for Cluster 2 to 1.72 (95% CI: 1.36-2.35; P < 0.001) for Cluster 3. Cluster 3, the highest risk profile, had features of adverse cardiovascular imaging with the greatest LV re-modelling, myocardial dysfunction, and decrease in arterial compliance. CONCLUSIONS: Clustering of clinical variables identified three distinct risk profiles and clinical trajectories of LVSD amongst initially asymptomatic subjects. Improved characterization may facilitate tailored interventions based on the LVSD sub-type and improve clinical outcomes.

Ischemic Heart Disease and Vascular Risk Factors Are Associated With Accelerated Brain Aging.

BACKGROUND: Ischemic heart disease (IHD) has been linked with poor brain outcomes. The brain magnetic resonance imaging-derived difference between predicted brain age and actual chronological age (brain-age delta in years, positive for accelerated brain aging) may serve as an effective means of communicating brain health to patients to promote healthier lifestyles. OBJECTIVES: The authors investigated the impact of prevalent IHD on brain aging, potential underlying mechanisms, and its relationship with dementia risk, vascular risk factors, cardiovascular structure, and function. METHODS: Brain age was estimated in subjects with prevalent IHD (n = 1,341) using a Bayesian ridge regression model with 25 structural (volumetric) brain magnetic resonance imaging features and built using UK Biobank participants with no prevalent IHD (n = 35,237). RESULTS: Prevalent IHD was linked to significantly accelerated brain aging (P < 0.001) that was not fully mediated by microvascular injury. Brain aging (positive brain-age delta) was associated with increased risk of dementia (OR: 1.13 [95% CI: 1.04-1.22]; P = 0.002), vascular risk factors (such as diabetes), and high adiposity. In the absence of IHD, brain aging was also associated with cardiovascular structural and functional changes typically observed in aging hearts. However, such alterations were not linked with risk of dementia. CONCLUSIONS: Prevalent IHD and coexisting vascular risk factors are associated with accelerated brain aging and risk of dementia. Positive brain-age delta representing accelerated brain aging may serve as an effective communication tool to show the impact of modifiable risk factors and disease supporting preventative strategies.

Machine learning and disease prediction in obstetrics.

Machine learning technologies and translation of artificial intelligence tools to enhance the patient experience are changing obstetric and maternity care. An increasing number of predictive tools have been developed with data sourced from electronic health records, diagnostic imaging and digital devices. In this review, we explore the latest tools of machine learning, the algorithms to establish prediction models and the challenges to assess fetal well-being, predict and diagnose obstetric diseases such as gestational diabetes, pre-eclampsia, preterm birth and fetal growth restriction. We discuss the rapid growth of machine learning approaches and intelligent tools for automated diagnostic imaging of fetal anomalies and to asses fetoplacental and cervix function using ultrasound and magnetic resonance imaging. In prenatal diagnosis, we discuss intelligent tools for magnetic resonance imaging sequencing of the fetus, placenta and cervix to reduce the risk of preterm birth. Finally, the use of machine learning to improve safety standards in intrapartum care and early detection of complications will be discussed. The demand for technologies to enhance diagnosis and treatment in obstetrics and maternity should improve frameworks for patient safety and enhance clinical practice.

Diagnosing and Preventing Instabilities in Recurrent Video Processing.

Recurrent models are a popular choice for video enhancement tasks such as video denoising or super-resolution. In this work, we focus on their stability as dynamical systems and show that they tend to fail catastrophically at inference time on long video sequences. To address this issue, we (1) introduce a diagnostic tool which produces input sequences optimized to trigger instabilities and that can be interpreted as visualizations of temporal receptive fields, and (2) propose two approaches to enforce the stability of a model during training: constraining the spectral norm or constraining the stable rank of its convolutional layers. We then introduce Stable Rank Normalization for Convolutional layers (SRN-C), a new algorithm that enforces these constraints. Our experimental results suggest that SRN-C successfully enforces stablility in recurrent video processing models without a significant performance loss.

Deep learning models to map osteocyte networks can successfully distinguish between young and aged bone.

Osteocytes, the most abundant and mechanosensitive cells in bone tissue, play a pivotal role in bone homeostasis and mechano-responsiveness, orchestrating the intricate balance between bone formation and resorption under daily activity. Studying osteocyte connectivity and understanding their intricate arrangement within the lacunar canalicular network (LCN) is essential for unraveling bone physiology. This is particularly true as our bones age, which is associated with decreased integrity of the osteocyte network, disrupted mass transport, and lower sensitivity to the mechanical stimuli that allow the skeleton to adapt to changing demands. Much work has been carried out to investigate this relationship, often involving high resolution microscopy of discrete fragments of this network, alongside advanced computational modelling of individual cells. However, traditional methods of segmenting and measuring osteocyte connectomics are time-consuming and labour-intensive, often hindered by human subjectivity and limited throughput. In this study, we explore the application of deep learning and computer vision techniques to automate the segmentation and measurement of osteocyte connectomics, enabling more efficient and accurate analysis. We compare several state-of-the-art computer vision models (U-Nets and Vision Transformers) to successfully segment the LCN, finding that an Attention U-Net model can accurately segment and measure 81.8% of osteocytes and 42.1% of dendritic processes, when compared to manual labelling. While further development is required, we demonstrate that this degree of accuracy is already sufficient to distinguish between bones of young (2 month old) and aged (36 month old) mice, as well as capturing the degeneration induced by genetic modification of osteocytes. By harnessing the power of these advanced technologies, further developments can unravel the complexities of osteocyte networks in unprecedented detail, revolutionising our understanding of bone health and disease.

FlexHDR: Modeling Alignment and Exposure Uncertainties for Flexible HDR Imaging.

High dynamic range (HDR) imaging is of fundamental importance in modern digital photography pipelines and used to produce a high-quality photograph with well exposed regions despite varying illumination across the image. This is typically achieved by merging multiple low dynamic range (LDR) images taken at different exposures. However, over-exposed regions and misalignment errors due to poorly compensated motion result in artefacts such as ghosting. In this paper, we present a new HDR imaging technique that specifically models alignment and exposure uncertainties to produce high quality HDR results. We introduce a strategy that learns to jointly align and assess the alignment and exposure reliability using an HDR-aware, uncertainty-driven attention map that robustly merges the frames into a single high quality HDR image. Further, we introduce a progressive, multi-stage image fusion approach that can flexibly merge any number of LDR images in a permutation-invariant manner. Experimental results show our method can produce better quality HDR images with up to 1.1dB PSNR improvement to the state-of-the-art, and subjective improvements in terms of better detail, colours, and fewer artefacts.

Learning Frequency Domain Priors for Image Demoireing.

Image demoireing is a multi-faceted image restoration task involving both moire pattern removal and color restoration. In this paper, we raise a general degradation model to describe an image contaminated by moire patterns, and propose a novel multi-scale bandpass convolutional neural network (MBCNN) for single image demoireing. For moire pattern removal, we propose a multi-block-size learnable bandpass filters (M-LBFs), based on a block-wise frequency domain transform, to learn the frequency domain priors of moire patterns. We also introduce a new loss function named Dilated Advanced Sobel loss (D-ASL) to better sense the frequency information. For color restoration, we propose a two-step tone mapping strategy, which first applies a global tone mapping to correct for a global color shift, and then performs local fine tuning of the color per pixel. To determine the most appropriate frequency domain transform, we investigate several transforms including DCT, DFT, DWT, learnable non-linear transform and learnable orthogonal transform. We finally adopt the DCT. Our basic model won the AIM2019 demoireing challenge. Experimental results on three public datasets show that our method outperforms state-of-the-art methods by a large margin.

A Continual Learning Survey: Defying Forgetting in Classification Tasks.

Artificial neural networks thrive in solving the classification problem for a particular rigid task, acquiring knowledge through generalized learning behaviour from a distinct training phase. The resulting network resembles a static entity of knowledge, with endeavours to extend this knowledge without targeting the original task resulting in a catastrophic forgetting. Continual learning shifts this paradigm towards networks that can continually accumulate knowledge over different tasks without the need to retrain from scratch. We focus on task incremental classification, where tasks arrive sequentially and are delineated by clear boundaries. Our main contributions concern: (1) a taxonomy and extensive overview of the state-of-the-art; (2) a novel framework to continually determine the stability-plasticity trade-off of the continual learner; (3) a comprehensive experimental comparison of 11 state-of-the-art continual learning methods; and (4) baselines. We empirically scrutinize method strengths and weaknesses on three benchmarks, considering Tiny Imagenet and large-scale unbalanced iNaturalist and a sequence of recognition datasets. We study the influence of model capacity, weight decay and dropout regularization, and the order in which the tasks are presented, and qualitatively compare methods in terms of required memory, computation time, and storage.

Optical imaging technology in colonoscopy: Is there a role for photometric stereo?

Colonoscopy screening for the detection and removal of colonic adenomas is central to efforts to reduce the morbidity and mortality of colorectal cancer. However, up to a third of adenomas may be missed at colonoscopy, and the majority of post-colonoscopy colorectal cancers are thought to arise from these. Adenomas have three-dimensional surface topographic features that differentiate them from adjacent normal mucosa. However, these topographic features are not enhanced by white light colonoscopy, and the endoscopist must infer these from two-dimensional cues. This may contribute to the number of missed lesions. A variety of optical imaging technologies have been developed commercially to enhance surface topography. However, existing techniques enhance surface topography indirectly, and in two dimensions, and the evidence does not wholly support their use in routine clinical practice. In this narrative review, co-authored by gastroenterologists and engineers, we summarise the evidence for the impact of established optical imaging technologies on adenoma detection rate, and review the development of photometric stereo (PS) for colonoscopy. PS is a machine vision technique able to capture a dense array of surface normals to render three-dimensional reconstructions of surface topography. This imaging technique has several potential clinical applications in colonoscopy, including adenoma detection, polyp classification, and facilitating polypectomy, an inherently three-dimensional task. However, the development of PS for colonoscopy is at an early stage. We consider the progress that has been made with PS to date and identify the obstacles that need to be overcome prior to clinical application.

Deep learning for single-molecule science.

Exploring and making predictions based on single-molecule data can be challenging, not only due to the sheer size of the datasets, but also because a priori knowledge about the signal characteristics is typically limited and poor signal-to-noise ratio. For example, hypothesis-driven data exploration, informed by an expectation of the signal characteristics, can lead to interpretation bias or loss of information. Equally, even when the different data categories are known, e.g., the four bases in DNA sequencing, it is often difficult to know how to make best use of the available information content. The latest developments in machine learning (ML), so-called deep learning (DL) offer interesting, new avenues to address such challenges. In some applications, such as speech and image recognition, DL has been able to outperform conventional ML strategies and even human performance. However, to date DL has not been applied much in single-molecule science, presumably in part because relatively little is known about the 'internal workings' of such DL tools within single-molecule science as a field. In this Tutorial, we make an attempt to illustrate in a step-by-step guide how one of those, a convolutional neural network (CNN), may be used for base calling in DNA sequencing applications. We compare it with a SVM as a more conventional ML method, and discuss some of the strengths and weaknesses of the approach. In particular, a 'deep' neural network has many features of a 'black box', which has important implications on how we look at and interpret data.