Journal of The Royal Society Interface最新文献

We maintain balance during gait using both proactive and reactive control strategies. Damage to the brain from a stroke impairs reactive balance, but little is known about how a stroke impacts proactive control during walking. Stroke-related impairments to proactive control could become targets for interventions designed to improve responses to predictable disturbances and reduce fall risk. Therefore, we determined whether proactive strategies during predictable treadmill accelerations differed between people post-stroke (n = 14) and people without stroke (n = 14). Both groups walked with accelerations at random (every one to five strides) and regular (every three strides) intervals. We quantified the effects of the perturbations as changes to the centre of mass (COM) speed and used mechanical leg work to quantify the proactive strategies to slow the COM. Participants without stroke reduced peak COM speed better than those with stroke when perturbations were regular (-0.016 m s^-1 versus +0.004 m s^-1; p = 0.007). They also reduced positive leg work more during the perturbation step than the group post-stroke (-5.7% versus +2.5%; p = 0.003). One implication of these findings is that people post-stroke may be more susceptible to falls during predictable gait disturbances, and future work should identify the underlying impairments that cause these deficits.

Chaos is omnipresent in nature, and its understanding provides enormous social and economic benefits. However, the unpredictability of chaotic systems is a textbook concept due to their sensitivity to initial conditions, aperiodic behaviour, fractal dimensions, nonlinearity and strange attractors. In this work, we introduce, for the first time, chaotic learning, a novel multiscale topological paradigm that enables accurate predictions from chaotic systems. We show that seemingly random and unpredictable chaotic dynamics counterintuitively offer unprecedented quantitative predictions. Specifically, we devise multiscale topological Laplacians to embed real-world data into a family of interactive chaotic dynamical systems, modulate their dynamical behaviours and enable the accurate prediction of the input data. As a proof of concept, we consider 28 datasets from four categories of realistic problems: 10 brain waves, four benchmark protein datasets, 13 single-cell RNA sequencing datasets and an image dataset, as well as two distinct chaotic dynamical systems, namely the Lorenz and Rossler attractors. We demonstrate chaotic learning predictions of the physical properties from chaos. Our new chaotic learning paradigm profoundly changes the textbook perception of chaos and bridges topology, chaos and learning for the first time.

Epidemiological models of mosquito-borne virus transmission often lack accurate estimates of host-to-vector transmission probability. Here, we estimated this probability for two strains of Zika virus (ZIKV)-one sylvatic and one human-endemic-from two monkey species to Aedes albopictus mosquitoes using experimental infection data. Viral dynamics did not differ between monkey species, although one (cynomolgus macaque) is a native ZIKV host and the other (squirrel monkey) a novel host, but did differ between strains, with viremia for the human-endemic strain peaking later and lower than the sylvatic strain. Only the sylvatic strain was transmitted to mosquitoes. Within mosquitoes, anatomical barriers influence viral progression to salivary glands, complicating host infectiousness estimation. We quantified the probability of viral dissemination to the legs in Ae. albopictus, which increased with host viral load and was higher after feeding on squirrel monkeys than on cynomolgus macaques. We also found a positive relationship between virus titre in mosquito legs and virus detection in saliva after a 14-day extrinsic incubation period. Combining these factors, we found that squirrel monkeys were on average 1.5 times more infectious to Ae. albopictus than cynomolgus macaques. These estimates will help assess ZIKV's potential to establish an enzootic, sylvatic cycle in the Americas.

Cardiovascular diseases (CVDs) are the major cause of death worldwide. Magnetic resonance imaging (MRI) is the gold standard modality for CVD diagnosis because of its ability to distinguish different types of soft tissues without the use of ionizing radiation. Cine MRI allows us to see the contractile function of the heart, and it is a safe method for patients with chronic kidney diseases. The aim of this work was to develop a deep learning model for automated classification of common CVDs from cine MRI while providing the model explainability. We investigated single-phase models based on either the end-diastolic (ED) or end-systolic (ES) phase using seven baseline deep learning models including ResNet, DenseNet and VGG. We then developed a multi-phase model including both ED and ES phases to incorporate cardiac function for CVD classification. While the single-phase model for the ED and ES phases yielded the highest test F1-scores of 71.0% and [Formula: see text] respectively, the multi-phase model achieved a test F1-score of [Formula: see text]. To better understand the model performance, we used explainability to visualize regions of the heart that exhibit characteristics of each disease. Our work has demonstrated that deep learning models can automatically and effectively classify CVDs from cine MRI while justifying classification, thus building trust from the clinical community.

The flow of cerebrospinal fluid (CSF) through perivascular spaces (PVSs) is an important part of the brain's system for clearing metabolic waste. Astrocyte endfeet ensheath the PVSs of penetrating arteries, separating them from brain extracellular space (ECS). Gaps between astrocyte endfeet could provide a low-resistance pathway for fluid transport across the endfoot wall. Recent research suggests that the astrocyte endfeet may also function as valves that rectify the CSF flow, allowing oscillatory pressures to drive net flows like those observed in experiments. In this study, we employ fluid-structure interaction modelling to investigate the endfoot valve mechanism. Due to the unavailability of precise in vivo measurements of gap shape and size, we explore three possible, though idealized, geometric arrangements: wedge-shaped gaps, overlapping endfeet of different sizes and curvature of the endfoot wall. For each, we quantify the dependence of net flow on oscillatory pressure amplitude, frequency and other key parameters. For all three, our simulations demonstrate effective flow rectification at frequencies associated with functional hyperaemia, respiration and cardiac pulsation.

Parameter inference is a critical step in the process of interpreting biological data using mathematical models. Inference provides a means of deriving quantitative, mechanistic insights from sparse, noisy data. While methods for parameter inference, parameter identifiability and model prediction are well developed for deterministic continuum models, working with biological applications often requires stochastic modelling approaches to capture inherent variability and randomness that can be prominent in biological measurements and data. Random walk models are especially useful for capturing spatio-temporal processes, such as ecological population dynamics, molecular transport phenomena and collective behaviour associated with multicellular phenomena. This review focuses on parameter inference, identifiability analysis and model prediction for a suite of biologically inspired, stochastic agent-based models relevent to animal dispersal and populations of biological cells. With a particular emphasis on model prediction, we highlight roles for numerical optimization and automatic differentiation. Open-source Julia code is provided to support scientific reproducibility. We encourage readers to use this code directly or adapt it to suit their interests and applications.

We propose a simple dynamic model of cancer development that captures carcinogenesis and subsequent cancer progression. A central idea of the model is to include the immune response to cancer, which leads to the emergence of an extinction threshold. We first identify the limitations of commonly used extinction threshold models from population biology in reproducing typical cancer progression. We then address these limitations by deriving a new model that incorporates: (i) random mutations of stem cells at a rate that increases with age and (ii) immune response whose strength may also vary over time. Our model accurately reproduces a wide range of real-world cancer data: the typical age-specific cumulative risk of most human cancers, the progression of breast cancer in mice and the unusual age-specific cumulative risk of breast cancer in women. In the last case, we model the different immune response at different phases of the menstrual cycle and menopausal treatment and show that this leads to a moving extinction threshold. This approach provides new insights into the effects of hormone replacement therapy and menstrual cycle length on breast cancer in women. More generally, it can be applied to a variety of other cancer scenarios where the immune response or other important factors vary over time.

Animal clustering takes place at a variety of spatial scales. While methods to quantify clustering already exist, many of these methods are either scale independent, not parameter-free, or model proximity as a binary function, which makes them unsuitable for anisotropic systems and is not representative of the perception neighbourhood of animals. We describe a method to quantify the degree of clustering of point-location data at different spatial scales, which uses kernel density estimation to construct a density function from the underlying point-location data. We build upon this method to automatically detect cluster diameters using smoothing kernels that better represent the perception neighbourhood of animals. Finally, we test our methods on artificial datasets with varying clustering characteristics, as well as on a dataset of African bush elephants. Our method correctly assigns higher clustering values to spatial scales with high degrees of clustering and accurately outputs a set of spatial scales that correspond to cluster diameters. The accuracy of our method is insensitive to the chosen kernel function. Combined with the parameter-free nature of our method, this allows for easy detection of clustering scales in anisotropic and hierarchically clustered systems, such as animal groups.

Hospital operating theatre suites are a particularly resource- and energy-intensive component of the health sector. Reducing their carbon footprint presents a significant challenge due to the necessity of maintaining patient safety. In this paper, we apply a multidisciplinary methodology to investigate and assess various strategies aimed at reducing the carbon footprint in hospital theatres. The strategies evaluated include (i) the duration of theatre ventilation operation, (ii) the efficiency of the ventilation strategy, and (iii) heat recovery systems and technologies. These approaches are assessed using a combination of theatre space monitoring (via building management systems), computational air-flow modelling and mathematical models. We develop a robust methodology that applies these modelling techniques to general hospital suites, enabling the estimation of reductions in CO₂ equivalent.

The paradox of enrichment stipulates that increasing the resources available to the prey population can lead to instability and a higher likelihood of population fluctuations. We study the converse situation where the prey's environment is degrading and ask if the dynamical interplay between this degradation and stochasticity can be beneficial to the stabilization of the prey population. The underlying systems are non-autonomous and subject to noise. We uncover a phenomenon pertinent to the paradox of enrichment: rare rarity. In particular, in a slow-fast ecosystem with a sole stable equilibrium, noise can induce dynamical excursions of a trajectory into a region with low species abundance, resulting in rarity. Surprisingly, it is the same noise that can facilitate a rapid recovery of the abundance of the rare species, shortening the duration of the rarity. As the environment continues to degrade, the occurrence of such rarity events can be non-uniform in time and even more rare. The intermittent occurrence of rare rarity is caused by the dynamical interplay between the phase-space distance from the stable equilibrium to the boundary separating two distinct regions of transient dynamics. The rare-rarity phenomenon can also arise in other natural systems such as the climate carbon-cycle system.