Publications

Publications from our Researchers

Several of our current PhD candidates and fellow researchers at the Data Science Institute have published, or in the proccess of publishing, papers to present their research.  

Search or filter publications

Filter by type:

Filter by publication type

Filter by year:

to

Results

  • Showing results for:
  • Reset all filters

Search results

  • Journal article
    Gan HM, Fernando S, Molina-Solana M, 2021,

    Scalable object detection pipeline for traffic cameras: Application to Tfl JamCams

    , Expert Systems with Applications, ISSN: 0957-4174

    With CCTV systems being installed in the transport infrastructures of many cities, there is an abundance of data to be extracted from the footage. This paper explores the application of the YOLOv3 object detection algorithm trained on the COCO dataset to the Transport for London’s (TfL) JamCam feed. The result, open-sourced and publicly available, is a series of easy to deploy Docker pipelines to create, store and serve (through a REST API) data on identified objects on that feed. The pipelines can be deployed to any Linux machine with an NVIDIA GPU to support accelerated computation. We studied how different confidence thresholds affect detections of relevant objects (cars, trucks and pedestrians) in London JamCam scenes. By running the system continuously for 3 weeks, we built a dataset of more than 2200 detection datapoints for each camera (̃6 datapoints an hour). We further visualised the detections on an animated geospatial map, showcasing their effectiveness in identifying traffic patterns typical of an urban city like London, portraying the variation on different object population levels throughout the day.
  • Journal article
    Huitzil I, Molina-Solana M, Gómez-Romero J, Bobillo Fet al., 2021,

    Minimalistic fuzzy ontology reasoning: An application to Building Information Modeling

    , Applied Soft Computing, Vol: 103, Pages: 1-15, ISSN: 1568-4946

    This paper presents a minimalistic reasoning algorithm to solve imprecise instance retrieval in fuzzy ontologies with application to querying Building Information Models (BIMs)—a knowledge representation formalism used in the construction industry. Our proposal is based on a novel lossless reduction of fuzzy to crisp reasoning tasks, which can be processed by any Description Logics reasoner. We implemented the minimalistic reasoning algorithm and performed an empirical evaluation of its performance in several tasks: interoperation with classical reasoners (Hermit and TrOWL), initialization time (comparing TrOWL and a SPARQL engine), and use of different data structures (hash tables, databases, and programming interfaces). We show that our software can efficiently solve very expressive queries not available nowadays in regular or semantic BIMs tools.
  • Journal article
    Tajnafoi G, Arcucci R, Mottet L, Vouriot C, Molina-Solana M, Pain C, Guo Y-Ket al., 2021,

    Variational Gaussian process for optimal sensor placement

    , Applications of Mathematics, Vol: 66, Pages: 287-317, ISSN: 0373-6725

    Sensor placement is an optimisation problem that has recently gained great relevance. In order to achieve accurate online updates of a predictive model, sensors are used to provide observations. When sensor location is optimally selected, the predictive model can greatly reduce its internal errors. A greedy-selection algorithm is used for locating these optimal spatial locations from a numerical embedded space. A novel architecture for solving this big data problem is proposed, relying on a variational Gaussian process. The generalisation of the model is further improved via the preconditioning of its inputs: Masked Autoregressive Flows are implemented to learn nonlinear, invertible transformations of the conditionally modelled spatial features. Finally, a global optimisation strategy extending the Mutual Information-based optimisation and fine-tuning of the selected optimal location is proposed. The methodology is parallelised to speed up the computational time, making these tools very fast despite the high complexity associated with both spatial modelling and placement tasks. The model is applied to a real three-dimensional test case considering a room within the Clarence Centre building located in Elephant and Castle, London, UK.
  • Journal article
    Balaban G, Halliday B, Bradley P, Bai W, Nygaard S, Owen R, Hatipoglu S, Ferreira ND, Izgi C, Tayal U, Corden B, Ware J, Pennell D, Rueckert D, Plank G, Rinaldi CA, Prasad SK, Bishop Met al., 2021,

    Late-gadolinium enhancement interface area and electrophysiological simulations predict arrhythmic events in non-ischemic dilated cardiomyopathy patients

    , JACC: Clinical Electrophysiology, Vol: 7, Pages: 238-249, ISSN: 2405-5018

    BACKGROUND: The presence of late-gadolinium enhancement (LGE) predicts life threatening ventricular arrhythmias in non-ischemic dilated cardiomyopathy (NIDCM); however, risk stratification remains imprecise. LGE shape and simulations of electrical activity may be able to provide additional prognostic information.OBJECTIVE: This study sought to investigate whether shape-based LGE metrics and simulations of reentrant electrical activity are associated with arrhythmic events in NIDCM patients.METHODS: CMR-LGE shape metrics were computed for a cohort of 156 NIDCM patients with visible LGE and tested retrospectively for an association with an arrhythmic composite end-point of sudden cardiac death and ventricular tachycardia. Computational models were created from images and used in conjunction with simulated stimulation protocols to assess the potential for reentry induction in each patient’s scar morphology. A mechanistic analysis of the simulations was carried out to explain the associations. RESULTS: During a median follow-up of 1611 [IQR 881-2341] days, 16 patients (10.3%) met the primary endpoint. In an inverse probability weighted Cox regression, the LGE-myocardial interface area (HR:1.75; 95% CI:1.24-2.47; p=0.001), number of simulated reentries (HR: 1.4; 95% CI: 1.23-1.59; p<0.01) and LGE volume (HR:1.44; 95% CI:1.07-1.94; p=0.02) were associated with arrhythmic events. Computational modeling revealed repolarisation heterogeneity and rate-dependent block of electrical wavefronts at the LGE-myocardial interface as putative arrhythmogenic mechanisms directly related to LGE interface area.CONCLUSION: The area of interface between scar and surviving myocardium, as well as simulated reentrant activity, are associated with an elevated risk of major arrhythmic events in NIDCM patients with LGE and represent novel risk predictors.
  • Journal article
    Xiong Z, Xia Q, Hu Z, Huang N, Bian C, Zheng Y, Vesal S, Ravikumar N, Maier A, Yang X, Heng P-A, Ni D, Li C, Tong Q, Si W, Puybareau E, Khoudli Y, Geraud T, Chen C, Bai W, Rueckert D, Xu L, Zhuang X, Luo X, Jia S, Sermesant M, Liu Y, Wang K, Borra D, Masci A, Corsi C, de Vente C, Veta M, Karim R, Preetha CJ, Engelhardt S, Qiao M, Wang Y, Tao Q, Nunez-Garcia M, Camara O, Savioli N, Lamata P, Zhao Jet al., 2021,

    A global benchmark of algorithms for segmenting the left atrium from late gadolinium-enhanced cardiac magnetic resonance imaging

    , Medical Image Analysis, Vol: 67, Pages: 1-14, ISSN: 1361-8415

    Segmentation of medical images, particularly late gadolinium-enhanced magnetic resonance imaging (LGE-MRI) used for visualizing diseased atrial structures, is a crucial first step for ablation treatment of atrial fibrillation. However, direct segmentation of LGE-MRIs is challenging due to the varying intensities caused by contrast agents. Since most clinical studies have relied on manual, labor-intensive approaches, automatic methods are of high interest, particularly optimized machine learning approaches. To address this, we organized the 2018 Left Atrium Segmentation Challenge using 154 3D LGE-MRIs, currently the world's largest atrial LGE-MRI dataset, and associated labels of the left atrium segmented by three medical experts, ultimately attracting the participation of 27 international teams. In this paper, extensive analysis of the submitted algorithms using technical and biological metrics was performed by undergoing subgroup analysis and conducting hyper-parameter analysis, offering an overall picture of the major design choices of convolutional neural networks (CNNs) and practical considerations for achieving state-of-the-art left atrium segmentation. Results show that the top method achieved a Dice score of 93.2% and a mean surface to surface distance of 0.7 mm, significantly outperforming prior state-of-the-art. Particularly, our analysis demonstrated that double sequentially used CNNs, in which a first CNN is used for automatic region-of-interest localization and a subsequent CNN is used for refined regional segmentation, achieved superior results than traditional methods and machine learning approaches containing single CNNs. This large-scale benchmarking study makes a significant step towards much-improved segmentation methods for atrial LGE-MRIs, and will serve as an important benchmark for evaluating and comparing the future works in the field. Furthermore, the findings from this study can potentially be extended to other imaging datasets and modalitie
  • Journal article
    Andersen MM, Schjoedt U, Price H, Rosas FE, Scrivner C, Clasen Met al., 2020,

    Playing with fear: a field study in recreational horror

    , Psychological Science, Vol: 31, Pages: 1497-1510, ISSN: 0956-7976

    Haunted attractions are illustrative examples of recreational fear in which people voluntarily seek out frightening experiences in pursuit of enjoyment. We present findings from a field study at a haunted-house attraction where visitors between the ages of 12 and 57 years (N = 110) were equipped with heart rate monitors, video-recorded at peak scare points during the attraction, and asked to report on their experience. Our results show that enjoyment has an inverted-U-shaped relationship with fear across repeated self-reported measures. Moreover, results from physiological data demonstrate that the experience of being frightened is a linear function of large-scale heart rate fluctuations, whereas there is an inverted-U-shaped relationship between participant enjoyment and small-scale heart rate fluctuations. These results suggest that enjoyment is related to forms of arousal dynamics that are “just right.” These findings shed light on how fear and enjoyment can coexist in recreational horror.
  • Journal article
    Rosas FE, Mediano PAM, Rassouli B, Barrett ABet al., 2020,

    An operational information decomposition via synergistic disclosure

    , Journal of Physics A: Mathematical and Theoretical, Vol: 53, Pages: 485001-485001, ISSN: 1751-8113

    Multivariate information decompositions hold promise to yield insight into complex systems, and stand out for their ability to identify synergistic phenomena. However, the adoption of these approaches has been hindered by there being multiple possible decompositions, and no precise guidance for preferring one over the others. At the heart of this disagreement lies the absence of a clear operational interpretation of what synergistic information is. Here we fill this gap by proposing a new information decomposition based on a novel operationalisation of informational synergy, which leverages recent developments in the literature of data privacy. Our decomposition is defined for any number of information sources, and its atoms can be calculated using elementary optimisation techniques. The decomposition provides a natural coarse-graining that scales gracefully with the system's size, and is applicable in a wide range of scenarios of practical interest.
  • Journal article
    Herzog R, Mediano PAM, Rosas FE, Carhart-Harris R, Perl YS, Tagliazucchi E, Cofre Ret al., 2020,

    A mechanistic model of the neural entropy increase elicited by psychedelic drugs

    , Scientific Reports, Vol: 10, ISSN: 2045-2322

    Psychedelic drugs, including lysergic acid diethylamide and other agonists of the serotonin 2A receptor (5HT2A-R), induce drastic changes in subjective experience, and provide a unique opportunity to study the neurobiological basis of consciousness. One of the most notable neurophysiological signatures of psychedelics, increased entropy in spontaneous neural activity, is thought to be of relevance to the psychedelic experience, mediating both acute alterations in consciousness and long-term effects. However, no clear mechanistic explanation for this entropy increase has been put forward so far. We sought to do this here by building upon a recent whole-brain model of serotonergic neuromodulation, to study the entropic effects of 5HT2A-R activation. Our results reproduce the overall entropy increase observed in previous experiments in vivo, providing the first model-based explanation for this phenomenon. We also found that entropy changes were not uniform across the brain: entropy increased in some regions and decreased in others, suggesting a topographical reconfiguration mediated by 5HT2A-R activation. Interestingly, at the whole-brain level, this reconfiguration was not well explained by 5HT2A-R density, but related closely to the topological properties of the brain's anatomical connectivity. These results help us understand the mechanisms underlying the psychedelic state and, more generally, the pharmacological modulation of whole-brain activity.
  • Journal article
    Bai W, Suzuki H, Huang J, Francis C, Wang S, Tarroni G, Guitton F, Aung N, Fung K, Petersen SE, Piechnik SK, Neubauer S, Evangelou E, Dehghan A, O'Regan DP, Wilkins MR, Guo Y, Matthews PM, Rueckert Det al., 2020,

    A population-based phenome-wide association study of cardiac and aortic structure and function

    , Nature Medicine, Vol: 26, Pages: 1654-1662, ISSN: 1078-8956

    Differences in cardiac and aortic structure and function are associated with cardiovascular diseases and a wide range of other types of disease. Here we analyzed cardiovascular magnetic resonance images from a population-based study, the UK Biobank, using an automated machine-learning-based analysis pipeline. We report a comprehensive range of structural and functional phenotypes for the heart and aorta across 26,893 participants, and explore how these phenotypes vary according to sex, age and major cardiovascular risk factors. We extended this analysis with a phenome-wide association study, in which we tested for correlations of a wide range of non-imaging phenotypes of the participants with imaging phenotypes. We further explored the associations of imaging phenotypes with early-life factors, mental health and cognitive function using both observational analysis and Mendelian randomization. Our study illustrates how population-based cardiac and aortic imaging phenotypes can be used to better define cardiovascular disease risks as well as heart–brain health interactions, highlighting new opportunities for studying disease mechanisms and developing image-based biomarkers.
  • Journal article
    Moriconi R, Deisenroth M, Karri S, 2020,

    High-dimensional Bayesian optimization usinglow-dimensional feature spaces

    , Machine Learning, Vol: 109, Pages: 1925-1943, ISSN: 0885-6125

    Bayesian optimization (BO) is a powerful approach for seeking the global optimum of expensive black-box functions and has proven successful for fine tuning hyper-parameters of machine learning models. However, BO is practically limited to optimizing 10–20 parameters. To scale BO to high dimensions, we usually make structural assumptions on the decomposition of the objective and/or exploit the intrinsic lower dimensionality of the problem, e.g. by using linear projections. We could achieve a higher compression rate with nonlinear projections, but learning these nonlinear embeddings typically requires much data. This contradicts the BO objective of a relatively small evaluation budget. To address this challenge, we propose to learn a low-dimensional feature space jointly with (a) the response surface and (b) a reconstruction mapping. Our approach allows for optimization of BO’s acquisition function in the lower-dimensional subspace, which significantly simplifies the optimization problem. We reconstruct the original parameter space from the lower-dimensional subspace for evaluating the black-box function. For meaningful exploration, we solve a constrained optimization problem.
  • Journal article
    Cofré R, Herzog R, Mediano PAM, Piccinini J, Rosas FE, Sanz Perl Y, Tagliazucchi Eet al., 2020,

    Whole-brain models to explore altered states of consciousness from the bottom up

    , Brain Sciences, Vol: 10, ISSN: 2076-3425

    The scope of human consciousness includes states departing from what most of us experience as ordinary wakefulness. These altered states of consciousness constitute a prime opportunity to study how global changes in brain activity relate to different varieties of subjective experience. We consider the problem of explaining how global signatures of altered consciousness arise from the interplay between large-scale connectivity and local dynamical rules that can be traced to known properties of neural tissue. For this purpose, we advocate a research program aimed at bridging the gap between bottom-up generative models of whole-brain activity and the top-down signatures proposed by theories of consciousness. Throughout this paper, we define altered states of consciousness, discuss relevant signatures of consciousness observed in brain activity, and introduce whole-brain models to explore the biophysics of altered consciousness from the bottom-up. We discuss the potential of our proposal in view of the current state of the art, give specific examples of how this research agenda might play out, and emphasize how a systematic investigation of altered states of consciousness via bottom-up modeling may help us better understand the biophysical, informational, and dynamical underpinnings of consciousness.
  • Journal article
    Fernando S, AmadorDíazLópez J, Şerban O, Gómez-Romero J, Molina-Solana M, Guo Yet al., 2020,

    Towards a large-scale twitter observatory for political events

    , Future Generation Computer Systems, Vol: 110, Pages: 976-983, ISSN: 0167-739X

    Explosion in usage of social media has made its analysis a relevant topic of interest, and particularly so in the political science area. Within Data Science, no other techniques are more widely accepted and appealing than visualisation. However, with datasets growing in size, visualisation tools also require a paradigm shift to remain useful in big data contexts. This work presents our proposal for a Large-Scale Twitter Observatory that enables researchers to efficiently retrieve, analyse and visualise data from this social network to gain actionable insights and knowledge related with political events. In addition to describing the supporting technologies, we put forward a working pipeline and validate the setup with different examples.
  • Journal article
    Meyer H, Dawes T, Serrani M, Bai W, Tokarczuk P, Cai J, Simoes Monteiro de Marvao A, Henry A, Lumbers T, Gierten J, Thumberger T, Wittbrodt J, Ware J, Rueckert D, Matthews P, Prasad S, Costantino M, Cook S, Birney E, O'Regan Det al., 2020,

    Genetic and functional insights into the fractal structure of the heart

    , Nature, Vol: 584, Pages: 589-594, ISSN: 0028-0836

    The inner surfaces of the human heart are covered by a complex network of muscular strands that is thought to be a vestigeof embryonic development.1,2 The function of these trabeculae in adults and their genetic architecture are unknown. Toinvestigate this we performed a genome-wide association study using fractal analysis of trabecular morphology as animage-derived phenotype in 18,096 UK Biobank participants. We identified 16 significant loci containing genes associatedwith haemodynamic phenotypes and regulation of cytoskeletal arborisation.3,4 Using biomechanical simulations and humanobservational data, we demonstrate that trabecular morphology is an important determinant of cardiac performance. Throughgenetic association studies with cardiac disease phenotypes and Mendelian randomisation, we find a causal relationshipbetween trabecular morphology and cardiovascular disease risk. These findings suggest an unexpected role for myocardialtrabeculae in the function of the adult heart, identify conserved pathways that regulate structural complexity, and reveal theirinfluence on susceptibility to disease
  • Journal article
    Chen J, Wang Z, Zhu T, Rosas FEet al., 2020,

    Recommendation algorithm in double-layer network based on vector dynamic evolution clustering and attention mechanism

    , Complexity, Vol: 2020, Pages: 1-19, ISSN: 1076-2787

    The purpose of recommendation systems is to help users find effective information quickly and conveniently and also to present the items that users are interested in. While the literature of recommendation algorithms is vast, most collaborative filtering recommendation approaches attain low recommendation accuracies and are also unable to track temporal changes of preferences. Additionally, previous differential clustering evolution processes relied on a single-layer network and used a single scalar quantity to characterise the status values of users and items. To address these limitations, this paper proposes an effective collaborative filtering recommendation algorithm based on a double-layer network. This algorithm is capable of fully exploring dynamical changes of user preference over time and integrates the user and item layers via an attention mechanism to build a double-layer network model. Experiments on Movielens, CiaoDVD, and Filmtrust datasets verify the effectiveness of our proposed algorithm. Experimental results show that our proposed algorithm can attain a better performance than other state-of-the-art algorithms.
  • Conference paper
    Chen C, Qin C, Qiu H, Ouyang C, Wang S, Chen L, Tarroni G, Bai W, Rueckert Det al., 2020,

    Realistic adversarial data augmentation for MR image segmentation

    , International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI)

    Neural network-based approaches can achieve high accuracy in various medicalimage segmentation tasks. However, they generally require large labelleddatasets for supervised learning. Acquiring and manually labelling a largemedical dataset is expensive and sometimes impractical due to data sharing andprivacy issues. In this work, we propose an adversarial data augmentationmethod for training neural networks for medical image segmentation. Instead ofgenerating pixel-wise adversarial attacks, our model generates plausible andrealistic signal corruptions, which models the intensity inhomogeneities causedby a common type of artefacts in MR imaging: bias field. The proposed methoddoes not rely on generative networks, and can be used as a plug-in module forgeneral segmentation networks in both supervised and semi-supervised learning.Using cardiac MR imaging we show that such an approach can improve thegeneralization ability and robustness of models as well as provide significantimprovements in low-data scenarios.

This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.

Request URL: http://wlsprd.imperial.ac.uk:80/respub/WEB-INF/jsp/search-t4-html.jsp Request URI: /respub/WEB-INF/jsp/search-t4-html.jsp Query String: id=607&limit=15&respub-action=search.html Current Millis: 1624150986985 Current Time: Sun Jun 20 02:03:06 BST 2021