A5070506182- Neural dynamics and brain function
- Advanced Memory and Neural Computing
- Photoreceptor and optogenetics research
- Cell Image Analysis Techniques
- Neuroscience and Neural Engineering
- Neural Networks and Applications
- Domain Adaptation and Few-Shot Learning
- Visual perception and processing mechanisms
- Image Processing Techniques and Applications
- Face Recognition and Perception
- Scientific Computing and Data Management
- Genetic Associations and Epidemiology
- Underwater Acoustics Research
- Genetic factors in colorectal cancer
- Wnt/β-catenin signaling in development and cancer
- Cancer-related gene regulation
- Molecular Biology Techniques and Applications
- Underwater Vehicles and Communication Systems
- Distributed and Parallel Computing Systems
- Hippo pathway signaling and YAP/TAZ
- Protein Structure and Dynamics
- Gene expression and cancer classification
- Evolution and Genetic Dynamics
- Advanced Radiotherapy Techniques
- Environmental Monitoring and Data Management
Cambridge University Hospitals NHS Foundation Trust
2024
University of Sussex
2022-2024
University of Bristol
2019-2023
Sussex County Community College
2023
University of Exeter
2020-2022
Living Systems (United States)
2020
Imperial College London
2015-2018
NIHR Imperial Biomedical Research Centre
2015-2018
Institute of Biomedical Science
2016
University of Oxford
2012-2015
Abstract Deep neural networks (DNNs) have had extraordinary successes in classifying photographic images of objects and are often described as the best models biological vision. This conclusion is largely based on three sets findings: (1) DNNs more accurate than any other model taken from various datasets, (2) do job predicting pattern human errors behavioral (3) brain signals response to datasets (e.g., single cell responses or fMRI data). However, these not test hypotheses regarding what...
Early esophagogastric cancer (OGC) stage presents with nonspecific symptoms.The aim of this study was to determine the accuracy a breath test for diagnosis OGC in multicenter validation study.Patient recruitment diagnostic conducted at 3 London hospital sites, samples returned central laboratory selected ion flow tube mass spectrometry (SIFT-MS) analysis. Based on 1:1 cancer:control ratio, and maintaining sensitivity specificity 80%, sample size required 325 patients. All patients were...
Chaste (Cancer, Heart And Soft Tissue Environment) is an open source simulation package for the numerical solution of mathematical models arising in physiology and biology. To date, development has been driven primarily by applications that include continuum modelling cardiac electrophysiology ('Cardiac Chaste'), discrete cell-based soft tissues ('Cell-based ventilation lungs ('Lung Chaste'). Cardiac addresses need a high-performance, generic, verified framework freely available to...
Abstract Wnt signaling regulates cell proliferation and differentiation as well migration polarity during development. However, it is still unclear how the ligand distribution precisely controlled to fulfil these functions. Here, we show that planar protein Vangl2 of by cytonemes. In zebrafish epiblast cells, mouse intestinal telocytes human gastric cancer activation generates extremely long cytonemes, which branch deliver multiple cells. The Vangl2-activated cytonemes increase Wnt/β-catenin...
Deep neural networks (DNNs) have had extraordinary successes in classifying photographic images of objects and are often described as the best models biological vision. This conclusion is largely based on three sets findings: (1) DNNs more accurate than any other model taken from various datasets, (2) do job predicting pattern human errors behavioral benchmark (3) brain signals response to datasets (e.g., single cell responses or fMRI data). However, most benchmarks report outcomes...
Antimicrobial Resistance is threatening our ability to treat common infectious diseases and overuse of antimicrobials human infections in hospitals accelerating this process. Clinical Decision Support Systems (CDSSs) have been proven enhance quality care by promoting change prescription practices through antimicrobial selection advice. However, bypassing an initial assessment determine the existence underlying disease that justifies need therapy might lead indiscriminate often unnecessary...
The quality and effectiveness of sensor information provided by mine-hunting autonomous underwater vehicles (AUVs) equipped with high-resolution sonars has improved drastically in recent years. In parallel, data rates have significantly increased resulting overload. Automatic target recognition (ATR) is regarded as a solution for this problem. This study describes specific ATR technique based on model matching application to data. A sonar generation synthetic aperture (SAS) images described...
Optogenetics has become a key tool for understanding the function of neural circuits and controlling their behavior. An array directly light driven opsins have been genetically isolated from several families organisms, with wide range temporal spectral properties. In order to characterize, understand apply these opsins, we present an integrated suite open-source, multi-scale computational tools called PyRhO. The purpose developing PyRhO is three-fold: (i) characterize new (and existing) by...
We present a reconfigurable neural processor for real-time simulation and prediction of opto-neural behaviour. combined detailed Hodgkin-Huxley CA3 neuron integrated with four-state Channelrhodopsin-2 (ChR2) model into silicon hardware. Our architecture consists Field Programmable Gated Array (FPGA) custom-built computing data-path, separate data management system memory approach based router. Advancements over previous work include the incorporation short long-term calcium light-dependent...
ORIGINAL RESEARCH article Front. Comput. Neurosci., 25 July 2012 Volume 6 - | https://doi.org/10.3389/fncom.2012.00046
This paper presents a new supervised classification approach for automated target recognition (ATR) in SAS images. The procedure starts with novel segmentation stage based on the Hilbert transform. A number of geometrical features are then extracted and used to classify observed objects against previously compiled database non-target features. proposed has been tested set 1528 simulated images created by NURC SIGMAS sonar model, achieving up 95% accuracy.
Convolutional neural networks (CNNs) are often described as promising models of human vision, yet they show many differences from abilities. We focus on a superhuman capacity top-performing CNNs, namely, their ability to learn very large datasets random patterns. verify that learning such tasks is extremely limited, even with few stimuli. argue the performance difference due CNNs' overcapacity and introduce biologically inspired mechanisms constrain it, while retaining good test set...
On several key issues we agree with the commentators. Perhaps most importantly, everyone seems to that psychology has an important role play in building better models of human vision, and (most) agrees (including us) deep neural networks (DNNs) will modelling vision going forward. But there are also disagreements about what for, how DNN-human correspondences should be evaluated, value alternative approaches, impact marketing hype literature. In our view, these latter contributing many...
Recent work shows that the developmental potential of progenitor cells in HH10 chick brain changes rapidly, accompanied by subtle morphology. This demands increased temporal resolution for studies at this stage, necessitating precise and unbiased staging. Here, we investigated whether could train a deep convolutional neural network to sub-stage brains using small dataset 151 expertly labelled images. By augmenting our images with biologically informed transformations data-driven...
This paper describes the design and operation of a system which can be used as Visual to Auditory Sensory Substitution Device (SSD), well front-end real-time retinal prosthesis (RP) or Vision Augmentation (VA) system. Such systems consist three components: sensory block capture visual scene, processing manage collected data generate stimulus patterns, an output block. For we use Dynamic Sensor (DVS) instead conventional camera. A microcontroller is block, receives asynchronous inputs from...
Abstract Clinical classification is essential for estimating disease prevalence but difficult, often requiring complex investigations. The widespread availability of population level genetic data makes novel stratification techniques a highly attractive alternative. We propose generalizable mathematical framework determining within cohort using risk scores. compare and evaluate methods based on the means scores’ distributions; Earth Mover’s Distance between linear combination kernel density...
Over successive stages, the ventral visual system of primate brain develops neurons that respond selectively to particular objects or faces with translation, size and view invariance. The powerful neural representations found in Inferotemporal cortex form a remarkably rapid robust basis for object recognition which belies difficulties faced by when learning natural environments. A central issue understanding process biological is how these learn separate from complex scenes composed multiple...
Abstract Clinical classification is essential for estimating disease prevalence but difficult, often requiring complex investigations. The widespread availability of population level genetic data makes novel stratification techniques a highly attractive alternative. We propose generalizable mathematical framework determining within cohort using risk scores. compare and evaluate methods based on the means scores’ distributions; Earth Mover’s Distance between linear combination kernel density...
Neurons in successive stages of the primate ventral visual pathway encode spatial structure objects. In this paper, we investigate through computer simulation how these cell firing properties may develop unsupervised visually-guided learning. Individual neurons model are shown to exploit statistical regularity and temporal continuity inputs during training learn that similar V4 TEO. conformation boundary contour elements at a particular position within an object regardless location on...
This demonstration shows a new type of front end for Retinal Prosthesis/Vision Augmentation (RP/VA) System, as well Visual to Auditory Sensory Substitution Device (SSD). Each system serves process visual scenes then present them in simplified form (augmented with auditory signals) assist visually impaired people. Both systems consist three components: sensory block capture the scene, processing manage collected data and generate stimulus patterns, an output block. Here we are presenting two...
We show how hand-centred visual representations could develop in the primate posterior parietal and premotor cortices during visually guided learning a self-organizing neural network model. The model incorporates trace feed-forward synaptic connections between successive neuronal layers. Trace encourages neurons to learn respond input images that tend occur close together time. assume sequences of eye movements are performed around individual scenes containing fixed hand-object...
Abstract Deep Convolutional Neural Networks (DNNs) have achieved superhuman accuracy on standard image classification benchmarks. Their success has reignited significant interest in their use as models of the primate visual system, bolstered by claims architectural and representational similarities. However, closer scrutiny these suggests that they rely various forms shortcut learning to achieve impressive performance, such using texture rather than shape information. Such superficial...