Showing 1–8 of 8 results for author: Irving, B

BASS. XXXVI. Constraining the Local Supermassive Black Hole - Halo Connection with BASS DR2 AGN

Authors: M. C. Powell, S. W. Allen, T. Caglar, N. Cappelluti, F. Harrison, B. E. Irving, M. J. Koss, A. B. Mantz, K. Oh, C. Ricci, R. J. Shaper, D. Stern, B. Trakhtenbrot, C. M. Urry, J. Wong

Abstract: We investigate the connection between supermassive black holes (SMBHs) and their host dark matter halos in the local universe using the clustering statistics and luminosity function of AGN from the Swift/BAT AGN Spectroscopic survey (BASS DR2). By forward-modeling AGN activity into snapshot halo catalogs from N-body simulations, we test a scenario in which SMBH mass correlates with dark matter (su… ▽ More We investigate the connection between supermassive black holes (SMBHs) and their host dark matter halos in the local universe using the clustering statistics and luminosity function of AGN from the Swift/BAT AGN Spectroscopic survey (BASS DR2). By forward-modeling AGN activity into snapshot halo catalogs from N-body simulations, we test a scenario in which SMBH mass correlates with dark matter (sub)halo mass for fixed stellar mass. We compare this to a model absent of this correlation, where stellar mass alone determines the SMBH mass. We find that while both simple models are able to largely reproduce the abundance and overall clustering of AGN, the model in which black hole mass is tightly correlated with halo mass is preferred by the data by $1.8σ$. When including an independent measurement on the black hole mass-halo mass correlation, this model is preferred by $4.6σ$. We show that the clustering trends with black hole mass can further break the degeneracies between the two scenarios, and that our preferred model reproduces the measured clustering differences on 1-halo scales between large and small black hole masses. These results indicate that the halo binding energy is fundamentally connected to the growth of supermassive black holes. △ Less

Submitted 14 October, 2022; v1 submitted 6 September, 2022; originally announced September 2022.

Comments: Accepted for publication in ApJ

Journal ref: ApJ 938, 77 (2022)

Using Deep Learning-based Features Extracted from CT scans to Predict Outcomes in COVID-19 Patients

Authors: Sai Vidyaranya Nuthalapati, Marcela Vizcaychipi, Pallav Shah, Piotr Chudzik, Chee Hau Leow, Paria Yousefi, Ahmed Selim, Keiran Tait, Ben Irving

Abstract: The COVID-19 pandemic has had a considerable impact on day-to-day life. Tackling the disease by providing the necessary resources to the affected is of paramount importance. However, estimation of the required resources is not a trivial task given the number of factors which determine the requirement. This issue can be addressed by predicting the probability that an infected patient requires Inten… ▽ More The COVID-19 pandemic has had a considerable impact on day-to-day life. Tackling the disease by providing the necessary resources to the affected is of paramount importance. However, estimation of the required resources is not a trivial task given the number of factors which determine the requirement. This issue can be addressed by predicting the probability that an infected patient requires Intensive Care Unit (ICU) support and the importance of each of the factors that influence it. Moreover, to assist the doctors in determining the patients at high risk of fatality, the probability of death is also calculated. For determining both the patient outcomes (ICU admission and death), a novel methodology is proposed by combining multi-modal features, extracted from Computed Tomography (CT) scans and Electronic Health Record (EHR) data. Deep learning models are leveraged to extract quantitative features from CT scans. These features combined with those directly read from the EHR database are fed into machine learning models to eventually output the probabilities of patient outcomes. This work demonstrates both the ability to apply a broad set of deep learning methods for general quantification of Chest CT scans and the ability to link these quantitative metrics to patient outcomes. The effectiveness of the proposed method is shown by testing it on an internally curated dataset, achieving a mean area under Receiver operating characteristic curve (AUC) of 0.77 on ICU admission prediction and a mean AUC of 0.73 on death prediction using the best performing classifiers. △ Less

Submitted 10 May, 2022; originally announced May 2022.

Deep Semi-Supervised Embedded Clustering (DSEC) for Stratification of Heart Failure Patients

Authors: Oliver Carr, Stojan Jovanovic, Luca Albergante, Fernando Andreotti, Robert Dürichen, Nadia Lipunova, Janie Baxter, Rabia Khan, Benjamin Irving

Abstract: Determining phenotypes of diseases can have considerable benefits for in-hospital patient care and to drug development. The structure of high dimensional data sets such as electronic health records are often represented through an embedding of the data, with clustering methods used to group data of similar structure. If subgroups are known to exist within data, supervised methods may be used to in… ▽ More Determining phenotypes of diseases can have considerable benefits for in-hospital patient care and to drug development. The structure of high dimensional data sets such as electronic health records are often represented through an embedding of the data, with clustering methods used to group data of similar structure. If subgroups are known to exist within data, supervised methods may be used to influence the clusters discovered. We propose to extend deep embedded clustering to a semi-supervised deep embedded clustering algorithm to stratify subgroups through known labels in the data. In this work we apply deep semi-supervised embedded clustering to determine data-driven patient subgroups of heart failure from the electronic health records of 4,487 heart failure and control patients. We find clinically relevant clusters from an embedded space derived from heterogeneous data. The proposed algorithm can potentially find new undiagnosed subgroups of patients that have different outcomes, and, therefore, lead to improved treatments. △ Less

Submitted 17 January, 2021; v1 submitted 24 December, 2020; originally announced December 2020.

Comments: 6 pages, 2 figures, HSYS workshop at ICML conference

Prediction of the onset of cardiovascular diseases from electronic health records using multi-task gated recurrent units

Authors: Fernando Andreotti, Frank S. Heldt, Basel Abu-Jamous, Ming Li, Avelino Javer, Oliver Carr, Stojan Jovanovic, Nadezda Lipunova, Benjamin Irving, Rabia T. Khan, Robert Dürichen

Abstract: In this work, we propose a multi-task recurrent neural network with attention mechanism for predicting cardiovascular events from electronic health records (EHRs) at different time horizons. The proposed approach is compared to a standard clinical risk predictor (QRISK) and machine learning alternatives using 5-year data from a NHS Foundation Trust. The proposed model outperforms standard clinical… ▽ More In this work, we propose a multi-task recurrent neural network with attention mechanism for predicting cardiovascular events from electronic health records (EHRs) at different time horizons. The proposed approach is compared to a standard clinical risk predictor (QRISK) and machine learning alternatives using 5-year data from a NHS Foundation Trust. The proposed model outperforms standard clinical risk scores in predicting stroke (AUC=0.85) and myocardial infarction (AUC=0.89), considering the largest time horizon. Benefit of using an \gls{mt} setting becomes visible for very short time horizons, which results in an AUC increase between 2-6%. Further, we explored the importance of individual features and attention weights in predicting cardiovascular events. Our results indicate that the recurrent neural network approach benefits from the hospital longitudinal information and demonstrates how machine learning techniques can be applied to secondary care. △ Less

Submitted 16 July, 2020; originally announced July 2020.

Comments: 5 pages, 2 figures, 2 tables, submitted at Healthcare Systems, Population Health, and the Role of Health-Tech - ICML 2020

Camouflage of the Phase Transition to Quark Matter in Neutron Stars

Authors: Wei Wei, Bryen Irving, Thomas Klähn, Prashanth Jaikumar

Abstract: It has been known for some time that compact stars containing quark matter can masquerade as neutron stars in the range of measured mass and radius, making it difficult to draw firm conclusions on the phases of matter present inside the star. Using the vector-enhanced Bag model (vBag), we examine mass-radius and mass-compactness relations with Maxwell and Gibbs construction for hybrid stars with t… ▽ More It has been known for some time that compact stars containing quark matter can masquerade as neutron stars in the range of measured mass and radius, making it difficult to draw firm conclusions on the phases of matter present inside the star. Using the vector-enhanced Bag model (vBag), we examine mass-radius and mass-compactness relations with Maxwell and Gibbs construction for hybrid stars with transitions from nuclear matter to two or three-flavor quark matter, including sequential transitions. Not only can stable hybrid stars with either two or three-flavor quark matter mimic neutron stars (the traditional masquerade), it appears as well difficult to distinguish two-flavor from three-flavor quark matter even in cases where a phase transition can be said to have occurred, as in the presence of a distinct kink in the mass-radius relation. Furthermore, allowing for sequential flavor transitions, we find that the transition into an unstable branch can be caused by either a transition from a nuclear to unstable quark matter or the sequential transition from nuclear to stable but "masquerading" two-flavor to unstable three-flavor quark matter. Addressing chiral restoration as well as quark deconfinement in a model of the phase transition, as the vBag does, adds further flexibility to the high-density equation of state, motivating caution in using even high-precision $M$-$R$ data to draw firm conclusions on the nature of phases and phase transitions in neutron stars. △ Less

Submitted 23 November, 2018; originally announced November 2018.

Comments: 10 pages, 5 figure panels

Extracting 3D Vascular Structures from Microscopy Images using Convolutional Recurrent Networks

Authors: Russell Bates, Benjamin Irving, Bostjan Markelc, Jakob Kaeppler, Ruth Muschel, Vicente Grau, Julia A. Schnabel

Abstract: Vasculature is known to be of key biological significance, especially in the study of cancer. As such, considerable effort has been focused on the automated measurement and analysis of vasculature in medical and pre-clinical images. In tumors in particular, the vascular networks may be extremely irregular and the appearance of the individual vessels may not conform to classical descriptions of vas… ▽ More Vasculature is known to be of key biological significance, especially in the study of cancer. As such, considerable effort has been focused on the automated measurement and analysis of vasculature in medical and pre-clinical images. In tumors in particular, the vascular networks may be extremely irregular and the appearance of the individual vessels may not conform to classical descriptions of vascular appearance. Typically, vessels are extracted by either a segmentation and thinning pipeline, or by direct tracking. Neither of these methods are well suited to microscopy images of tumor vasculature. In order to address this we propose a method to directly extract a medial representation of the vessels using Convolutional Neural Networks. We then show that these two-dimensional centerlines can be meaningfully extended into 3D in anisotropic and complex microscopy images using the recently popularized Convolutional Long Short-Term Memory units (ConvLSTM). We demonstrate the effectiveness of this hybrid convolutional-recurrent architecture over both 2D and 3D convolutional comparators. △ Less

Submitted 26 May, 2017; originally announced May 2017.

Comments: The article has been submitted to IEEE TMI

maskSLIC: Regional Superpixel Generation with Application to Local Pathology Characterisation in Medical Images

Authors: Benjamin Irving

Abstract: Supervoxel methods such as Simple Linear Iterative Clustering (SLIC) are an effective technique for partitioning an image or volume into locally similar regions, and are a common building block for the development of detection, segmentation and analysis methods. We introduce maskSLIC an extension of SLIC to create supervoxels within regions-of-interest, and demonstrate,on examples from 2-dimension… ▽ More Supervoxel methods such as Simple Linear Iterative Clustering (SLIC) are an effective technique for partitioning an image or volume into locally similar regions, and are a common building block for the development of detection, segmentation and analysis methods. We introduce maskSLIC an extension of SLIC to create supervoxels within regions-of-interest, and demonstrate,on examples from 2-dimensions to 4-dimensions, that maskSLIC overcomes issues that affect SLIC within an irregular mask. We highlight the benefits of this method through examples, and show that it is able to better represent underlying tumour subregions and achieves significantly better results than SLIC on the BRATS 2013 brain tumour challenge data (p=0.001) - outperforming SLIC on 18/20 scans. Finally, we show an application of this method for the analysis of functional tumour subregions and demonstrate that it is more effective than voxel clustering. △ Less

Submitted 9 February, 2017; v1 submitted 30 June, 2016; originally announced June 2016.

Comments: The article has been submitted to IEEE TPAMI

Pieces-of-parts for supervoxel segmentation with global context: Application to DCE-MRI tumour delineation

Authors: Benjamin Irving, James M Franklin, Bartlomiej W Papiez, Ewan M Anderson, Ricky A Sharma, Fergus V Gleeson, Sir Michael Brady, Julia A Schnabel

Abstract: Rectal tumour segmentation in dynamic contrast-enhanced MRI (DCE-MRI) is a challenging task, and an automated and consistent method would be highly desirable to improve the modelling and prediction of patient outcomes from tissue contrast enhancement characteristics - particularly in routine clinical practice. A framework is developed to automate DCE-MRI tumour segmentation, by introducing: perfus… ▽ More Rectal tumour segmentation in dynamic contrast-enhanced MRI (DCE-MRI) is a challenging task, and an automated and consistent method would be highly desirable to improve the modelling and prediction of patient outcomes from tissue contrast enhancement characteristics - particularly in routine clinical practice. A framework is developed to automate DCE-MRI tumour segmentation, by introducing: perfusion-supervoxels to over-segment and classify DCE-MRI volumes using the dynamic contrast enhancement characteristics; and the pieces-of-parts graphical model, which adds global (anatomic) constraints that further refine the supervoxel components that comprise the tumour. The framework was evaluated on 23 DCE-MRI scans of patients with rectal adenocarcinomas, and achieved a voxelwise area-under the receiver operating characteristic curve (AUC) of 0.97 compared to expert delineations. Creating a binary tumour segmentation, 21 of the 23 cases were segmented correctly with a median Dice similarity coefficient (DSC) of 0.63, which is close to the inter-rater variability of this challenging task. A sec- ond study is also included to demonstrate the method's generalisability and achieved a DSC of 0.71. The framework achieves promising results for the underexplored area of rectal tumour segmentation in DCE-MRI, and the methods have potential to be applied to other DCE-MRI and supervoxel segmentation problems △ Less

Submitted 18 April, 2016; originally announced April 2016.

Comments: accepted for publication in the Journal of Medical Image Analysis