Search

The progenitors of Type Ia supernovae (SNe Ia) are debated, particularly the evolutionary state of the binary companion that donates mass to the exploding carbon-oxygen white dwarf. In previous work, we presented hydrodynamic models and optically thin radio synchrotron light-curves of SNe Ia interacting with detached, confined shells of CSM, representing CSM shaped by novae. In this work, we extend these light-curves to the optically thick regime, considering both synchrotron self-absorption and free-free absorption. We obtain simple formulae to describe the evolution of optical depth seen in the simulations, allowing optically thick light-curves to be approximated for arbitrary shell properties. We then demonstrate the use of this tool by interpreting published radio data. First, we consider the non-detection of PTF11kx - an SN Ia known to have a detached, confined shell - and find that the non-detection is consistent with current models for its CSM, and that observations at a later time would have been useful for this event. Secondly, we statistically analyze an ensemble of radio non-detections for SNe Ia with no signatures of interaction, and find that shells with masses $(10^{-4}-0.3)~M_\odot$ located $(10^{15}-10^{16})$ cm from the progenitor are currently not well constrained by radio datasets, due to their dim, rapidly-evolving light-curves., 16 pages, 7 figures, ApJ accepted; companion Python tools available on github (see text)

Breast cancer is the second most common type of cancer of women in the U.S. behind skin cancer. Early detection and characterization of breast masses is critical for effective diagnosis and treatment of breast cancer. Computer-aided breast mass characterization methods would help to improve the accuracy of diagnoses, their reproducibility, and the throughput of breast cancer screening workflows. In this work, we introduce sparse representations of deep learning features for separation of malignant from benign breast masses in mammograms. We expect that the use of deep feature-based dictionaries will produce better benign/malignant class separation than straightforward sparse representation techniques, and fine-tuned convolutional neural networks (CNNs). We performed 10- and 30-fold cross-validation experiments for classification of benign and malignant breast masses on the MIAS and DDSM mammographic datasets. The results show that the proposed deep feature sparse analysis produces better classification rates than conventional sparse representations and fine-tuned CNNs. The top areas under the curve (AUC) for the receiver operating curve are 80.64% for 10-fold and 97.44% for 30-fold cross-validation in MIAS, and 77.29% for 10-fold and 76.02% for 30-fold cross-validation in DDSM. The main advantages of this approach are that it employs dictionaries of deep network features that are sparse in nature and that it alleviates the need for large volumes of training data and lengthy training procedures. The interesting results from this work prompt further exploration of the relationship between sparse optimization problems and deep learning.

Observations of core-collapse supernovae (CCSNe) reveal a wealth of information about the dynamics of the supernova ejecta and its composition but very little direct information about the progenitor. Constraining properties of the progenitor and the explosion requires coupling the observations with a theoretical model of the explosion. Here, we begin with the CCSN simulations of Couch et al 2020 ApJ 890 127, which use a non-parametric treatment of the neutrino transport while also accounting for turbulence and convection. In this work we use the SuperNova Explosion Code to evolve the CCSN hydrodynamics to later times and compute bolometric light curves. Focusing on SNe IIP, we then (1) directly compare the theoretical STIR explosions to observations and (2) assess how properties of the progenitor's core can be estimated from optical photometry in the plateau phase alone. First, the distribution of plateau luminosities (L$_{50}$) and ejecta velocities achieved by our simulations is similar to the observed distributions. Second, we fit our models to the light curves and velocity evolution of some well-observed SNe. Third, we recover well-known correlations, as well as the difficulty of connecting any one SN property to zero-age main sequence mass. Finally, we show that there is a usable, linear correlation between iron core mass and L$_{50}$ such that optical photometry alone of SNe IIP can give us insights into the cores of massive stars. Illustrating this by application to a few SNe, we find iron core masses of 1.3-1.5 solar masses with typical errors of ~0.05 solar masses. Data are publicly available online (\url{https://doi.org/10.5281/zenodo.6631964})., 23 pages, Accepted to ApJ. Data available online https://doi.org/10.5281/zenodo.6631964

The topic of sparse representation of samples in high dimensional spaces has attracted growing interest during the past decade. In this work, we develop sparse representation-based methods for classification of clinical imaging patterns into healthy and diseased states. We propose a spatial block decomposition method to address irregularities of the approximation problem and to build an ensemble of classifiers that we expect to yield more accurate numerical solutions than conventional sparse analyses of the complete spatial domain of the images. We introduce two classification decision strategies based on maximum a posteriori probability (BBMAP), or a log likelihood function (BBLL) and an approach to adjusting the classification decision criteria. To evaluate the performance of the proposed approach we used cross-validation techniques on imaging datasets with disease class labels. We first applied the proposed approach to diagnosis of osteoporosis using bone radiographs. In this problem we assume that changes in trabecular bone connectivity can be captured by intensity patterns. The second application domain is separation of breast lesions into benign and malignant categories in mammograms. The object classes in both of these applications are not linearly separable, and the classification accuracy may depend on the lesion size in the second application. Our results indicate that the proposed integrative sparse analysis addresses the ill-posedness of the approximation problem and produces very good class separation for trabecular bone characterization and for breast lesion characterization. Our approach yields higher classification rates than conventional sparse classification and previously published convolutional neural networks (CNNs) that we fine-tuned for our datasets, or utilized for feature extraction. The BBLL technique also produced higher classification rates than learners using hand-crafted texture features, and the Bag of Keypoints, which is a sophisticated patch-based method. Furthermore, our comparative experiments showed that the BBLL function may yield more accurate classification than BBMAP, because BBLL accounts for possible estimation bias.

Explaining the observed diversity of supernovae (SNe) and the physics of explosion requires knowledge of their progenitor stars, which can be obtained by constraining the circumstellar medium (CSM). Models of the SN ejecta colliding with CSM are necessary to infer the structure of the CSM and tie it back to a progenitor model. Recent SNe I revealed CSM concentrated at a distance $r\sim10^16$ cm, for which models of SN interaction are extremely limited. In this paper, we assume the concentrated region is a "wall" representing swept-up material, and unswept material lies outside the wall. We simulate one-dimensional hydrodynamics of SNe Ia & Ib impacting 300 unique CSM configurations using RT1D, which captures the Rayleigh-Taylor instability. We find that the density ratio between the wall and ejecta -- denoted $A_0$ or "wall height" -- is key, and higher walls deviate more from self-similar evolution. Functional fits accounting for $A_0$ are presented for the forward shock radius evolution. We show that higher walls have more degeneracy between CSM properties in the deceleration parameter, slower shocks, deeper-probing reverse shocks, slower shocked ejecta, less ejecta mass than CSM in the shock, and more mixing of ejecta into the CSM at early times. We analyze observations of SN 2014C (Type Ib) and suggest that it had a moderately high wall ($10 < A_0 < 200$) and wind-like outer CSM. We also postulate an alternate interpretation for the radio data of SN 2014C, that the radio rise occurs in the wind rather than the wall. Finally, we find that hydrodynamic measurements at very late times cannot distinguish the presence of a wall, except perhaps as an anomalously wide shock region., 17 pages, 13 figures, accepted to ApJ

Rationale The topic of sparse representation of samples in high dimensional spaces has attracted growing interest during the past decade. In this work, we develop sparse representation-based methods for classification of radiological imaging patterns of breast lesions into benign and malignant states. Methods We propose a spatial block decomposition method to address irregularities of the approximation problem and to build an ensemble of classifiers (CL) that we expect to yield more accurate numerical solutions than conventional whole-region of interest (ROI) sparse analyses. We introduce two classification decision strategies based on maximum a posteriori probability (BBMAP-S), or a log likelihood function (BBLL-S). Results To evaluate the performance of the proposed approach we used cross-validation techniques on imaging datasets with disease class labels. We utilized the proposed approach for separation of breast lesions into benign and malignant categories in mammograms. The level of difficulty is high in this application and the accuracy may depend on the lesion size. Our results indicate that the proposed integrative sparse analysis addresses the ill-posedness of the approximation problem, producing AUC (area under the receiver operating curve) value of 89.1% for randomized 30-fold cross-validation. Conclusions Furthermore, our comparative experiments showed that the BBLL-S decision function may yield more accurate classification than BBMAP-S because BBLL-S accounts for possible estimation bias.

Despite their cosmological utility, the progenitors of Type Ia supernovae (SNe Ia) are still unknown, with many efforts focused on whether accretion from a nondegenerate companion can grow a carbon-oxygen white dwarf to near the Chandrasekhar mass. The association of SNe Ia resembling SN 1991T ("91T-like") with circumstellar interaction may be evidence for this "single-degenerate" channel. However, the observed circumstellar medium (CSM) in these interacting systems is unlike a stellar wind -- of particular interest, it is sometimes detached from the stellar surface, residing at $\sim 10^{16}~{\rm cm}$. A Hubble Space Telescope (HST) program to discover detached CSM around 91T-like SNe Ia successfully discovered interaction nearly two years after explosion in SN 2015cp (Graham et al., 2018). In this work, we present radio and X-ray follow-up observations of SN 2015cp and analyze them in the framework of Harris, Nugent, & Kasen (2016) to limit the properties of a constant-density CSM shell in this system. Assuming the HST detection was shortly after the shock crossed the CSM, we constrain the total CSM mass in this system to be $< 0.5~{\rm M_\odot}$. This limit is comparable to the CSM mass of supernova PTF11kx, but does not rule out lower masses predicted for recurrent novae. From lessons learned modeling PTF11kx and SN 2015cp, we suggest a strategy for future observations of these events to increase the sample of known interacting SNe Ia., 13 pages, 5 figures, published in The Astrophysical Journal

Some active galactic nuclei (AGN) are surrounded by extended emission-line regions (EELRs), which trace both the illumination pattern of escaping radiation and its history over the light-travel time from the AGN to the gas. From a new set of such EELRs, we present evidence that the AGN in many Seyfert galaxies undergo luminous episodes 20,000-200,000 years in duration. Motivated by the discovery of the spectacular nebula known as Hanny's Voorwerp, ionized by a powerful AGN which has apparently faded dramatically within ~ 100,000 years, Galaxy Zoo volunteers have carried out both targeted and serendipitous searches for similar emission-line clouds around low-redshift galaxies.We present the resulting list of candidates and describe spectroscopy identifying 19 galaxies with AGN-ionized regions at projected radii > 10 kpc. This search recovered known EELRs and identified additional previously unknown cases, one with detected emission to r = 37 kpc. At least 14/19 are in interacting or merging systems; tidal tails are a prime source of extraplanar ionized gas. We see a mix of one- and two-sided structures, with observed cone angles from 23-112 degrees. We consider the energy balance in the ionized clouds, with lower and upper bounds on ionizing luminosity from recombination and ionization-parameter arguments, and estimate the luminosity of the core from the far-infrared data. The implied ratio of ionizing radiation seen by the clouds to that emitted by the nucleus, for a constant nuclear source, ranges from 0.02 to > 12; 7/19 exceed unity. Small values imply heavily obscured AGN. However, large values may require that the AGN has faded over tens of thousands of years, giving us several examples of systems in which such dramatic long-period variation has occurred; this is the only current technique for addressing these timescales in AGN history. (Abridged)

We report on the discovery of SN2014J in the nearby galaxy M82. Given its proximity, it offers the best opportunity to date to study a thermonuclear supernova over a wide range of the electromagnetic spectrum. The first set of optical, near-IR and mid-IR observations of SN2014J, orchestrated by the intermediate Palomar Transient Factory (iPTF), show that SN2014J is a spectroscopically normal Type Ia supernova, albeit exhibiting high-velocity features in its spectrum and heavily reddened by dust in the host galaxy. Our earliest detections start just hours after the fitted time of explosion. We use high-resolution optical spectroscopy to analyze the dense intervening material and do not detect any evolution in the resolved absorption features during the lightcurve rise. Similarly to other highly reddened Type Ia supernovae, a low value of total-to-selective extinction, Rv < 2, provides the best match to our observations. We also study pre-explosion optical and near-IR images from HST with special emphasis on the sources nearest to the SN location., Accepted for publication in ApJL

For decades, a wide variety of observations spanning the radio through optical and on to the x-ray have attempted to uncover signs of type Ia supernovae (SNe Ia) interacting with a circumstellar medium (CSM). The goal of these studies is to constrain the nature of the hypothesized SN Ia mass-donor companion. A continuous CSM is typically assumed when interpreting observations of interaction. However, while such models have been successfully applied to core-collapse SNe, the assumption of continuity may not be accurate for SNe Ia, as shells of CSM could be formed by pre-supernova eruptions (novae). In this work, we model the interaction of SNe with a spherical, low density, finite-extent CSM and create a suite of synthetic radio synchrotron light curves. We find that CSM shells produce sharply peaked light curves, and identify a fiducial set of models that all obey a common evolution and can be used to generate radio light curves for interaction with an arbitrary shell. The relations obeyed by the fiducial models can be used to deduce CSM properties from radio observations; we demonstrate this by applying them to the non-detections of SN 2011fe and SN 2014J. Finally, we explore a multiple shell CSM configuration and describe its more complicated dynamics and resultant radio light curves., 15 pages, 11 figures, ApJ accepted

We derive spatially resolved stellar kinematics for a sample of 84 out of 104 observed local (0.02 < z < 0.09) galaxies hosting type-1 active galactic nuclei (AGNs), based on long-slit spectra obtained at the 10 m W. M. Keck-1 Telescope. In addition to providing central stellar velocity dispersions, we measure major axis rotation curves and velocity dispersion profiles using three separate wavelength regions, including the prominent Ca H&K, Mg Ib, and Ca II NIR stellar features. In this paper, we compare kinematic measurements of stellar velocity dispersion obtained for different apertures, wavelength regions, and signal-to-noise ratios, and provide recipes to cross-calibrate the measurements reducing systematic effects to the level of a few percent. We also provide simple recipes based on readily observable quantities such as global colors and Ca H&K equivalent width that will allow observers of high-redshift AGN hosts to increase the probability of obtaining reliable stellar kinematic measurements from unresolved spectra in the region surrounding the Ca H&K lines. In subsequent papers in this series, we will combine this unprecedented spectroscopic data set with surface photometry and black hole mass measurements to study in detail the scaling relations between host galaxy properties and black hole mass.