Search

In the upcoming decades large facilities, such as the SKA, will provide resolved observations of the kinematics of millions of galaxies. In order to assist in the timely exploitation of these vast datasets we explore the use of a self-supervised, physics aware neural network capable of Bayesian kinematic modelling of galaxies. We demonstrate the network's ability to model the kinematics of cold gas in galaxies with an emphasis on recovering physical parameters and accompanying modelling errors. The model is able to recover rotation curves, inclinations and disc scale lengths for both CO and HI data which match well with those found in the literature. The model is also able to provide modelling errors over learned parameters thanks to the application of quasi-Bayesian Monte-Carlo dropout. This work shows the promising use of machine learning, and in particular self-supervised neural networks, in the context of kinematically modelling galaxies. This work represents the first steps in applying such models for kinematic fitting and we propose that variants of our model would seem especially suitable for enabling emission-line science from upcoming surveys with e.g. the SKA, allowing fast exploitation of these large datasets.

Next generation interferometers, such as the Square Kilometre Array, are set to obtain vast quantities of information about the kinematics of cold gas in galaxies. Given the volume of data produced by such facilities astronomers will need fast, reliable, tools to informatively filter and classify incoming data in real time. In this paper, we use machine learning techniques with a hydrodynamical simulation training set to predict the kinematic behaviour of cold gas in galaxies and test these models on both simulated and real interferometric data. Using the power of a convolutional autoencoder we embed kinematic features, unattainable by the human eye or standard tools, into a three-dimensional space and discriminate between disturbed and regularly rotating cold gas structures. Our simple binary classifier predicts the circularity of noiseless, simulated, galaxies with a recall of $85\%$ and performs as expected on observational CO and HI velocity maps, with a heuristic accuracy of $95\%$. The model output exhibits predictable behaviour when varying the level of noise added to the input data and we are able to explain the roles of all dimensions of our mapped space. Our models also allow fast predictions of input galaxies' position angles with a $1\sigma$ uncertainty range of $\pm17^{\circ}$ to $\pm23^{\circ}$ (for galaxies with inclinations of $82.5^{\circ}$ to $32.5^{\circ}$, respectively), which may be useful for initial parameterisation in kinematic modelling samplers. Machine learning models, such as the one outlined in this paper, may be adapted for SKA science usage in the near future., Comment: 15 pages, 11 figures

Identifying image features that are robust with respect to segmentation variability and domain shift is a tough challenge in radiomics. So far, this problem has mainly been tackled in test-retest analyses. In this work we analyze radiomics feature stability based on probabilistic segmentations. Based on a public lung cancer dataset, we generate an arbitrary number of plausible segmentations using a Probabilistic U-Net. From these segmentations, we extract a high number of plausible feature vectors for each lung tumor and analyze feature variance with respect to the segmentations. Our results suggest that there are groups of radiomic features that are more (e.g. statistics features) and less (e.g. gray-level size zone matrix features) robust against segmentation variability. Finally, we demonstrate that segmentation variance impacts the performance of a prognostic lung cancer survival model and propose a new and potentially more robust radiomics feature selection workflow., Comment: accepted at ISBI 2020

We present a method for highly efficient landmark detection that combines deep convolutional neural networks with well established model-based fitting algorithms. Motivated by established model-based fitting methods such as active shapes, we use a PCA of the landmark positions to allow generative modeling of facial landmarks. Instead of computing the model parameters using iterative optimization, the PCA is included in a deep neural network using a novel layer type. The network predicts model parameters in a single forward pass, thereby allowing facial landmark detection at several hundreds of frames per second. Our architecture allows direct end-to-end training of a model-based landmark detection method and shows that deep neural networks can be used to reliably predict model parameters directly without the need for an iterative optimization. The method is evaluated on different datasets for facial landmark detection and medical image segmentation. PyTorch code is freely available at https://github.com/justusschock/shapenet, Comment: https://github.com/justusschock/shapenet