Persistent animal identification leveraging non-visual markers.

Our objective is to locate and provide a unique identifier for each mouse in a cluttered home-cage environment through time, as a precursor to automated behaviour recognition for biological research. This is a very challenging problem due to (i) the lack of distinguishing visual features for each mouse, and (ii) the close confines of the scene with constant occlusion, making standard visual tracking approaches unusable. However, a coarse estimate of each mouse's location is available from a unique RFID implant, so there is the potential to optimally combine information from (weak) tracking with coarse information on identity. To achieve our objective, we make the following key contributions: (a) the formulation of the object identification problem as an assignment problem (solved using Integer Linear Programming), (b) a novel probabilistic model of the affinity between tracklets and RFID data, and (c) a curated dataset with per-frame BB and regularly spaced ground-truth annotations for evaluating the models. The latter is a crucial part of the model, as it provides a principled probabilistic treatment of object detections given coarse localisation. Our approach achieves 77% accuracy on this animal identification problem, and is able to reject spurious detections when the animals are hidden.

Inference and Learning for Generative Capsule Models.

Capsule networks (see Hinton et al., 2018) aim to encode knowledge of and reason about the relationship between an object and its parts. In this letter, we specify a generative model for such data and derive a variational algorithm for inferring the transformation of each model object in a scene and the assignments of observed parts to the objects. We derive a learning algorithm for the object models, based on variational expectation maximization (Jordan et al., 1999). We also study an alternative inference algorithm based on the RANSAC method of Fischler and Bolles (1981). We apply these inference methods to data generated from multiple geometric objects like squares and triangles ("constellations") and data from a parts-based model of faces. Recent work by Kosiorek et al. (2019) has used amortized inference via stacked capsule autoencoders to tackle this problem; our results show that we significantly outperform them where we can make comparisons (on the constellations data).

On Suspicious Coincidences and Pointwise Mutual Information.

Barlow (1985) hypothesized that the co-occurrence of two events A and B is "suspicious" if P(A,B)≫P(A)P(B). We first review classical measures of association for 2 × 2 contingency tables, including Yule's Y (Yule, 1912), which depends only on the odds ratio λ and is independent of the marginal probabilities of the table. We then discuss the mutual information (MI) and pointwise mutual information (PMI), which depend on the ratio P(A,B)/P(A)P(B), as measures of association. We show that once the effect of the marginals is removed, MI and PMI behave similarly to Y as functions of λ. The pointwise mutual information is used extensively in some research communities for flagging suspicious coincidences. We discuss the pros and cons of using it in this way, bearing in mind the sensitivity of the PMI to the marginals, with increased scores for sparser events.

The Effect of Class Imbalance on Precision-Recall Curves.

In this note, I study how the precision of a binary classifier depends on the ratio r of positive to negative cases in the test set, as well as the classifier's true and false-positive rates. This relationship allows prediction of how the precision-recall curve will change with r, which seems not to be well known. It also allows prediction of how Fß and the precision gain and recall gain measures of Flach and Kull (2015) vary with r.

Assessing the utility of autofluorescence-based pulmonary optical endomicroscopy to predict the malignant potential of solitary pulmonary nodules in humans.

Solitary pulmonary nodules are common, often incidental findings on chest CT scans. The investigation of pulmonary nodules is time-consuming and often leads to protracted follow-up with ongoing radiological surveillance, however, clinical calculators that assess the risk of the nodule being malignant exist to help in the stratification of patients. Furthermore recent advances in interventional pulmonology include the ability to both navigate to nodules and also to perform autofluorescence endomicroscopy. In this study we assessed the efficacy of incorporating additional information from label-free fibre-based optical endomicrosopy of the nodule on assessing risk of malignancy. Using image analysis and machine learning approaches, we find that this information does not yield any gain in predictive performance in a cohort of patients. Further advances with pulmonary endomicroscopy will require the addition of molecular tracers to improve information from this procedure.

Artefact in Physiological Data Collected from Patients with Brain Injury: Quantifying the Problem and Providing a Solution Using a Factorial Switching Linear Dynamical Systems Approach.

INTRODUCTION: High-resolution, artefact-free and accurately annotated physiological data are desirable in patients with brain injury both to inform clinical decision-making and for intelligent analysis of the data in applications such as predictive modelling. We have quantified the quality of annotation surrounding artefactual events and propose a factorial switching linear dynamical systems (FSLDS) approach to automatically detect artefact in physiological data collected in the neurological intensive care unit (NICU). METHODS: Retrospective analysis of the BrainIT data set to discover potential hypotensive events corrupted by artefact and identify the annotation of associated clinical interventions. Training of an FSLDS model on clinician-annotated artefactual events in five patients with severe traumatic brain injury. RESULTS: In a subset of 187 patients in the BrainIT database, 26.5 % of potential hypotensive events were abandoned because of artefactual data. Only 30 % of these episodes could be attributed to an annotated clinical intervention. As assessed by the area under the receiver operating characteristic curve metric, FSLDS model performance in automatically identifying the events of blood sampling, arterial line damping and patient handling was 0.978, 0.987 and 0.765, respectively. DISCUSSION: The influence of artefact on physiological data collected in the NICU is a significant problem. This pilot study using an FSLDS approach shows real promise and is under further development.

Localisation microscopy with quantum dots using non-negative matrix factorisation.

We propose non-negative matrix factorisation with iterative restarts (iNMF) to model a noisy dataset of highly overlapping fluorophores with intermittent intensities. We can recover high-resolution images of individual sources from the optimised model, despite their high mutual overlap in the original data. Each source can have an arbitrary, unknown shape of the PSF and blinking behaviour. This allows us to use quantum dots as bright and stable fluorophores for localisation microscopy. We compare the iNMF results to CSSTORM, 3B and bSOFI. iNMF shows superior performance in the challenging task of super-resolution imaging using quantum dots. We can also retrieve axial localisation of the sources from the shape of the recovered PSF.

Autoregressive hidden Markov models for the early detection of neonatal sepsis.

Late onset neonatal sepsis is one of the major clinical concerns when premature babies receive intensive care. Current practice relies on slow laboratory testing of blood cultures for diagnosis. A valuable research question is whether sepsis can be reliably detected before the blood sample is taken. This paper investigates the extent to which physiological events observed in the patient's monitoring traces could be used for the early detection of neonatal sepsis. We model the distribution of these events with an autoregressive hidden Markov model (AR-HMM). Both learning and inference carefully use domain knowledge to extract the baby's true physiology from the monitoring data. Our model can produce real-time predictions about the onset of the infection and also handles missing data. We evaluate the effectiveness of the AR-HMM for sepsis detection on a dataset collected from the Neonatal Intensive Care Unit at the Royal Infirmary of Edinburgh.

Greedy learning of binary latent trees.

Inferring latent structures from observations helps to model and possibly also understand underlying data generating processes. A rich class of latent structures is the latent trees, i.e., tree-structured distributions involving latent variables where the visible variables are leaves. These are also called hierarchical latent class (HLC) models. Zhang and Kocka proposed a search algorithm for learning such models in the spirit of Bayesian network structure learning. While such an approach can find good solutions, it can be computationally expensive. As an alternative, we investigate two greedy procedures: the BIN-G algorithm determines both the structure of the tree and the cardinality of the latent variables in a bottom-up fashion. The BIN-A algorithm first determines the tree structure using agglomerative hierarchical clustering, and then determines the cardinality of the latent variables as for BIN-G. We show that even with restricting ourselves to binary trees, we obtain HLC models of comparable quality to Zhang's solutions (in terms of cross-validated log-likelihood), while being generally faster to compute. This claim is validated by a comprehensive comparison on several data sets. Furthermore, we demonstrate that our methods are able to estimate interpretable latent structures on real-world data with a large number of variables. By applying our method to a restricted version of the 20 newsgroups data, these models turn out to be related to topic models, and on data from the PASCAL Visual Object Classes (VOC) 2007 challenge, we show how such treestructured models help us understand how objects co-occur in images. For reproducibility of all experiments in this paper, all code and data sets (or links to data) are available at http://people.kyb.tuebingen.mpg.de/harmeling/code/ltt-1.4.tar.

Factorial switching linear dynamical systems applied to physiological condition monitoring.

Condition monitoring often involves the analysis of systems with hidden factors that switch between different modes of operation in some way. Given a sequence of observations, the task is to infer the filtering distribution of the switch setting at each time step. In this paper, we present factorial switching linear dynamical systems as a general framework for handling such problems. We show how domain knowledge and learning can be successfully combined in this framework, and introduce a new factor (the "X-factor") for dealing with unmodeled variation. We demonstrate the flexibility of this type of model by applying it to the problem of monitoring the condition of a premature baby receiving intensive care. The state of health of a baby cannot be observed directly, but different underlying factors are associated with particular patterns of physiological measurements and artifacts. We have explicit knowledge of common factors and use the X-factor to model novel patterns which are clinically significant but have unknown cause. Experimental results are given which show the developed methods to be effective on typical intensive care unit monitoring data.

A regularized discriminative model for the prediction of protein-peptide interactions.

MOTIVATION: Short well-defined domains known as peptide recognition modules (PRMs) regulate many important protein-protein interactions involved in the formation of macromolecular complexes and biochemical pathways. Since high-throughput experiments like yeast two-hybrid and phage display are expensive and intrinsically noisy, it would be desirable to more specifically target or partially bypass them with complementary in silico approaches. In the present paper, we present a probabilistic discriminative approach to predicting PRM-mediated protein-protein interactions from sequence data. The model is motivated by the discriminative model of Segal and Sharan as an alternative to the generative approach of Reiss and Schwikowski. In our evaluation, we focus on predicting the interaction network. As proposed by Williams, we overcome the problem of susceptibility to over-fitting by adopting a Bayesian a posteriori approach based on a Laplacian prior in parameter space. RESULTS: The proposed method was tested on two datasets of protein-protein interactions involving 28 SH3 domain proteins in Saccharmomyces cerevisiae, where the datasets were obtained with different experimental techniques. The predictions were evaluated with out-of-sample receiver operator characteristic (ROC) curves. In both cases, Laplacian regularization turned out to be crucial for achieving a reasonable generalization performance. The Laplacian-regularized discriminative model outperformed the generative model of Reiss and Schwikowski in terms of the area under the ROC curve on both datasets. The performance was further improved with a hybrid approach, in which our model was initialized with the motifs obtained with the method of Reiss and Schwikowski. AVAILABILITY: Software and supplementary material is available from http://lehrach.com/wolfgang/dmf.

The impact of using related individuals for haplotype reconstruction in population studies.

Recent studies have highlighted the dangers of using haplotypes reconstructed directly from population data for a fine-scale mapping analysis. Family data may help resolve ambiguity, yet can be costly to obtain. This study is concerned with the following question: How much family data (if any) should be used to facilitate haplotype reconstruction in a population study? We conduct a simulation study to evaluate how changes in family information can affect the accuracy of haplotype frequency estimates and phase reconstruction. To reconstruct haplotypes, we introduce an EM-based algorithm that can efficiently accommodate unrelated individuals, parent-child trios, and arbitrarily large half-sib pedigrees. Simulations are conducted for a diverse set of haplotype frequency distributions, all of which have been previously published in empirical studies. A wide variety of important results regarding the effectiveness of using pedigree data in a population study are presented in a coherent, unified framework. Insight into the different properties of the haplotype frequency distribution that can influence experimental design is provided. We show that a preliminary estimate of the haplotype frequency distribution can be valuable in large population studies with fixed resources.

How to pretend that correlated variables are independent by using difference observations.

In many areas of data modeling, observations at different locations (e.g.,time frames or pixel locations) are augmented by differences of nea r by observations (e.g., delta features in speech recognition, Gabor jets in image analysis). These augmented observations are then often modeled as being independent. How can this make sense?We provide two interpretations,showing (1) that the likelihood of data generated from an auto regressive process can be computed in terms of "independent" augmented observations and (2) that the augmented observations can be given a coherent treatment in terms of the products of experts model (Hinton, 1999).

Greedy learning of multiple objects in images using robust statistics and factorial learning.

We consider data that are images containing views of multiple objects. Our task is to learn about each of the objects present in the images. This task can be approached as a factorial learning problem, where each image must be explained by instantiating a model for each of the objects present with the correct instantiation parameters. A major problem with learning a factorial model is that as the number of objects increases, there is a combinatorial explosion of the number of configurations that need to be considered. We develop a method to extract object models sequentially from the data by making use of a robust statistical method, thus avoiding the combinatorial explosion, and present results showing successful extraction of objects from real images.