Patch-Based Sparse Representation For Bacterial Detection

Abstract

In this paper, we propose an unsupervised approach for bacterial detection in optical endomicroscopy images. This approach splits each image into a set of overlapping patches and assumes that observed intensities are linear combinations of the actual intensity values associated with background image structures, corrupted by additive Gaussian noise and potentially by a sparse outlier term modelling anomalies (which are considered to be candidate bacteria). The actual intensity term representing background structures is modelled as a linear combination of a few atoms drawn from a dictionary which is learned from bacteria-free data and then fixed while analyzing new images. The bacteria detection task is formulated as a minimization problem and an alternating direction method of multipliers (ADMM) is then used to estimate the unknown parameters. Simulations conducted using two ex vivo lung datasets show good detection and correlation performance between bacteria counts identified by a trained clinician and those of the proposed method.

EmiNet: Annotation Free Moving Bacteria Detection on Optical Endomicroscopy Images

“…Due to the limited availability of annotated images, many approaches have resorted to unsupervised learning methods for bacteria detection. For instance, in [19] and [20], algorithms were proposed to detect bacteria in fluorescence intensity images using the OEM imaging system. These algorithms treated input images as a combination of actual intensity values related to background structures, contaminated by additive Gaussian noise, and possibly some sparse anomalies representing potential bacteria.…”

“…These algorithms treated input images as a combination of actual intensity values related to background structures, contaminated by additive Gaussian noise, and possibly some sparse anomalies representing potential bacteria. In [19], the bacteria detection problem was framed as a minimization problem, and an alternating direction method of multipliers (ADMM) was employed to estimate the unknown parameters. On the other hand, [20] utilized Hierarchical Bayesian models (HBM) to detect bacteria in the modeled images, incorporating Markov Random Fields (MRF) to describe the background intensity.…”

Applications of Machine Learning in time-domain Fluorescence Lifetime Imaging: a Review

“…In contrast, unsupervised methods, like the Bayesian approach for detecting microorganisms in pulmonary OEM imaging, treat bacterial objects as outliers added to actual pixel values disturbed by Gaussian noise [ 157 ]. Another approach divided images into overlapping patches, transforming bacteria detection into a minimization problem [ 158 ].…”

Bayesian Activity Estimation and Uncertainty Quantification of Spent Nuclear Fuel Using Passive Gamma Emission Tomography

“…Moreover, ρ, α > 0 are user defined parameters. In Equation (13), it is possible to use a block Gibbs sampler( with proximal MCMC step in the Poisson noise case) to sample according to the conditional distributions of each of the unknown model parameters. In practice, strong correlations appear between z and t. Moreover, as z can be sparse, sampling f (t, s 2 |y, ∆ \t , Φ \(s 2 ) ), where H \u denotes the parameter vector H whose parameter u is omitted, via Gibbs sampling results in very slow convergence.…”

“…Most of state-of-the-art algorithms considered to solve linear inverse image restoration problems (irrespective of the noise model) are either optimization or simulation-based methods. Optimization-based approaches primarily rely on log-concave Bayesian models, such as [8][9][10][11][12][13][14][15], and have been proposed to perform maximum-a-posteriori (MAP) estimation. For example, PIDAL, which stands for Poisson image deblurring using augmented Lagrangian [8], and SALSA, which stands for split augmented Lagrangian shrinkage algorithm [10], are Poisson and Gaussian image restoration algorithms based on a totalvariation loss ot sparsity-promoting prior, which solves the restoration problem using an alternating direction method of multipliers (ADMM).…”