Automated colitis detection from endoscopic biopsies as a tissue screening tool in diagnostic pathology

McCann, Michael T.; Bhagavatula, Ramamurthy; Fickus, Matthew; Ozolek, John A.; Kovačević, Jelena

doi:10.1109/icip.2012.6467483

Cited by 10 publications

(9 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Of the remaining articles, none of them present an extensive comparison to other methods, their data sets are either non-public [35][36][37][38], no longer available [39], or the reference is not complete [40,41], which make them unsuitable for comparison. Jia et al [42] presented a segmentation method, and is therefore not comparable.…”

Section: Related Work and State-of-the-artmentioning

confidence: 99%

A Probabilistic Bag-to-Class Approach to Multiple-Instance Learning

Møllersen

Hardeberg

Godtliebsen

2020

Data

View full text Add to dashboard Cite

Multi-instance (MI) learning is a branch of machine learning, where each object (bag) consists of multiple feature vectors (instances)—for example, an image consisting of multiple patches and their corresponding feature vectors. In MI classification, each bag in the training set has a class label, but the instances are unlabeled. The instances are most commonly regarded as a set of points in a multi-dimensional space. Alternatively, instances are viewed as realizations of random vectors with corresponding probability distribution, where the bag is the distribution, not the realizations. By introducing the probability distribution space to bag-level classification problems, dissimilarities between probability distributions (divergences) can be applied. The bag-to-bag Kullback–Leibler information is asymptotically the best classifier, but the typical sparseness of MI training sets is an obstacle. We introduce bag-to-class divergence to MI learning, emphasizing the hierarchical nature of the random vectors that makes bags from the same class different. We propose two properties for bag-to-class divergences, and an additional property for sparse training sets, and propose a dissimilarity measure that fulfils them. Its performance is demonstrated on synthetic and real data. The probability distribution space is valid for MI learning, both for the theoretical analysis and applications.

show abstract

Section: Related Work and State-of-the-artmentioning

confidence: 99%

A Probabilistic Bag-to-Class Approach to Multiple-Instance Learning

Møllersen

Hardeberg

Godtliebsen

2020

Data

View full text Add to dashboard Cite

show abstract

“…1) Experimental Setup: As application-specific features we use the histopathology vocabulary (HV) [1], [4]. These features emulate the visual cues used by expert histopathologists [1], [2], [4], and are thus physiologically relevant. From the HV, we use nucleus size (1D), nucleus eccentricity (1D), nucleus density (1D), nucleus color (3D), red blood cell coverage (1D), and background color (3D).…”

Section: B Hande Data Setmentioning

confidence: 99%

Classification with reject option using contextual information

Condessa

Bioucas-Dias

Castro

et al. 2013

2013 IEEE 10th International Symposium on Biomedical Imaging

Self Cite

View full text Add to dashboard Cite

We propose a new algorithm for classification that merges classification with reject option with classification using contextual information. A reject option is desired in many image-classification applications requiring a robust classifier and when the need for high classification accuracy surpasses the need to classify the entire image. Moreover, our algorithm improves the classifier performance by including local and nonlocal contextual information, at the expense of rejecting a fraction of the samples. As a probabilistic model, we adopt a multinomial logistic regression. We use a discriminative random model for the description of the problem; we introduce reject option into the classification problem through association potential, and contextual information through interaction potential. We validate the method on the images of H&E-stained teratoma tissues and show the increase in the classifier performance when rejecting part of the assigned class labels.

show abstract

“…We refer to [15] for an introduction to frame theory and to [8] for an overview of the current research in the field. Frames have traditionally played a significant role in the theory of signal processing, but today they have found application to packet based network communication [7,18], wireless sensor networks [9,10,11,12], distributed processing [7], quantum information theory, bio-medical engineering [2,25], compressed sensing [3,14], fingerprinting [26], spectral theory [6,19,20], and much more.…”

Section: Introductionmentioning

confidence: 99%

Frame scalings: A condition number approach

Casazza

2017

Linear Algebra and its Applications

View full text Add to dashboard Cite

Abstract. Scaling frame vectors is a simple and noninvasive way to construct tight frames. However, not all frames can be modifed to tight frames in this fashion, so in this case we explore the problem of finding the best conditioned frame by scaling, which is crucial for applications like signal processing. We conclude that this problem is equivalent to solving a convex optimization problem involving the operator norm, which is unconventional since this problem was only studied in the perspective of Frobenius norm before. We also further study the Frobenius norm case in relation to the condition number of the frame operator, and the convexity of optimal scalings.

show abstract

Automated colitis detection from endoscopic biopsies as a tissue screening tool in diagnostic pathology

Cited by 10 publications

References 11 publications

A Probabilistic Bag-to-Class Approach to Multiple-Instance Learning

A Probabilistic Bag-to-Class Approach to Multiple-Instance Learning

Classification with reject option using contextual information

Frame scalings: A condition number approach

Contact Info

Product

Resources

About