Dataset Information

A Novel Cascade Classifier for Automatic Microcalcification Detection.

ABSTRACT: In this paper, we present a novel cascaded classification framework for automatic detection of individual and clusters of microcalcifications (?C). Our framework comprises three classification stages: i) a random forest (RF) classifier for simple features capturing the second order local structure of individual ?Cs, where non-?C pixels in the target mammogram are efficiently eliminated; ii) a more complex discriminative restricted Boltzmann machine (DRBM) classifier for ?C candidates determined in the RF stage, which automatically learns the detailed morphology of ?C appearances for improved discriminative power; and iii) a detector to detect clusters of ?Cs from the individual ?C detection results, using two different criteria. From the two-stage RF-DRBM classifier, we are able to distinguish ?Cs using explicitly computed features, as well as learn implicit features that are able to further discriminate between confusing cases. Experimental evaluation is conducted on the original Mammographic Image Analysis Society (MIAS) and mini-MIAS databases, as well as our own Seoul National University Bundang Hospital digital mammographic database. It is shown that the proposed method outperforms comparable methods in terms of receiver operating characteristic (ROC) and precision-recall curves for detection of individual ?Cs and free-response receiver operating characteristic (FROC) curve for detection of clustered ?Cs.

SUBMITTER: Shin SY

PROVIDER: S-EPMC4668028 | biostudies-literature | 2015

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A Novel Cascade Classifier for Automatic Microcalcification Detection.

Shin Seung Yeon SY Lee Soochahn S Yun Il Dong ID Jung Ho Yub HY Heo Yong Seok YS Kim Sun Mi SM Lee Kyoung Mu KM

PloS one 20151202 12

In this paper, we present a novel cascaded classification framework for automatic detection of individual and clusters of microcalcifications (μC). Our framework comprises three classification stages: i) a random forest (RF) classifier for simple features capturing the second order local structure of individual μCs, where non-μC pixels in the target mammogram are efficiently eliminated; ii) a more complex discriminative restricted Boltzmann machine (DRBM) classifier for μC candidates determined ...[more]

PMID: 26630496

Similar Datasets

Project description:BackgroundMounting evidence suggests several diseases and biological processes target transcription termination to misregulate gene expression. Disruption of transcription termination leads to readthrough transcription past the 3' end of genes, which can result in novel transcripts, changes in epigenetic states and altered 3D genome structure.ResultsWe developed Automatic Readthrough Transcription Detection (ARTDeco), a tool to detect and analyze multiple features of readthrough transcription from RNA-seq and other next-generation sequencing (NGS) assays that profile transcriptional activity. ARTDeco robustly quantifies the global severity of readthrough phenotypes, and reliably identifies individual genes that fail to terminate (readthrough genes), are aberrantly transcribed due to upstream termination failure (read-in genes), and novel transcripts created as a result of readthrough (downstream of gene or DoG transcripts). We used ARTDeco to characterize readthrough transcription observed during influenza A virus (IAV) infection, validating its specificity and sensitivity by comparing its performance in samples infected with a mutant virus that fails to block transcription termination. We verify ARTDeco's ability to detect readthrough as well as identify read-in genes from different experimental assays across multiple experimental systems with known defects in transcriptional termination, and show how these results can be leveraged to improve the interpretation of gene expression and downstream analysis. Applying ARTDeco to a gene expression data set from IAV-infected monocytes from different donors, we find strong evidence that read-in gene-associated expression quantitative trait loci (eQTLs) likely regulate genes upstream of read-in genes. This indicates that taking readthrough transcription into account is important for the interpretation of eQTLs in systems where transcription termination is blocked.ConclusionsARTDeco aids researchers investigating readthrough transcription in a variety of systems and contexts.

Dataset Information

A Novel Cascade Classifier for Automatic Microcalcification Detection.

Publications

A Novel Cascade Classifier for Automatic Microcalcification Detection.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets