Dataset Information

Classifying tissue samples from measurements on cells with within-class tissue sample heterogeneity.

ABSTRACT: We consider here the problem of classifying a macro-level object based on measurements of embedded (micro-level) observations within each object, for example, classifying a patient based on measurements on a collection of a random number of their cells. Classification problems with this hierarchical, nested structure have not received the same statistical understanding as the general classification problem. Some heuristic approaches have been developed and a few authors have proposed formal statistical models. We focus on the problem where heterogeneity exists between the macro-level objects within a class. We propose a model-based statistical methodology that models the log-odds of the macro-level object belonging to a class using a latent-class variable model to account for this heterogeneity. The latent classes are estimated by clustering the macro-level object density estimates. We apply this method to the detection of patients with cervical neoplasia based on quantitative cytology measurements on cells in a Papanicolaou smear. Quantitative cytology is much cheaper and potentially can take less time than the current standard of care. The results show that the automated quantitative cytology using the proposed method is roughly equivalent to clinical cytopathology and shows significant improvement over a statistical model that does not account for the heterogeneity of the data.

SUBMITTER: Yamal JM

PROVIDER: S-EPMC3169670 | biostudies-literature | 2011 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Classifying tissue samples from measurements on cells with within-class tissue sample heterogeneity.

Yamal Jose-Miguel JM Follen Michele M Guillaud Martial M Cox Dennis D DD

Biostatistics (Oxford, England) 20110603 4

We consider here the problem of classifying a macro-level object based on measurements of embedded (micro-level) observations within each object, for example, classifying a patient based on measurements on a collection of a random number of their cells. Classification problems with this hierarchical, nested structure have not received the same statistical understanding as the general classification problem. Some heuristic approaches have been developed and a few authors have proposed formal stat ...[more]

PMID: 21642388

Similar Datasets

Project description:Genomic tools have revealed genetically diverse pathogens within some hosts. Within-host pathogen diversity, which we refer to as "complex infection", is increasingly recognized as a determinant of treatment outcome for infections like tuberculosis. Complex infection arises through two mechanisms: within-host mutation (which results in clonal heterogeneity) and reinfection (which results in mixed infections). Estimates of the frequency of within-host mutation and reinfection in populations are critical for understanding the natural history of disease. These estimates influence projections of disease trends and effects of interventions. The genotyping technique MLVA (multiple loci variable-number tandem repeats analysis) can identify complex infections, but the current method to distinguish clonal heterogeneity from mixed infections is based on a rather simple rule. Here we describe ClassTR, a method which leverages MLVA information from isolates collected in a population to distinguish mixed infections from clonal heterogeneity. We formulate the resolution of complex infections into their constituent strains as an optimization problem, and show its NP-completeness. We solve it efficiently by using mixed integer linear programming and graph decomposition. Once the complex infections are resolved into their constituent strains, ClassTR probabilistically classifies isolates as clonally heterogeneous or mixed by using a model of tandem repeat evolution. We first compare ClassTR with the standard rule-based classification on 100 simulated datasets. ClassTR outperforms the standard method, improving classification accuracy from 48% to 80%. We then apply ClassTR to a sample of 436 strains collected from tuberculosis patients in a South African community, of which 92 had complex infections. We find that ClassTR assigns an alternate classification to 18 of the 92 complex infections, suggesting important differences in practice. By explicitly modeling tandem repeat evolution, ClassTR helps to improve our understanding of the mechanisms driving within-host diversity of pathogens like Mycobacterium tuberculosis.

Project description:BackgroundComputer-aided machine learning models are being actively developed with clinically available biomarkers to diagnose Alzheimer's disease (AD) in living persons. Despite considerable work with cross-sectional in vivo data, many models lack validation against postmortem AD neuropathological data.ObjectiveTrain machine learning models to classify the presence or absence of autopsy-confirmed severe AD neuropathology using clinically available features.MethodsAD neuropathological status are assessed at postmortem for participants from the National Alzheimer's Coordinating Center (NACC). Clinically available features are utilized, including demographics, Apolipoprotein E(APOE) genotype, and cortical thicknesses derived from ante-mortem MRI scans encompassing AD meta regions of interest (meta-ROI). Both logistic regression and random forest models are trained to identify linearly and nonlinearly separable features between participants with the presence (N = 91, age-at-MRI = 73.6±9.24, 38 women) or absence (N = 53, age-at-MRI = 68.93±19.69, 24 women) of severe AD neuropathology. The trained models are further validated in an external data set against in vivo amyloid biomarkers derived from PET imaging (amyloid-positive: N = 71, age-at-MRI = 74.17±6.37, 26 women; amyloid-negative: N = 73, age-at-MRI = 71.59±6.80, 41 women).ResultsOur models achieve a cross-validation accuracy of 84.03% in classifying the presence or absence of severe AD neuropathology, and an external-validation accuracy of 70.14% in classifying in vivo amyloid positivity status.ConclusionsOur models show that clinically accessible features, including APOE genotype and cortical thinning encompassing AD meta-ROIs, are able to classify both postmortem confirmed AD neuropathological status and in vivo amyloid status with reasonable accuracies. These results suggest the potential utility of AD meta-ROIs in determining AD neuropathological status in living persons.

Dataset Information

Classifying tissue samples from measurements on cells with within-class tissue sample heterogeneity.

Publications

Classifying tissue samples from measurements on cells with within-class tissue sample heterogeneity.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets