Dataset Information

Model-based clustering for assessing the prognostic value of imaging biomarkers and mixed type tests.

ABSTRACT: A model-based clustering method is proposed to address two research aims in Alzheimer's disease (AD): to evaluate the accuracy of imaging biomarkers in AD prognosis, and to integrate biomarker information and standard clinical test results into the diagnoses. One challenge in such biomarker studies is that it is often desired or necessary to conduct the evaluation without relying on clinical diagnoses or some other standard references. This is because (1) biomarkers may provide prognostic information long before any standard reference can be acquired; (2) these references are often based on or provide unfair advantage to standard tests. Therefore, they can mask the prognostic value of a useful biomarker, especially when the biomarker is much more accurate than the standard tests. In addition, the biomarkers and existing tests may be of mixed type and vastly different distributions. A model-based clustering method based on finite mixture modeling framework is introduced. The model allows for the inclusion of mixed typed manifest variables with possible differential covariates to evaluate the prognostic value of biomarkers in addition to standard tests without relying on potentially inaccurate reference diagnoses. Maximum likelihood parameter estimation is carried out via the EM algorithm. Accuracy measures and the ROC curves of the biomarkers are derived subsequently. Finally, the method is illustrated with a real example in AD.

SUBMITTER: Wang Z

PROVIDER: S-EPMC5613685 | biostudies-literature | 2017 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Model-based clustering for assessing the prognostic value of imaging biomarkers and mixed type tests.

Wang Zheyu Z Sebestyen Krisztian K Monsell Sarah E SE

Computational statistics & data analysis 20161102

A model-based clustering method is proposed to address two research aims in Alzheimer's disease (AD): to evaluate the accuracy of imaging biomarkers in AD prognosis, and to integrate biomarker information and standard clinical test results into the diagnoses. One challenge in such biomarker studies is that it is often desired or necessary to conduct the evaluation without relying on clinical diagnoses or some other standard references. This is because (1) biomarkers may provide prognostic inform ...[more]

PMID: 28966420

Similar Datasets

Project description:BackgroundAmide proton transfer (APT) imaging has been gradually applied to cervical cancer, yet the relationships between APT and multiple model diffusion-weighted imaging (DWI) have yet to be investigated. This study attempted to evaluate the added value of 3-dimensional (3D) APT imaging to multiple model DWI for assessing prognostic factors of cervical cancer.MethodsThis prospective diagnostic study was conducted in The First Affiliated Hospital of Zhengzhou University. A total of 88 consecutive patients with cervical cancer underwent APT imaging and DWI with 11 b-values (0-2,000 s/mm2). The apparent diffusion coefficient (ADC), pure molecular diffusion (D), perfusion fraction (f), pseudo-diffusion (D*), mean kurtosis (MK), and mean diffusivity (MD) were calculated based on mono-exponential, bi-exponential, and kurtosis models. The mean, minimum, and maximum values of APT signal intensity (APT SI) and DWI-derived metrics were compared based on tumor stages, subtypes, grades, and lymphovascular space invasion status by Student's t-test or Mann-Whitney U test. Receiver operating characteristic (ROC) curve analysis was used to evaluate the diagnostic performance of the parameters.ResultsAPT SImax, APT SImin, MKmean, and MKmax showed significant differences between adenocarcinoma (AC) and squamous cell carcinoma (SCC) (all P<0.05). APT SImean, APT SImax, and MKmax were higher and ADCmin, Dmean, Dmin, and MDmin were lower in the high-grade tumor than in low-grade tumor (all P<0.05). For distinguishing lymphovascular space invasion, only MKmean showed significant difference (P=0.010). APT SImax [odds ratio (OR) =2.347, P=0.029], APT SImin (OR =0.352; P=0.024), and MKmean (OR =6.523; P=0.001) were the independent predictors for tumor subtype, and APT SImax (OR =2.885; P=0.044), MDmin (OR =0.155, P=0.012) were the independent predictors for histological grade of cervical cancer. When APT SImin and APT SImax was combined with MKmean and MKmax, the diagnostic performance was significantly improved for differentiating AC and AC [area under the curve (AUC): 0.908, sensitivity: 87.5%; specificity: 83.3%; P<0.001]. The combination of APT SImean, APT SImax, ADCmin, MKmax, and MDmin demonstrated the highest diagnostic performance for predicting tumor grade (AUC: 0.903, sensitivity: 78.6%; specificity: 88.9%; P<0.001).ConclusionsAddition of APT to DWI may improve the ability to noninvasively predict poor prognostic factors of cervical cancer.

Project description:Analysis of data measured on different scales is a relevant challenge. Biomedical studies often focus on high-throughput datasets of, e.g., quantitative measurements. However, the need for integration of other features possibly measured on different scales, e.g. clinical or cytogenetic factors, becomes increasingly important. The analysis results (e.g. a selection of relevant genes) are then visualized, while adding further information, like clinical factors, on top. However, a more integrative approach is desirable, where all available data are analyzed jointly, and where also in the visualization different data sources are combined in a more natural way. Here we specifically target integrative visualization and present a heatmap-style graphic display. To this end, we develop and explore methods for clustering mixed-type data, with special focus on clustering variables. Clustering of variables does not receive as much attention in the literature as does clustering of samples. We extend the variables clustering methodology by two new approaches, one based on the combination of different association measures and the other on distance correlation. With simulation studies we evaluate and compare different clustering strategies. Applying specific methods for mixed-type data proves to be comparable and in many cases beneficial as compared to standard approaches applied to corresponding quantitative or binarized data. Our two novel approaches for mixed-type variables show similar or better performance than the existing methods ClustOfVar and bias-corrected mutual information. Further, in contrast to ClustOfVar, our methods provide dissimilarity matrices, which is an advantage, especially for the purpose of visualization. Real data examples aim to give an impression of various kinds of potential applications for the integrative heatmap and other graphical displays based on dissimilarity matrices. We demonstrate that the presented integrative heatmap provides more information than common data displays about the relationship among variables and samples. The described clustering and visualization methods are implemented in our R package CluMix available from https://cran.r-project.org/web/packages/CluMix.

Project description:BackgroundDisease prevalence is rarely explicitly considered in the early stages of the development of novel prognostic tests. Rather, researchers use the area under the receiver operating characteristic (AUROC) as the key metric to gauge and report predictive performance ability. Because this statistic does not account for disease prevalence, proposed tests may not appropriately address clinical requirements. This ultimately impedes the translation of prognostic tests into clinical practice.MethodsA method to express positive- and/or negative predictive value criteria (PPV, NPV) within the ROC space is presented. Equations are derived for so-called equi-PPV (and equi-NPV) lines. Herewith it is possible, for any given prevalence, to plot a series of sensitivity-specificity pairs which meet a specified PPV (or NPV) criterion onto the ROC space.This concept is introduced by firstly reviewing the well-established "mechanics", strengths and limitations of the ROC analysis in the context of developing prognostic models. Then, the use of PPV (and/or) NPV criteria to augment the ROC analysis is elaborated.Additionally, an interactive web tool was also created to enable people to explore the dynamics of lines of equi-predictive value in function of prevalence. The web tool also allows to gauge what ROC curve shapes best meet specific positive and/or negative predictive value criteria (http://d4ta.link/ppvnpv/).ResultsTo illustrate the merits and implications of this concept, an example on the prediction of pre-eclampsia risk in low-risk nulliparous pregnancies is elaborated.ConclusionsIn risk stratification, the clinical usefulness of a prognostic test can be expressed in positive- and negative predictive value criteria; the development of novel prognostic tests will be facilitated by the possibility to co-visualise such criteria together with ROC curves. To achieve clinically meaningful risk stratification, the development of separate tests to meet either a pre-specified positive value (rule-in) or a negative predictive value (rule-out) criteria should be considered: the characteristics of successful rule-in and rule-out tests may markedly differ.

Project description:Background Reflectance confocal microscopy (RCM), VISIA, and dermoscopy have emerged as promising tools for objective diagnosis and assessment of rosacea. However, little is known about the diagnostic value of these imaging systems for rosacea. Objectives To assess the diagnostic value of RCM, VISIA, and dermoscopy for rosacea by establishing a novel multilayer perceptron (MLP) model. Methods A total of 520 patients with rosacea and other facial diseases were included in this study. A total of 474 samples of dermoscopy data, 374 samples of RCM data, 434 samples of VISIA data, and 291 samples containing three data sources were collected. An MLP model was built with the total data to explore the association between the imageological features of each instrument and the probability of rosacea. Results Our MLP model revealed that the area under the receiver operating characteristic curve (AUROC) values of RCM, VISIA and dermoscopy for diagnosing rosacea were 0.5233, 0.5646 and 0.7971, respectively. The integration of these three tools with clinical data could further improve the accuracy of the predictive diagnosis to 0.8385. For the imageological features of each tool, abnormalities (hyperkeratosis or parakeratosis) in the stratum corneum were effective variables for excluding rosacea (odds ratio [OR], 0.4333) under RCM. The indicators of rosacea under VISIA included overall severity of erythema, erythema involving the cheek or superciliary arch, visible red blood vessels, and papules (OR = 2.2745, 3.1592, 1.8365, 2.8647, and 1.4260, respectively). The candidate variables of dermoscopy included yellow background, white background, uniform distribution of vessels, branched vessels, and reticular blood vessels (OR = 0.4259, 0.4949, 2.2858, 3.7444, and 2.4576, respectively). Conclusions RCM, dermoscopy, and VISIA each can present several imageological features and were of certain value for assisting rosacea diagnosis. The combined analysis of these three tools using our MLP model may be useful for improving the accuracy of diagnosing rosacea.

Dataset Information

Model-based clustering for assessing the prognostic value of imaging biomarkers and mixed type tests.

Publications

Model-based clustering for assessing the prognostic value of imaging biomarkers and mixed type tests.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets