PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Model interpretability”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 343 records · Page 19Linked to original sources
PubMed1985

The use of a logistic model for the quantitative interpretation of indirect sandwich enzyme labelled immunosorbent assays (ELISA) for antibodies and antigens in foot and mouth disease.

A three parameter logistic model is described for the analysis of profiles of optical density vs log dose for indirect sandwich ELISA tests in foot and mouth disease. The model describes the observed phenomenon of saturability with increasing dose, and its parameters can be interpreted in terms of molecular binding events. A computer program to fit the model is described. An approximate statistical test is developed which can be used to test for departures from equivalence for replicate profiles. It is found that correlation of the optical density values to a standard reference reaction considerably improves reproducibility.

Animals↗
PubMed1995

A linear modelling approach to automatic interpretation of quality control measurements in mammography.

An approach to the automated interpretation of quality control data is described in which the relationship between small changes in system performance and the resulting small changes in the measured quality control test results are approximated as a set of linear equations in matrix form. The inverse of this forward matrix is then used to identify causal changes from the set of measured changes. This approach has been investigated for the case of mammographic screening quality assurance and shown to provide a potentially useful tool to assist in the interpretation task.

Autoanalysis↗
PubMed1997

Fourier models and the loci of adaptation.

First measures of sensitivity and the need for a model to interpret them are addressed. Then modeling in the Fourier domain is promoted by a demonstration of how much an approach explains spatial sensitization and its dependence on luminance. Then the retinal illuminance and receptor absorptions produced by various stimuli are derived to foster interpretation of the neural mechanisms underlying various psychophysical phenomena. Finally, the sequence and the anatomical loci of the processes controlling visual sensitivity are addressed. It is concluded that multiplicative adaptation often has effects identical to response compression followed by subtractive adaptation and that, perhaps as a consequence, there is no evidence of retinal gain changes in human cone vision until light levels are well above those available in natural scenes and in most contemporary psychophysical experiments; that contrast gain control fine tunes sensitivity to patterns at all luminances; and that response compression, modulated by subtractive adaptation, predominates in the control of sensitivity in human cone vision.

Adaptation, Physiological↗
PubMed1999

Electrostatic Contribution to the Enthalpy of Charging at Hematite/Electrolyte Interface.

A calorimetry study of hematite surface reactions is reported, with special emphasis on the estimation of the electrostatic contribution to the enthalpy of charging. The calorimetry titrations were performed outside the point of zero charge region where the electrostatic contribution is significant. The results were interpreted by the surface complexation model. The interpretation enabled the evaluation of standard protonation and deprotonation enthalpies as well as the electrostatic contribution to these quantities. Copyright 1999 Academic Press.

Journal Article↗
PubMed1994

The GALEN project.

The GALEN project is developing language independent concept representation systems as the foundations for the next generation of multilingual coding systems. It aims to support the flexibility required to cope with the diversity amongst medical applications, while ensuring the coherence necessary for integration and re-use of terminologies. GALEN is developing a fully compositional and generative formal system for modelling concepts: the GALEN Representation and Integration Language (GRAIL) Kernel. Its goal is to overcome many of the problems with traditional coding and classification systems, in particular the combinatorial explosion of terms in enumerative systems and the generation of nonsensical terms in partially compositional systems. It will also provide a clean separation between the concept model and linguistic mechanisms which interpret that model (i.e., the words in a specific language, syntax, alternative phrasings, etc.) in order to allow the development of multilingual systems. GRAIL aims to be formally sound and produce models that are verifiable and contain no contradictions or ambiguities, with realistic human effort. A Coding Reference (CORE) Model of medical terminology covering is being developed which aims to represent the core concepts in for example pathology, anatomy and therapeutics, that have widespread applicability in medical applications. It should also provide the basis for specialist extensions according to the formal principles of GRAIL. The main results of GALEN will be delivered as a Terminology Server (TeS) which encapsulates and coordinates the functionality of the concept module, multilingual module, and code conversion module, and also provides a uniform applications programming interface and network services for use by external applications.

Databases, Bibliographic↗
PubMed1998

Incorporation of a set enumeration trees-based classifier into a hybrid computer-assisted diagnosis scheme for mass detection.

RATIONALE AND OBJECTIVES: The authors evaluated whether a hybrid classifier of two independent computer-aided diagnosis (CAD) schemes, the set enumeration (SE) trees approach and an artificial neural network (ANN), could improve the detection of masses on digitized mammograms. The potential benefits resulting from the interpretability of the SE trees model was also explored. MATERIALS AND METHODS: Two hundred thirty verified mass regions and 230 negative but suspicious regions were randomly selected from 618 digitized mammograms. Each region was represented by a 24-parameter feature vector. These features were used as input data for the SE trees and ANN-based schemes. After the positive and negative regions were randomly segmented into five exclusive partitions, a fivefold cross-validation method was applied to evaluate and compare the performance of the SE trees, ANN, and hybrid system in the identification of masses. RESULTS: The performance of the SE trees approach was comparable to that of the ANN. The average area under the receiver operating characteristic (ROC) curves for all five partitions was 0.88 (standard deviation, 0.04). Owing to the relatively low correlation between the region-based results of the SE trees and ANN methods, the hybrid classifier yielded a significantly improved performance, with an area under the ROC curve of 0.94 (standard deviation, 0.02; P < .05). CONCLUSION: The hybrid CAD scheme significantly improved performance. The amenability of the SE trees models to interpretation may aid in the assessment of the importance of specific features.

Artificial Intelligence↗
PubMed1998

Model selection for covariance structures analysis in nursing research.

Covariance structures analysis is often used in nursing research to appraise statistical models reflecting complex human health processes. The model selection approach in covariance structures analysis is designed to select the "best" model from a specified set of theoretically defensible, competing alternatives, all of which are viewed as approximations. Model selection criteria explicitly incorporate both model misfit in the population and sampling error to evaluate the set of models. The result is that interpretability of model parameters and goodness-of-fit are enhanced simultaneously. Relative merits of the model selection approach are identified in light of technical concerns, parsimony, and use of scientific theory in nursing.

Analysis of Variance↗
PubMed2000

Light scattering and absorption model for the quantitative interpretation of human blood platelet spectral data.

Multiwavelength ultraviolet-visible (UV-Vis) transmission spectroscopy is a relatively simple technique that can provide considerable quantitative information on the properties of micron and submicron particle suspensions. Two important particle properties are particle size distribution (PSD) and chemical composition. These properties provide characteristics for the identification and classification of biological systems ranging in size and composition from proteins and nucleic acids to cells. By measuring the complete UV-Vis spectrum, the combined scattering and absorption properties are obtained as a function of wavelength. The quantitative evaluation of the size distribution and chemical composition is accomplished through the application of light-scattering theory. This paper reports on the estimation of the optical properties of human blood platelets and their use in the interpretation of platelet UV-Vis spectra within the context of Mie theory. The model developed herein provides reliable and accurate estimates for the PSD and particle number of platelet suspensions. One potential application of this characterization method is in the analysis of platelet activation by thrombin. Quantification of spectral data with respect to average particle size and particle number provides a real-time description of the dramatic changes that accompany the platelet activation process.

Blood Platelets↗
PubMed1999

A model for the description and interpretation of suicide prevention.

Views about suicide prevention are based on underlying beliefs about the origins of problems and basic concepts about humankind. These implicit theories have an effect on prevention practices. Making these views explicit is one of the keys for the further development of suicide prevention. In this study a paradigm for analyzing suicide prevention of means of a coding frame and interpreting findings by means of theoretical models of prevention was elaborated. The analysis was based on empirical data consisting of definitions of prevention given by psychologists (N = 34) participating in the national suicide prevention project in Finland. The study demonstrates that suicide prevention can be differentiated at the operational level by means of the analysis method generated. Moreover, the findings can be interpreted according to theoretical criteria. Views expressed by the psychologists seemed to correspond largely to central features of current prevention models. Furthermore, the data can be seen to serve as an empirical validation of these models. Suicide prevention proved to be a multifactorial concept manifesting mainly process theory and interactional explanations of suicidality, and prevention practices fell into a simple typology of four categories.

Adult↗
PubMed2002

Representation primitives, process models and patient data in computer-interpretable clinical practice guidelines: a literature review of guideline representation models.

Representation of clinical practice guidelines in a computer-interpretable format is a critical issue for guideline development, implementation, and evaluation. We studied 11 types of guideline representation models that can be used to encode guidelines in computer-interpretable formats. We have consistently found in all reviewed models that primitives for representation of actions and decisions are necessary components of a guideline representation model. Patient states and execution states are important concepts that closely relate to each other. Scheduling constraints on representation primitives can be modeled as sequences, concurrences, alternatives, and loops in a guideline's application process. Nesting of guidelines provides multiple views to a guideline with different granularities. Integration of guidelines with electronic medical records can be facilitated by the introduction of a formal model for patient data. Data collection, decision, patient state, and intervention constitute four basic types of primitives in a guideline's logic flow. Decisions clarify our understanding on a patient's clinical state, while interventions lead to the change from one patient state to another.

Artificial Intelligence↗
PubMed1998

Animal models of atherosclerosis and interpretation of drug intervention studies.

Atherosclerosis has often been defined as a multifactoral disease; however, a common risk factor associated with accelerated vascular disease in man or animals is an elevated plasma cholesterol level. Even though there is no one perfect animal model that completely replicates the stages of human atherosclerosis, cholesterol feeding and mechanical endothelial injury are two common features shared by most models of atherosclerosis. The models may differ with respect to degree of dietary cholesterol supplementation, length of hypercholesterolemia, dietary regimen and type, duration and degree of mechanical endothelial injury. With the advent of genetic engineering, transgenic mouse models have supplemented the classical dietary cholesterol induced disease models such as the cholesterol-fed hamster, rabbit, pig and monkey. The desire to limit the progression of atherosclerosis has spawned numerous drug intervention studies. Biochemical as well as morphologic and morphometric changes in the extent, structure and composition of atherosclerotic lesions following drug intervention have become major endpoints of in vivo drug intervention studies. Interpretations of alterations in vascular pathology following drug administration are often confounded by associated changes in plasma lipids and lipoproteins, limitation of the animal models and additional properties of compounds unrelated to their primary mode of action. Thus, the current review will summarize the pathology of atherosclerosis, describe various animal models of vascular disease and provide a critical review of the methods utilized and conclusions drawn when evaluating pharmacologic agents in animals.

Animals↗
PubMed2026

Comprehensive Evaluation and Explainable Interpretation of Peptide-HLA Binding Prediction Tools.

Accurate prediction of peptide binding to human leukocyte antigen class I (HLA-I) molecules is critical for advancing immunological research, particularly in vaccine design and immunotherapy. However, limitations in model performance, interpretability, and dataset quality impede the widespread adoption of existing predictive tools. Here, we present a comprehensive evaluation of 17 HLA-I peptide binding prediction models, utilizing a meticulously curated dataset comprising over 290,000 peptides spanning 44 HLA-I alleles. We assessed model accuracy, robustness, and interpretability, employing explainability techniques such as SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations) to elucidate underlying prediction mechanisms. Our results reveal substantial performance disparities, with self-attention-based models, including STMHCpan and BigMHC, exhibiting superior accuracy. Notably, the capsule network model CapsNet-MHC_AN demonstrated robust performance. Models trained on eluted ligand datasets outperformed those relying on binding affinity data, underscoring the critical role of high-quality training data. Ensemble and multi-algorithm approaches further improved prediction reliability. These findings highlight the need for ongoing innovation in model architecture, integration of diverse and high-quality datasets, and incorporation of structural predictors to develop more accurate, interpretable, and clinically applicable HLA-I peptide binding prediction tools.

HLA-I binding↗
PubMed1998

Investigation of phase transitions in bilayer membranes.

This article described three techniques used to study phase transitions in phospholipid bilayers. The complementarity of the three techniques in characterizing the thermotropic and structural properties of phospholipid bilayers has been demonstrated by describing their use to characterize a series of mixed-chain-length PCs. It has been shown that an understanding of the energetics that govern the packing of phospholipid chains in the gel phase can be used to construct a model to interpret thermodynamic data of the PCs. This model, in turn, provided a framework for designing and interpreting the Raman spectroscopic and X-ray diffraction experiments on this series of phospholipids. The result was a complete description of the phase transitions and gel phase packing properties of the mixed-chain-length PCs. The phase diagram of Fig. 5B has been expanded to include the mixed-chain-length PC series C18C18PC through C18C0PC. Furthermore, the phase diagram and the chain inequivalence parameter have been shown to describe the behavior of any mixed-chain-length PC, irrespective of the lengths of the hydrocarbon chains or the position of the chains on the glycerol backbone. This is demonstrated by the additional mixed-chain-length PCs plotted in Fig. 5B. With minor modifications, the phase diagram also accurately describes the behavior of mixed-chain-length phosphatidylethanolamines, sphingomyelins, and unsaturated PCs. Finally, it has been demonstrated that a correlation exists between the thermodynamic and the Raman spectroscopic parameters determined for the phase transition of phospholipid bilayers. This correlation is based on the common chain energetics being measured by these two techniques.

Calorimetry, Differential Scanning↗
PubMed2001

Statistical methods for the analysis and presentation of the results of bone marrow transplants. Part 2: Regression modeling.

In this paper, we address methods of multivariate regression. We discuss the value of regression compared to matched pairs analysis, methods of coding variables, basic concepts of the Cox model and interpretation of results of the Cox model. We present methods of handling variables whose effect changes with time. We present methods to check the assumptions of the Cox regression. Finally, and perhaps most importantly, we provide suggestions for presenting the results in clear and thorough tables and graphs.

Bone Marrow Transplantation↗
PubMed1978

Corneal temperatures--a study of normal and laser-injured corneas in the Dutch belted rabbit.

Air Force laser safety standards are developed from laser-exposure data obtained in studies using experimental animals and from biomathematical modeling procedures. Interpretation of research data and predictive modeling calculations are enhanced by knowledge of the temperature values of the tissue absorption sites. Corneal temperature values for the normal, the anesthetized, and the laser-injured Dutch belted rabbit are presented and compared with values obtained in other studies. The corneal temperatures were measured by infrared radiometry.

Anesthesia, Intravenous↗
PubMed1999

A computer-aided approach to the structural analysis and modification of a large circulatory system model.

The purpose of this study is to show an approach to making an intelligent support system for understanding and modifying a large circulatory system model using techniques of system analysis. Structural analysis makes it possible to visualize hierarchies of Coleman's circulatory model Human. Two techniques are successively applied for structural analysis, model reduction and graph analysis by interpretative structural modeling (ISM). First, the analysis for model reduction removes input-output relations with an input-output gain less than a given threshold, and second, the ISM technique applied to the reduced model of Human provides hierarchical directed graphs. The proposed approach: 1) enables visualization of a hierarchy graph of cause and effect relations of the large circulatory model, 2) suggests control and diagnostic information to the model by tracing back a path in the hierarchy, and 3) allows the user to modify the circulatory model. The efficiency and performance of the proposed approach demonstrates technical indications of success in analyzing and justifying experimental evidences with the online help of the system.

Algorithms↗