Biomedical subjects

M C Jaulent

Publications and source records attributed to M C Jaulent.

At least 19 recordsLinked to original sources

PubMed2001

Integrating medical applications in an open architecture through generic and reusable components.

Allowing exchange of information and cooperation among network-wide distributed and heterogeneous applications is a major need of current health care information systems. It forces the development of open and modular integration architectures. Major issues in the development include defining a flexible and robust federation model, developing interaction and communication facilities as well as the mechanism insuring semantic interoperability. We developed generic and reusable software components to ease the construction of any integration platform. The Pilot and the Mediator Service components facilitate the execution of services and the meaningful transformation of information. They have been tested in the context of the SynEx European project to construct a multi-agents based integration architecture. The possibility of such architectures to take into account the issue of semantic interoperability is further discussed.

Computer Communication Networks↗

PubMed2001

Expression and meaning of medical language: building an epistemological framework for the study of semantic distance.

The use of language for communication purposes introduces a semantic gap in data processing. We have developed a new tool called semantic distance to overcome this gap. A previous presentation of our tool was based on an empirical approach. Here, we examine the properties and theoretical foundation of semantic distance from a rationalistic perspective. We present an epistemological framework to explain the meaning of semantic distance. As a tool, the purpose of semantic distance is the exchange of data or knowledge based on linguistic expressions between software components. We propose a review of the relationship between language and its representation for data processing based on the underlying philosophical assumptions. Description logic is a new paradigm in the medical informatics community to express relations between terms in "is-a" hierarchies. The utility of semantic distance is compared to description logic for communication purposes in different use cases.

Knowledge↗

PubMed2001

From paper-based to electronic guidelines: application to French guidelines.

In order to develop an internet-based decision support system, making available for French general practitioners several prevention guidelines is was necessary to implement paper based guideline. We propose a framework allowing to transform paper based practice guideline into their electronic form. Three different problems were identified: computability (e.g. determinism of the eCPG), logic (e.g. ambiguities when combining Booleans operators) and external validity (i.e. stability of decision for variations around thresholds and proportion of subjects classified in the various terminal nodes). The last problem concerned documentation of evidence: the level of evidence was associated only with the terminal decision node and not with the pathway through the decision tree. We concluded that computerisation of guidelines is not possible without expertise or authors advice. To improve computability it is necessary to provide authors with a framework that checks ambiguities, and logical errors.

Algorithms↗

PubMed2001

Logistic regression model: an assessment of variability of predictions.

Risk prediction models available for cardiovascular prevention are statistical or based on machine learning methods. This paper investigates whether the logistic regression method can be considered as reference for validation of other methods. In order to test the stability of the predictions using this method, we performed two types of analyses on 50 random training and test samples drawn from the same database. In first analyses three models were obtained by forced entry of different sets of four variables. In second analyses, models were built with increasing number of predictive variables. The predictive performance was assessed by the area under the ROC curve. Although across-samples variability is low for a given model, it is large enough to lead to wrong conclusions when comparing different prediction methods. We also suggest that a low events-per-variable ratio alters the stability of a model's coefficients but does not affect the variability of prediction performance.

Area Under Curve↗

PubMed2001

Modeling uncertainty in computerized guidelines using fuzzy logic.

Computerized Clinical Practice Guidelines (CPGs) improve quality of care by assisting physicians in their decision making. A number of problems emerges since patients with close characteristics are given contradictory recommendations. In this article, we propose to use fuzzy logic to model uncertainty due to the use of thresholds in CPGs. A fuzzy classification procedure has been developed that provides for each message of the CPG, a strength of recommendation that rates the appropriateness of the recommendation for the patient under consideration. This work is done in the context of a CPG for the diagnosis and the management of hypertension, published in 1997 by the French agency ANAES. A population of 82 patients with mild to moderate hypertension was selected and the results of the classification system were compared to whose given by a classical decision tree. Observed agreement is 86.6% and the variability of recommendations for patients with close characteristics is reduced.

Decision Making, Computer-Assisted↗

PubMed2001

A consensus approach to maintain a knowledge based system in pathology.

The IDEM (Images and Diagnosis from Example in Medicine) software is a computerized environment able to store unambiguous descriptions of histopathologic images from pathologists. Medical imaging could benefit from such environments if they can easily and continuously be maintained. Within the IDEM environment, we developed a knowledge management module coupled with a consensus module to support knowledge acquisition and maintenance by the experts of the domain. Two pathologists, a senior and junior pathologist, reviewed fifty-three cases of breast pathology. Our findings indicate 1) that the IDEM knowledge management module allows experts to describe images by selecting terms and defining new ones if necessary, allowing the construction of a glossary for the domain and 2) that the consensus module, developed to store valid multi-experts cases, contributes also to validate new terms of the glossary and to refine semantic distance between terms. Such methodology could be applied to others highly evolving medical domains.

Artificial Intelligence↗

PubMed2000

[Can the degree of renal artery stenosis be automatically quantified?].

The objective of the reported study is to validate a computer system, QUASAR, dedicated to the quantification of renal artery stenoses. This system estimates automatically the reference diameter and calculates the minimum diameter to compute a degree of stenosis. A hundred and eighty images of atheromatous stenoses between 10% and 80% were collected from two French independent protocols. For the 49 images of the EMMA protocol, the results from QUASAR were compared with the visual estimation of an initial investigator and with the results from a reference method based on a panel of fixe experienced experts. For the 131 images of the ASTARTE protocol, the results from QUASAR were compared with those from a semi-automatic quantification system and with those from a system based on densitometric analysis. The present work validates QUASAR in a population of narrow atheromatous stenoses (> 50%). In the context of the EMMA protocol, QUASAR is not significantly different from the mean of the fixe experts. It is unbiased and more precise than the estimation of a single investigator. In the context of the ASTARTE protocol, there is no significant difference between the three methods for the stenoses higher than 50%, however, globally, QUASAR surestimates significantly (up to 10%) the degree of stenosis.

Aged↗

PubMed2000

Using semantic distance for the efficient coding of medical concepts.

OBJECTIVE: To use the notion of semantic distance to find the nearest neighbors of a medical concept in a controlled vocabulary. MATERIAL AND METHOD: 392 concepts from the cardiovascular chapter of the ICD-10 were projected on the axes of SNOMED III. Distances were measured on each axis and the resulting distance was found using a Lp norm. RESULTS: The distance between a set of ischemic diseases and a set of non-ischemic diseases was significant (p < 0.0001). Our method was validated by finding the k nearest neighbors of ten different diagnoses from the ICD-10 cardiovascular chapter. DISCUSSION: The availability of SNOMED-RT should improve our method. Several more steps are necessary to provide an ideal coding tool.

Cardiovascular Diseases↗

PubMed2000

Models to predict cardiovascular risk: comparison of CART, multilayer perceptron and logistic regression.

The estimate of a multivariate risk is now required in guidelines for cardiovascular prevention. Limitations of existing statistical risk models lead to explore machine-learning methods. This study evaluates the implementation and performance of a decision tree (CART) and a multilayer perceptron (MLP) to predict cardiovascular risk from real data. The study population was randomly splitted in a learning set (n = 10,296) and a test set (n = 5,148). CART and the MLP were implemented at their best performance on the learning set and applied on the test set and compared to a logistic model. Implementation, explicative and discriminative performance criteria are considered, based on ROC analysis. Areas under ROC curves and their 95% confidence interval are 0.78 (0.75-0.81), 0.78 (0.75-0.80) and 0.76 (0.73-0.79) respectively for logistic regression, MLP and CART. Given their implementation and explicative characteristics, these methods can complement existing statistical models and contribute to the interpretation of risk.

Artificial Intelligence↗

PubMed2000

A property concept frame representation for flexible image-content retrieval in histopathology databases.

In histopathology databases, images descriptions are collections of properties provided by experts. Image content retrieval implies comparison of such properties. The objective of this work is to enrich the traditional attribute-value representation of properties in order to take into account the polymorphism and subjectivity of properties and to manage the comparison process. In this paper we define a property concept frame (PCF) representation based on fuzzy logic to handle both representation and comparison. Seven quantifiable morphological characteristics were selected from histopathological reports to illustrate the variety of fuzzy predicates and linguistic terms in properties. The PCF representation has been tested in the context of breast pathology. It is concluded that the PCF representation provides a unification scheme to retrieve in images morphological characteristics that are described in different ways. It may enhance the relevancy of applications in various contexts such as image content-based retrieval or case-based reasoning from images.

Breast↗

PubMed2000

Towards content-based image retrieval in a HIS-integrated PACS.

Development of a Picture Archiving and Communications System (PACS) with a large and diverse set of medical images will lead to large digital libraries that can be accessed to provide improved support for patient care, research and education. New representational and retrieval models for clinical images are required to address these issues. The PACS at the Georges Pompidou Hospital (GPH) is integrated in the hospital information system (HIS), and several modalities from medical imaging departments have been attached to it. The two main axes of the GPH PACS project were 1) HIS-integration to allow hospital-wide access to the images based on demographic and procedure-type information and 2) the development of content-based image retrieval to enhance the medical impact of image retrieval in daily practice.

Breast↗

PubMed2000

[Cardiovascular risk and management of patients with hypertension].

Whereas cardiovascular diseases remain a priority amongst preventable diseases and that their risk factors, especially hypertension, remain inadequately controlled, new tools, such as the cardiovascular risk, would allow better targeting of treatment on high risk patients. All the evidence is in favour of prevention based on the estimation of the risk, but this article summarises the problems which this strategy continues to pose. In particular, necessity of validation at several levels of the equation or equations used (exactitude of the estimated risk, its accuracy and transportability); influence of the mode of presentation of the risk on the perception and decision of the physician and patient; practical application of the strategy; choice of decisional threshold respecting the requirements of different age groups, and presentation to physicians (recommendations and/or computerisation?).

Cardiovascular Diseases↗

PubMed1999

Interobserver variability in the interpretation of renal digital subtraction angiography.

OBJECTIVE: Our purpose was to analyze interobserver variability in the interpretation of renal digital subtraction angiography and to describe the main factors associated with observer discrepancies. MATERIALS AND METHODS: Forty-nine cases of unilateral atheromatous renal artery stenosis of more than 60% were quantified first by local investigators in a multicenter study and then by five other radiologists. Differences between radiologists for the minimum diameter (Dmin), the reference diameter (Dref), and the percentage of stenosis of the renal arteries were analyzed. Interpretations by the local investigators were then compared with the gold standard, defined as the mean for the five radiologists. RESULTS: The average SD for estimation of all renal artery stenoses by all radiologists was 7% for stenosis percentage, 0.5 mm for Dmin, and 0.7 mm for Dref. Main discrepancies occurred more frequently in cases of weakly opacified renal artery stenosis and poststenotic dilatation. The observations of local investigators disagreed by more than two SDs (14%) with the gold standard for 11 of 49 cases (22%). CONCLUSION: The accuracy of digital subtraction angiography in renal artery interpretations is poor because of variations in evaluating both Dmin and Dref. Precise and reproducible methods for quantification of renal artery stenosis are required.

Adult↗

PubMed1999

Quantifying stenosis in renal arteriograms: a fuzzy syntactic analysis.

The introduction of fuzzy logic improves a system for the automatic quantification of renal artery lesions seen in digital subtraction angiograms. A two-step approach has been followed. An earlier system based on non-fuzzy syntactic analysis provided a clear symbolic description of the stenotic lesions. Although this system worked correctly, it did not take into account the variability and uncertainty inherent to image processing and to knowledge on the reference diameter. This system has been improved by the introduction of fuzzy logic in the representation of the reference diameter. It provides a description of the stenosis in terms of fuzzy quantities. To illustrate the benefits of the fuzzy approach, the results of the two systems have been compared by plotting the differences of an index of variability. It appears that the differences are statistically different when using a two-tailed paired t-test (t = 2.37; p = 0.025). The result shows that the fuzzy approach is better than a non-fuzzy approach in the sense that the index of variability is reduced significantly.

Angiography, Digital Subtraction↗

PubMed1999

Decision aids for triage of patients with chest pain: a systematic review of field evaluation studies.

We performed an overview of published controlled trials to assess the overall effectiveness of decision aids directed at improving triage of patients with acute chest pain. Searches of the Medline database identified 11 randomized or quasi-randomized controlled trials testing various decision aids: risk stratification system (n = 6), practice guidelines (n = 3), and formalized protocols of care (n = 2). Sensitivity, specificity of the decision aid and length of stay (LOS) in the intensive care unit (ICU) were the main outcomes. Decision aids slightly modified sensitivity and specificity (available in 5 studies), but sensitivity was already high in reference groups. Among the 9 studies providing information on LOS, 7 showed a statistically significant difference favoring the decision aid. The level of evidence concerning the efficacy of decision aids in this domain is relatively low. Larger and appropriately designed clinical trials are required to show an impact on acute cardiac ischaemia complications and mortality.

Humans↗

PubMed1999

A visual coding system in histopathology and its consensual acquisition.

Divergent descriptions of histopathologic images induce inter- and intra-observer variability in diagnosis. Even though a controlled terminology exists to describe medical imaging, pathologists do not always agree on the visual representation of the descriptive terms. The main purpose of our work is to define a methodology to build a standardized visual coding system unambiguously characterizing the terms of a microglossary. The methodology follows two steps: 1) the acquisition of experts' descriptions of images using the microglossary and 2) a consensus derivation. The procedure was applied on a set of 85 histopathological images of breast tumors described by two experts. Among the 339 objects selected in images, 176 were detected by both experts, 77% localized at the same place and 25% also identically labeled. The microglossary was enriched and illustrated via the resulting consensual descriptions. The contribution of this work supports relevant indexing of biomedical images and image-related information.

Breast Neoplasms↗

PubMed1999

Decision aids for triage of patients with chest pain: a systematic review of field evaluation studies.

Chest Pain↗

PubMed1998

IDEM: a Web application of case-based reasoning in histopathology.

Different software engineering and artificial intelligence methods can be used to design Internet retrieval of prototypical medical images. We used the case-based reasoning (CBR) approach to provide an 'intelligent' access to a collection of illustrated medical cases through the Internet. This paper presents a Web interface for the CBR system IDEM (image and diagnosis from examples in medicine) in the domain of breast pathology. Thanks to the definition of a similarity measure between the descriptions of cases we propose a flexible querying of the case-base and a quantitative browsing among cases through similarity links. The resemblance rates provided by the system argue for the quality and the relevancy of the retrieved data. The flexibility of the querying process is robust to missing information and could be adapted to a daily practice. The CBR approach is a promising method for a clinical relevant and an efficient retrieval of reference images and diagnosis clues through Internet.

Artificial Intelligence↗