gilles cohen - Academia.edu (original) (raw)

Papers by gilles cohen

Antimicrobial Resistance and Infection Control, May 30, 2013

Background: Targeted screening of patients at high risk for methicillin-resistant Staphylococcus ... more Background: Targeted screening of patients at high risk for methicillin-resistant Staphylococcus aureus (MRSA) carriage is an important component of MRSA control programs, which rely on prediction tools to identify those high-risk patients. Most previous risk studies reported a substantial rate of patients who are eligible for screening, but failed to be enrolled. The characteristics of these missed patients are seldom described. We aimed to determine the rate and characteristics of patients who were missed by a MRSA screening programme at our institution to see how the failure to include these patients might impact the accuracy of clinical prediction tools. Findings: From March-June 2010 all patients admitted to 13 internal medicine wards at the University of Geneva Hospital (HUG) were prospectively screened for MRSA carriage. Of 1968 patients admitted to the ward, 267 patients (13.6%) failed to undergo appropriate MRSA screening. Forty-one (2.4%) screened patients were MRSA carriers at admission. On multivariate regression, patients who were missed by screening were more likely to be aged < 50 years (OR 2.4 [1.4-3.9]), transferred to internal medicine from another ward in the hospital (OR 2.8 [1.1-7.1]), and have a history of malignancy ). There was no significant difference in the rate of previous MRSA carriage between screened and unscreened patients. Conclusions: Our findings highlight the potential bias that "missed" patients may introduce into MRSA risk scores. Reporting on the proportions and characteristics of missed patients is essential for accurate interpretation of MRSA prediction tools.

Antimicrobial Resistance and Infection Control, 2013

Background The incidence of extended-spectrum beta-lactamase producing-enterobacteriacae (ESBL-E)... more Background The incidence of extended-spectrum beta-lactamase producing-enterobacteriacae (ESBL-E) infection is rising worldwide. We aimed to determine the prevalence and nosocomial acquisition rate of ESBL-E as well as the risk factors for ESBL-E carriage and acquisition amongst patients consecutively admitted to 13 internal medicine units at our hospital who were not previously known to be ESBL-E carriers. Findings We screened all patients admitted or transferred to internal medicine units for ESBL-E on admission and discharge using rectal swabs. Of 1072 patients screened, 51 (4.8%) were carriers of an ESBL-E at admission. Of 473 patients who underwent admission and discharge screening, 21 (4.4%) acquired an ESBL-E. On multivariate analysis, diabetes mellitus without end-organ complications (OR 2.87 [1.09-7.08]), connective tissue disease (OR 7.22 [1.17-44.59]), and liver failure (OR 8.39 [1.55-45.45]) were independent risk factors for carriage of an ESBL-E upon admission to hospit...

In hospitalized populations, there is significant heterogeneity in patient characteristics, disea... more In hospitalized populations, there is significant heterogeneity in patient characteristics, disease severity, and treatment responses, which generally translates into very different related outcomes and costs. A better understanding of this heterogeneity can lead to better management, more effective and efficient treatments by personalizing care to better meet patients' profiles. Thus, identifying distinct clinical profiles among patients can lead to more homogenous subgroups of patients. Super-utilizers (SUs) are such a group, who contribute a substantial proportion of health care costs and utilize a disproportionately high share of health care resources. This study uses cost, utilization metrics and clinical information to segment the population of patients (N=32,759) admitted to the University Hospital of Geneva in 2019 and thus identifies the characteristics of its SUs group using Latent Class Analysis. This study shows how cluster analysis might be valuable to hospitals for identifying super-utilizers within their patient population and understanding their characteristics.

Lecture Notes in Computer Science, 2004

This paper addresses the problem of tuning hyperparameters in support vector machine modeling. A ... more This paper addresses the problem of tuning hyperparameters in support vector machine modeling. A Direct Simplex Search (DSS) method, which seeks to evolve hyperparameter values using an empirical error estimate as steering criterion, is proposed and experimentally evaluated on real-world datasets. DSS is a robust hill climbing scheme, a popular derivative-free optimization method, suitable for low-dimensional optimization problems for which the computation of the derivatives is impossible or difficult. Our experiments show that DSS attains performance levels equivalent to that of GS while dividing computational cost by a minimum factor of 4.

Journal of Digital Imaging, 2008

In this paper, we compare five common classifier families in their ability to categorize six lung... more In this paper, we compare five common classifier families in their ability to categorize six lung tissue patterns in high-resolution computed tomography (HRCT) images of patients affected with interstitial lung diseases (ILD) and with healthy tissue. The evaluated classifiers are naive Bayes, k-nearest neighbor, J48 decision trees, multilayer perceptron, and support vector machines (SVM). The dataset used contains 843 regions of interest (ROI) of healthy and five pathologic lung tissue patterns identified by two radiologists at the University Hospitals of Geneva. Correlation of the feature space composed of 39 texture attributes is studied. A grid search for optimal parameters is carried out for each classifier family. Two complementary metrics are used to characterize the performances of classification. These are based on McNemar's statistical tests and global accuracy. SVM reached best values for each metric and allowed a mean correct prediction rate of 88.3% with high class-specific precision on testing sets of 423 ROIs.

Studies in Health Technology and Informatics

In hospitalized populations, there is significant heterogeneity in patients’ characteristics, dis... more In hospitalized populations, there is significant heterogeneity in patients’ characteristics, disease severity, and treatment responses, which translates into different related outcomes and costs. Identifying inpatient clusters with similar clinical profiles could lead to better quality and personalized care while improving clinical resources used. Super-utilizers (SUs) are one such a group, who contribute a substantial proportion of health care costs and utilize a disproportionately high share of health care resources. This study uses cost, utilization metrics and clinical information to segment the population of patients (N=32,759) admitted to the University Hospitals of Geneva per year in 2017 - 2019. Using Latent Class Analysis it identifies 8 subgroups with highly similar patients demographics, medical conditions, types of service and costs within groups and which are highly different between groups. As such 82% of all SU patients, 99% of all patients less than 20 years old and...

Objective: An important problem that arises in hospitals is the monitoring and detection of nosoc... more Objective: An important problem that arises in hospitals is the monitoring and detection of nosocomial or hospital acquired infections (NIs). This paper describes a retrospective analysis of a prevalence survey of NIs done in the Geneva University Hospital. Our goal is to identify patients with one or more NIs on the basis of clinical and other data collected during the survey.

Nosocomial or hospital-acquired infections (NIs) have become a major concern not only in health c... more Nosocomial or hospital-acquired infections (NIs) have become a major concern not only in health care institutions but also among the gen-eral public. Since 1994 the Geneva University Hospital has been undertaking yearly prevalence studies in order to monitor and detect NIs. This paper describes a retrospective analysis of the results of one such study. Our goal is to identify patients with one or more NIs on the basis of clinical and other data collected during the survey. In this classification task, the main difficulty resides in the significant imbalance between positive or infected (11%) and negative (89%) cases. To cope with class imbalance, we investigate a support vector algorithm in which asymmetrical margins are tuned to improve recognition of rare positive cases. Experiments have shown this approach to be effective for the NI detection problem: we obtained a sensitivity rate of 92%, significantly better than the highest sensitivity (87%) obtained via novel resampling strat...

Machine Learning Techniques and Data Science

In hospitalized populations, there is significant heterogeneity in patient characteristics, disea... more In hospitalized populations, there is significant heterogeneity in patient characteristics, disease severity, and treatment responses, which generally translates into very different related outcomes and costs. A better understanding of this heterogeneity can lead to better management, more effective and efficient treatments by personalizing care to better meet patients' profiles. Thus, identifying distinct clinical profiles among patients can lead to more homogenous subgroups of patients. Super-utilizers (SUs) are such a group, who contribute a substantial proportion of health care costs and utilize a disproportionately high share of health care resources. This study uses cost, utilization metrics and clinical information to segment the population of patients (N=32,759) admitted to the University Hospital of Geneva in 2019 and thus identifies the characteristics of its SUs group using Latent Class Analysis. This study shows how cluster analysis might be valuable to hospitals for...

Studies in health technology and informatics, 2005

This paper addresses the model selection problem for Support Vector Machines. A hybrid genetic al... more This paper addresses the model selection problem for Support Vector Machines. A hybrid genetic algorithm guided by Direct Simplex Search to evolves hyperparameter values using an empirical error estimate as a steering criterion. This approach is specificaly tailored and experimentally evaluated on a health care problem which involves discriminating 11 % nosocomially infected patients from 89 % non infected patients. The combination of Direct Search Simplex with GAs is shown to improve the performance of GAs in terms of solution quality and computational efficiency. Unlike most other hyperparameter tuning techniques, our hybrid approach does not require supplementary effort such as computation of derivatives, making them well suited for practical purposes. This method produces encouraging results: it exhibits high performance and good convergence properties.

Task I . Our approach uses thesaural resources (from the UMLS) together with a variant of the Por... more Task I . Our approach uses thesaural resources (from the UMLS) together with a variant of the Porter stemmer for string normalization. Gene and Protein Entities (GPE) of the collection (525,938 MedLine citations) were simply marked up by dictionary look up during the indexing in order to avoid erroneous conflation: strings not found in the UMLS Specialist lexicon (augmented with various English lexical resources) were considered as GPE and were moderately overweighted. In the same spirit like other TREC competitors [23] for task I, an overweighting factor was also applied to features belonging to Medical Subject Headings (MeSH) and found in MedLine citations using a MeSH mapping tool [1]. A standard vector space IR engine with tf-idf parameters was used for indexing the Genomic collection: article’s ti tles, MeSH and RN terms, and abstact fields were selected. Best average precisions were obtained with atc.ntn (using the SMART notation) schemes: 16.71 (standard) vs. 17.02 (using UML...

Antimicrobial Resistance and Infection Control, 2013

Background Targeted screening of patients at high risk for methicillin-resistant Staphylococcus a... more Background Targeted screening of patients at high risk for methicillin-resistant Staphylococcus aureus (MRSA) carriage is an important component of MRSA control programs, which rely on prediction tools to identify those high-risk patients. Most previous risk studies reported a substantial rate of patients who are eligible for screening, but failed to be enrolled. The characteristics of these missed patients are seldom described. We aimed to determine the rate and characteristics of patients who were missed by a MRSA screening programme at our institution to see how the failure to include these patients might impact the accuracy of clinical prediction tools. Findings From March-June 2010 all patients admitted to 13 internal medicine wards at the University of Geneva Hospital (HUG) were prospectively screened for MRSA carriage. Of 1968 patients admitted to the ward, 267 patients (13.6%) failed to undergo appropriate MRSA screening. Forty-one (2.4%) screened patients were MRSA carriers...

Infection Control and Hospital Epidemiology, 2014

Objective. To test the hypothesis that methicillin-susceptible Staphylococcus aureus (MSSA) carri... more Objective. To test the hypothesis that methicillin-susceptible Staphylococcus aureus (MSSA) carriage may protect against nosocomial methicillin-resistant S. aureus (MRSA) acquisition by competing for colonization of the anterior nares. Design. Prospective cohort and nested case-control study. Setting. Swiss university hospital. Patients. All adult patients admitted to 14 wards of the general medicine division between April 1 and October 31, 2007. Methods. Patients were screened for MRSA and MSSA carriage at admission to and discharge from the division. Associations between nosocomial MRSA acquisition and MSSA colonization at admission and other confounders were analyzed by univariable and multivariable analysis. Results. Of 898 patients included, 183 (20%) were treated with antibiotics. Nosocomial MRSA acquisition occurred in 70 (8%) of the patients (case patients); 828 (92%) of the patients (control subjects) were free of MRSA colonization at discharge. MSSA carriage at admission w...

Abstract. An important problem that arises in hospitals is the monitoring and detection of nosoco... more Abstract. An important problem that arises in hospitals is the monitoring and detection of nosocomial or hospital acquired infections (NIs). This paper describes a retrospective analysis of a prevalence survey of NIs done in the Geneva University Hospital. Our goal is to identify patients with one or more NIs on the basis of clinical and other data collected during the survey. In this classification task, the main difficulty resides in the significant imbalance between positive or infected (11%) and negative (89%) cases. To remedy class imbalance, we propose a novel approach in which both oversampling of rare positives and undersampling of the non infected majority rely on synthetic cases generated via class-specific subclustering. Experiments have shown this approach to be remarkably more effective than classical random resampling methods. 1

Antimicrobial resistance and infection control, Jan 30, 2013

Targeted screening of patients at high risk for methicillin-resistant Staphylococcus aureus (MRSA... more Targeted screening of patients at high risk for methicillin-resistant Staphylococcus aureus (MRSA) carriage is an important component of MRSA control programs, which rely on prediction tools to identify those high-risk patients. Most previous risk studies reported a substantial rate of patients who are eligible for screening, but failed to be enrolled. The characteristics of these missed patients are seldom described. We aimed to determine the rate and characteristics of patients who were missed by a MRSA screening programme at our institution to see how the failure to include these patients might impact the accuracy of clinical prediction tools. From March-June 2010 all patients admitted to 13 internal medicine wards at the University of Geneva Hospital (HUG) were prospectively screened for MRSA carriage. Of 1968 patients admitted to the ward, 267 patients (13.6%) failed to undergo appropriate MRSA screening. Forty-one (2.4%) screened patients were MRSA carriers at admission. On mu...

Studies in health technology and informatics, 2011

This paper considers the model selection problem for Support Vector Machines. A well-known deriva... more This paper considers the model selection problem for Support Vector Machines. A well-known derivative Pattern Search method, which aims to tune hyperparameter values using an empirical error estimate as a steering criterion, is proposed. This approach is experimentally evaluated on a health care problem which involves discriminating nosocomially infected patients from non-infected patients. The Hooke and Jeeves Pattern Search (HJPS) method is shown to improve the results achieved by Grid Search (GS) in terms of solution quality and computational efficiency. Unlike most other parameter tuning techniques, our approach does not require supplementary effort such as computation of derivatives, making them well suited for practical purposes. This method produces encouraging results: it exhibits good performance and convergence properties.

Nosocomial infections (NIs)-those acquired in health care settingsrepresent one of the major caus... more Nosocomial infections (NIs)-those acquired in health care settingsrepresent one of the major causes of increased mortality in hospitalized patients. As they are a real problem for both patients and health authorities, the development of an effective surveillance system to monitor and detect them is of paramount importance. This paper presents a retrospective analysis of a prevalence survey of NIs done in the Geneva University Hospital. The objective is to identify patients with one or more NIs based on clinical and other data collected during the survey. In this classification task, the main difficulty lies in the significant imbalance between positive and negative cases. To overcome this problem, we investigate one-class Parzen density estimator which can be trained to differentiate two classes taking examples from a single class. The results obtained are encouraging: whereas standard 2-class SVMs scored a baseline sensitivity of 50.6% on this problem, the one-class approach increased sensitivity to as much as 88.6%. These results suggest that one-class Parzen density estimator can provide an effective and efficient way of overcoming data imbalance in classification problems.

Image management, analysis, and retrieval are currently very active research fields mainly becaus... more Image management, analysis, and retrieval are currently very active research fields mainly because of the large amount of visual data being produced in modern hospitals, and the lack of applications dealing with these data. Most often, the goal is to aid the diagnostic process. Unfortunately, only very few medical image retrieval systems are currently used in clinical routine. One application domain with a high potential for automatic image retrieval is the analysis and retrieval of lung CTs. A first user study in the United States (Purdue University) shows that these systems allow improving the diagnostic quality significantly. This article describes an approach to an aid for lung CT diagnostics. The analysis incorporates several steps and the goal is to automate the process as much as possible for easy integration into clinical processes. Thus, several automatic steps are proposed from a selection of the most characteristic slices, to an automatic segmentation of the lung tissue and a classification on the segmented area into visual observation classes. Feedback to the MD is given in the form of marked regions in the images that appear to be different from the norm of healthy tissue. We currently work on a small set of training images with marked and annotated regions but a larger set of images for the evaluation of our algorithm is in work. The article currently only contains a short quantitative evaluation. For most tasks we use existing open source software such as Weka, GIFT, and itk. This allows an easy reproduction of the search results and limits the need for costly redevelopments.

This article deals with data on nosocomial infections acquired in the Geneva University Hospitals... more This article deals with data on nosocomial infections acquired in the Geneva University Hospitals. Goal of the work is to derive a model from a hospital-acquired infection (HAI) prevalence survey of year Y and apply them to a prevalence survey of years Y+1, Y+2. This analysis permits to evaluate the effectiveness of preventive measures taken after the prevalence survey in year Y. It also analyzes the robustness of the SVM algorithm on time-variable attributes. The model build on the dataset of year Y gives better results than in a previous study. The application of the model on the Y+1 and Y+2 prevalence surveys shows simultaneously improvements and deteriorations of 5 performance measures. This highlights the effectiveness of prevention and reduces the risk of HAI after the prevalence survey of year Y. We introduce a new method to detect redundancy in a dataset with the SVM algorithm.

Antimicrobial Resistance and Infection Control, May 30, 2013

Background: Targeted screening of patients at high risk for methicillin-resistant Staphylococcus ... more Background: Targeted screening of patients at high risk for methicillin-resistant Staphylococcus aureus (MRSA) carriage is an important component of MRSA control programs, which rely on prediction tools to identify those high-risk patients. Most previous risk studies reported a substantial rate of patients who are eligible for screening, but failed to be enrolled. The characteristics of these missed patients are seldom described. We aimed to determine the rate and characteristics of patients who were missed by a MRSA screening programme at our institution to see how the failure to include these patients might impact the accuracy of clinical prediction tools. Findings: From March-June 2010 all patients admitted to 13 internal medicine wards at the University of Geneva Hospital (HUG) were prospectively screened for MRSA carriage. Of 1968 patients admitted to the ward, 267 patients (13.6%) failed to undergo appropriate MRSA screening. Forty-one (2.4%) screened patients were MRSA carriers at admission. On multivariate regression, patients who were missed by screening were more likely to be aged < 50 years (OR 2.4 [1.4-3.9]), transferred to internal medicine from another ward in the hospital (OR 2.8 [1.1-7.1]), and have a history of malignancy ). There was no significant difference in the rate of previous MRSA carriage between screened and unscreened patients. Conclusions: Our findings highlight the potential bias that "missed" patients may introduce into MRSA risk scores. Reporting on the proportions and characteristics of missed patients is essential for accurate interpretation of MRSA prediction tools.

Antimicrobial Resistance and Infection Control, 2013

Background The incidence of extended-spectrum beta-lactamase producing-enterobacteriacae (ESBL-E)... more Background The incidence of extended-spectrum beta-lactamase producing-enterobacteriacae (ESBL-E) infection is rising worldwide. We aimed to determine the prevalence and nosocomial acquisition rate of ESBL-E as well as the risk factors for ESBL-E carriage and acquisition amongst patients consecutively admitted to 13 internal medicine units at our hospital who were not previously known to be ESBL-E carriers. Findings We screened all patients admitted or transferred to internal medicine units for ESBL-E on admission and discharge using rectal swabs. Of 1072 patients screened, 51 (4.8%) were carriers of an ESBL-E at admission. Of 473 patients who underwent admission and discharge screening, 21 (4.4%) acquired an ESBL-E. On multivariate analysis, diabetes mellitus without end-organ complications (OR 2.87 [1.09-7.08]), connective tissue disease (OR 7.22 [1.17-44.59]), and liver failure (OR 8.39 [1.55-45.45]) were independent risk factors for carriage of an ESBL-E upon admission to hospit...

In hospitalized populations, there is significant heterogeneity in patient characteristics, disea... more In hospitalized populations, there is significant heterogeneity in patient characteristics, disease severity, and treatment responses, which generally translates into very different related outcomes and costs. A better understanding of this heterogeneity can lead to better management, more effective and efficient treatments by personalizing care to better meet patients' profiles. Thus, identifying distinct clinical profiles among patients can lead to more homogenous subgroups of patients. Super-utilizers (SUs) are such a group, who contribute a substantial proportion of health care costs and utilize a disproportionately high share of health care resources. This study uses cost, utilization metrics and clinical information to segment the population of patients (N=32,759) admitted to the University Hospital of Geneva in 2019 and thus identifies the characteristics of its SUs group using Latent Class Analysis. This study shows how cluster analysis might be valuable to hospitals for identifying super-utilizers within their patient population and understanding their characteristics.

Lecture Notes in Computer Science, 2004

This paper addresses the problem of tuning hyperparameters in support vector machine modeling. A ... more This paper addresses the problem of tuning hyperparameters in support vector machine modeling. A Direct Simplex Search (DSS) method, which seeks to evolve hyperparameter values using an empirical error estimate as steering criterion, is proposed and experimentally evaluated on real-world datasets. DSS is a robust hill climbing scheme, a popular derivative-free optimization method, suitable for low-dimensional optimization problems for which the computation of the derivatives is impossible or difficult. Our experiments show that DSS attains performance levels equivalent to that of GS while dividing computational cost by a minimum factor of 4.

Journal of Digital Imaging, 2008

In this paper, we compare five common classifier families in their ability to categorize six lung... more In this paper, we compare five common classifier families in their ability to categorize six lung tissue patterns in high-resolution computed tomography (HRCT) images of patients affected with interstitial lung diseases (ILD) and with healthy tissue. The evaluated classifiers are naive Bayes, k-nearest neighbor, J48 decision trees, multilayer perceptron, and support vector machines (SVM). The dataset used contains 843 regions of interest (ROI) of healthy and five pathologic lung tissue patterns identified by two radiologists at the University Hospitals of Geneva. Correlation of the feature space composed of 39 texture attributes is studied. A grid search for optimal parameters is carried out for each classifier family. Two complementary metrics are used to characterize the performances of classification. These are based on McNemar's statistical tests and global accuracy. SVM reached best values for each metric and allowed a mean correct prediction rate of 88.3% with high class-specific precision on testing sets of 423 ROIs.

Studies in Health Technology and Informatics

In hospitalized populations, there is significant heterogeneity in patients’ characteristics, dis... more In hospitalized populations, there is significant heterogeneity in patients’ characteristics, disease severity, and treatment responses, which translates into different related outcomes and costs. Identifying inpatient clusters with similar clinical profiles could lead to better quality and personalized care while improving clinical resources used. Super-utilizers (SUs) are one such a group, who contribute a substantial proportion of health care costs and utilize a disproportionately high share of health care resources. This study uses cost, utilization metrics and clinical information to segment the population of patients (N=32,759) admitted to the University Hospitals of Geneva per year in 2017 - 2019. Using Latent Class Analysis it identifies 8 subgroups with highly similar patients demographics, medical conditions, types of service and costs within groups and which are highly different between groups. As such 82% of all SU patients, 99% of all patients less than 20 years old and...

Objective: An important problem that arises in hospitals is the monitoring and detection of nosoc... more Objective: An important problem that arises in hospitals is the monitoring and detection of nosocomial or hospital acquired infections (NIs). This paper describes a retrospective analysis of a prevalence survey of NIs done in the Geneva University Hospital. Our goal is to identify patients with one or more NIs on the basis of clinical and other data collected during the survey.

Nosocomial or hospital-acquired infections (NIs) have become a major concern not only in health c... more Nosocomial or hospital-acquired infections (NIs) have become a major concern not only in health care institutions but also among the gen-eral public. Since 1994 the Geneva University Hospital has been undertaking yearly prevalence studies in order to monitor and detect NIs. This paper describes a retrospective analysis of the results of one such study. Our goal is to identify patients with one or more NIs on the basis of clinical and other data collected during the survey. In this classification task, the main difficulty resides in the significant imbalance between positive or infected (11%) and negative (89%) cases. To cope with class imbalance, we investigate a support vector algorithm in which asymmetrical margins are tuned to improve recognition of rare positive cases. Experiments have shown this approach to be effective for the NI detection problem: we obtained a sensitivity rate of 92%, significantly better than the highest sensitivity (87%) obtained via novel resampling strat...

Machine Learning Techniques and Data Science

In hospitalized populations, there is significant heterogeneity in patient characteristics, disea... more In hospitalized populations, there is significant heterogeneity in patient characteristics, disease severity, and treatment responses, which generally translates into very different related outcomes and costs. A better understanding of this heterogeneity can lead to better management, more effective and efficient treatments by personalizing care to better meet patients' profiles. Thus, identifying distinct clinical profiles among patients can lead to more homogenous subgroups of patients. Super-utilizers (SUs) are such a group, who contribute a substantial proportion of health care costs and utilize a disproportionately high share of health care resources. This study uses cost, utilization metrics and clinical information to segment the population of patients (N=32,759) admitted to the University Hospital of Geneva in 2019 and thus identifies the characteristics of its SUs group using Latent Class Analysis. This study shows how cluster analysis might be valuable to hospitals for...

Studies in health technology and informatics, 2005

This paper addresses the model selection problem for Support Vector Machines. A hybrid genetic al... more This paper addresses the model selection problem for Support Vector Machines. A hybrid genetic algorithm guided by Direct Simplex Search to evolves hyperparameter values using an empirical error estimate as a steering criterion. This approach is specificaly tailored and experimentally evaluated on a health care problem which involves discriminating 11 % nosocomially infected patients from 89 % non infected patients. The combination of Direct Search Simplex with GAs is shown to improve the performance of GAs in terms of solution quality and computational efficiency. Unlike most other hyperparameter tuning techniques, our hybrid approach does not require supplementary effort such as computation of derivatives, making them well suited for practical purposes. This method produces encouraging results: it exhibits high performance and good convergence properties.

Task I . Our approach uses thesaural resources (from the UMLS) together with a variant of the Por... more Task I . Our approach uses thesaural resources (from the UMLS) together with a variant of the Porter stemmer for string normalization. Gene and Protein Entities (GPE) of the collection (525,938 MedLine citations) were simply marked up by dictionary look up during the indexing in order to avoid erroneous conflation: strings not found in the UMLS Specialist lexicon (augmented with various English lexical resources) were considered as GPE and were moderately overweighted. In the same spirit like other TREC competitors [23] for task I, an overweighting factor was also applied to features belonging to Medical Subject Headings (MeSH) and found in MedLine citations using a MeSH mapping tool [1]. A standard vector space IR engine with tf-idf parameters was used for indexing the Genomic collection: article’s ti tles, MeSH and RN terms, and abstact fields were selected. Best average precisions were obtained with atc.ntn (using the SMART notation) schemes: 16.71 (standard) vs. 17.02 (using UML...

Antimicrobial Resistance and Infection Control, 2013

Background Targeted screening of patients at high risk for methicillin-resistant Staphylococcus a... more Background Targeted screening of patients at high risk for methicillin-resistant Staphylococcus aureus (MRSA) carriage is an important component of MRSA control programs, which rely on prediction tools to identify those high-risk patients. Most previous risk studies reported a substantial rate of patients who are eligible for screening, but failed to be enrolled. The characteristics of these missed patients are seldom described. We aimed to determine the rate and characteristics of patients who were missed by a MRSA screening programme at our institution to see how the failure to include these patients might impact the accuracy of clinical prediction tools. Findings From March-June 2010 all patients admitted to 13 internal medicine wards at the University of Geneva Hospital (HUG) were prospectively screened for MRSA carriage. Of 1968 patients admitted to the ward, 267 patients (13.6%) failed to undergo appropriate MRSA screening. Forty-one (2.4%) screened patients were MRSA carriers...

Infection Control and Hospital Epidemiology, 2014

Objective. To test the hypothesis that methicillin-susceptible Staphylococcus aureus (MSSA) carri... more Objective. To test the hypothesis that methicillin-susceptible Staphylococcus aureus (MSSA) carriage may protect against nosocomial methicillin-resistant S. aureus (MRSA) acquisition by competing for colonization of the anterior nares. Design. Prospective cohort and nested case-control study. Setting. Swiss university hospital. Patients. All adult patients admitted to 14 wards of the general medicine division between April 1 and October 31, 2007. Methods. Patients were screened for MRSA and MSSA carriage at admission to and discharge from the division. Associations between nosocomial MRSA acquisition and MSSA colonization at admission and other confounders were analyzed by univariable and multivariable analysis. Results. Of 898 patients included, 183 (20%) were treated with antibiotics. Nosocomial MRSA acquisition occurred in 70 (8%) of the patients (case patients); 828 (92%) of the patients (control subjects) were free of MRSA colonization at discharge. MSSA carriage at admission w...

Abstract. An important problem that arises in hospitals is the monitoring and detection of nosoco... more Abstract. An important problem that arises in hospitals is the monitoring and detection of nosocomial or hospital acquired infections (NIs). This paper describes a retrospective analysis of a prevalence survey of NIs done in the Geneva University Hospital. Our goal is to identify patients with one or more NIs on the basis of clinical and other data collected during the survey. In this classification task, the main difficulty resides in the significant imbalance between positive or infected (11%) and negative (89%) cases. To remedy class imbalance, we propose a novel approach in which both oversampling of rare positives and undersampling of the non infected majority rely on synthetic cases generated via class-specific subclustering. Experiments have shown this approach to be remarkably more effective than classical random resampling methods. 1

Antimicrobial resistance and infection control, Jan 30, 2013

Targeted screening of patients at high risk for methicillin-resistant Staphylococcus aureus (MRSA... more Targeted screening of patients at high risk for methicillin-resistant Staphylococcus aureus (MRSA) carriage is an important component of MRSA control programs, which rely on prediction tools to identify those high-risk patients. Most previous risk studies reported a substantial rate of patients who are eligible for screening, but failed to be enrolled. The characteristics of these missed patients are seldom described. We aimed to determine the rate and characteristics of patients who were missed by a MRSA screening programme at our institution to see how the failure to include these patients might impact the accuracy of clinical prediction tools. From March-June 2010 all patients admitted to 13 internal medicine wards at the University of Geneva Hospital (HUG) were prospectively screened for MRSA carriage. Of 1968 patients admitted to the ward, 267 patients (13.6%) failed to undergo appropriate MRSA screening. Forty-one (2.4%) screened patients were MRSA carriers at admission. On mu...

Studies in health technology and informatics, 2011

This paper considers the model selection problem for Support Vector Machines. A well-known deriva... more This paper considers the model selection problem for Support Vector Machines. A well-known derivative Pattern Search method, which aims to tune hyperparameter values using an empirical error estimate as a steering criterion, is proposed. This approach is experimentally evaluated on a health care problem which involves discriminating nosocomially infected patients from non-infected patients. The Hooke and Jeeves Pattern Search (HJPS) method is shown to improve the results achieved by Grid Search (GS) in terms of solution quality and computational efficiency. Unlike most other parameter tuning techniques, our approach does not require supplementary effort such as computation of derivatives, making them well suited for practical purposes. This method produces encouraging results: it exhibits good performance and convergence properties.

Nosocomial infections (NIs)-those acquired in health care settingsrepresent one of the major caus... more Nosocomial infections (NIs)-those acquired in health care settingsrepresent one of the major causes of increased mortality in hospitalized patients. As they are a real problem for both patients and health authorities, the development of an effective surveillance system to monitor and detect them is of paramount importance. This paper presents a retrospective analysis of a prevalence survey of NIs done in the Geneva University Hospital. The objective is to identify patients with one or more NIs based on clinical and other data collected during the survey. In this classification task, the main difficulty lies in the significant imbalance between positive and negative cases. To overcome this problem, we investigate one-class Parzen density estimator which can be trained to differentiate two classes taking examples from a single class. The results obtained are encouraging: whereas standard 2-class SVMs scored a baseline sensitivity of 50.6% on this problem, the one-class approach increased sensitivity to as much as 88.6%. These results suggest that one-class Parzen density estimator can provide an effective and efficient way of overcoming data imbalance in classification problems.

Image management, analysis, and retrieval are currently very active research fields mainly becaus... more Image management, analysis, and retrieval are currently very active research fields mainly because of the large amount of visual data being produced in modern hospitals, and the lack of applications dealing with these data. Most often, the goal is to aid the diagnostic process. Unfortunately, only very few medical image retrieval systems are currently used in clinical routine. One application domain with a high potential for automatic image retrieval is the analysis and retrieval of lung CTs. A first user study in the United States (Purdue University) shows that these systems allow improving the diagnostic quality significantly. This article describes an approach to an aid for lung CT diagnostics. The analysis incorporates several steps and the goal is to automate the process as much as possible for easy integration into clinical processes. Thus, several automatic steps are proposed from a selection of the most characteristic slices, to an automatic segmentation of the lung tissue and a classification on the segmented area into visual observation classes. Feedback to the MD is given in the form of marked regions in the images that appear to be different from the norm of healthy tissue. We currently work on a small set of training images with marked and annotated regions but a larger set of images for the evaluation of our algorithm is in work. The article currently only contains a short quantitative evaluation. For most tasks we use existing open source software such as Weka, GIFT, and itk. This allows an easy reproduction of the search results and limits the need for costly redevelopments.

This article deals with data on nosocomial infections acquired in the Geneva University Hospitals... more This article deals with data on nosocomial infections acquired in the Geneva University Hospitals. Goal of the work is to derive a model from a hospital-acquired infection (HAI) prevalence survey of year Y and apply them to a prevalence survey of years Y+1, Y+2. This analysis permits to evaluate the effectiveness of preventive measures taken after the prevalence survey in year Y. It also analyzes the robustness of the SVM algorithm on time-variable attributes. The model build on the dataset of year Y gives better results than in a previous study. The application of the model on the Y+1 and Y+2 prevalence surveys shows simultaneously improvements and deteriorations of 5 performance measures. This highlights the effectiveness of prevention and reduces the risk of HAI after the prevalence survey of year Y. We introduce a new method to detect redundancy in a dataset with the SVM algorithm.