Artif Intell Med.

Morphometric analysis of brain images with reduced number of statistical tests: A study on the gender-related differentiation of the corpus callosum.

Publication Date: 2009 Jun 24 PMID: 19559582
Authors: Kontos, D. - Megalooikonomou, V. - Gee, J. C.
Journal: Artif Intell Med

OBJECTIVE: We evaluate the feasibility of applying dynamic recursive partitioning (DRP), an image analysis technique, to perform morphometric analysis. We apply DRP to detect and characterize discriminative morphometric characteristics between anatomical brain structures from different groups of subjects. Our method reduces the number of statistical tests, commonly required by pixel-wise statistics, alleviating the effect of the multiple comparison problem. METHODS AND MATERIALS: The main idea of DRP is to partition the two-dimensional (2D) image adaptively into progressively smaller subregions until statistically significant discriminative regions are detected. The partitioning process is guided by statistical tests applied on groups of pixels. By performing statistical tests on groups of pixels rather than on individual pixels, the number of statistical tests is effectively reduced. This reduction of statistical tests restricts the effect of the multiple comparison problem (i.e., type-I error). We demonstrate an application of DRP for detecting gender-related morphometric differentiation of the corpus callosum. DRP was applied to template deformation fields computed from registered magnetic resonance images of the corpus callosum to detect regions of significant expansion or contraction between female and male subjects. RESULTS: DRP was able to detect regions comparable to those of pixel-wise analysis, while reducing the number of required statistical tests up to almost 50%. The detected regions were in agreement with findings previously reported in the literature. Statistically significant discriminative morphological variability was detected in the posterior corpus callosum region, the isthmus and the anterior corpus callosum. In addition, by operating on groups of pixels, DRP appears to be less prone to detecting spatially diffused and isolated outlier pixels as significant. CONCLUSION: DRP can be a viable approach for detecting discriminative morphometric characteristics among groups of subjects, having the potential to alleviate the multiple comparisons' effect by significantly reducing the number of required statistical tests.

post to: CiteULike

An intelligent model for liver disease diagnosis.

Publication Date: 2009 Jun 18 PMID: 19540738
Authors: Lin, R. H.
Journal: Artif Intell Med

Objectives: Liver disease, the most common disease in Taiwan, is not easily discovered in its initial stage; early diagnosis of this leading cause of mortality is therefore highly important. The design of an effective diagnosis model is therefore an important issue in liver disease treatment. This study accordingly employs classification and regression tree (CART) and case-based reasoning (CBR) techniques to structure an intelligent diagnosis model aiming to provide a comprehensive analytic framework to raise the accuracy of liver disease diagnosis. Methods: Based on the advice and assistance of doctors and medical specialists of liver conditions, 510 outpatient visitors using ICD-9 (International Classification of Diseases, 9th Revision) codes at a medical center in Taiwan from 2005 to 2006 were selected as the cases in the data set for liver disease diagnosis. Data on 340 patients was utilized for the development of the model and on 170 patients utilized to perform comparative analysis of the models. This paper accordingly suggests an intelligent model for the diagnosis of liver diseases which integrates CART and CBR. The major steps in applying the model include: (1) adopting CART to diagnose whether a patient suffers from liver disease; (2) for patients diagnosed with liver disease in the first step, employing CBR to diagnose the types of liver diseases. Results: In the first phase, CART is used to extract rules from health examination data to show whether the patient suffers from liver disease. The results indicate that the CART rate of accuracy is 92.94%. In the second phase, CBR is developed to diagnose the type of liver disease, and the new case triggers the CBR system to retrieve the most similar case from the case base in order to support the treatment of liver disease. The new case is supported by a similarity ratio, and the CBR diagnostic accuracy rate is 90.00%. Actual implementation shows that the intelligent diagnosis model is capable of integrating CART and CBR techniques to examine liver diseases with considerable accuracy. The model can be used as a supporting system in making decisions regarding liver disease diagnosis and treatment. The rules extracted from CART are helpful to physicians in diagnosing liver diseases. CBR can retrieve the most similar case from the case base in order to solve a new liver disease problem and can be of great assistance to physicians in identifying the type of liver disease, reducing diagnostic errors and improving the quality and effectiveness of medical treatment.

post to: CiteULike

An interpretable fuzzy rule-based classification methodology for medical diagnosis.

Publication Date: 2009 Jun 17 PMID: 19540096
Authors: Gadaras, I. - Mikhailov, L.
Journal: Artif Intell Med

OBJECTIVE: The aim of this paper is to present a novel fuzzy classification framework for the automatic extraction of fuzzy rules from labeled numerical data, for the development of efficient medical diagnosis systems. METHODS AND MATERIALS: The proposed methodology focuses on the accuracy and interpretability of the generated knowledge that is produced by an iterative, flexible and meaningful input partitioning mechanism. The generated hierarchical fuzzy rule structure is composed by linguistic; multiple consequent fuzzy rules that considerably affect the model comprehensibility. RESULTS AND CONCLUSION: The performance of the proposed method is tested on three medical pattern classification problems and the obtained results are compared against other existing methods. It is shown that the proposed variable input partitioning leads to a flexible decision making framework and fairly accurate results with a small number of rules and a simple, fast and robust training process.

post to: CiteULike

A comparison of three computational modelling methods for the prediction of virological response to combination HIV therapy.

Publication Date: 2009 Jun 11 PMID: 19524413
Authors: Wang, D. - Larder, B. - Revell, A. - Montaner, J. - Harrigan, R. - De Wolf, F. - Lange, J. - Wegner, S. - Ruiz, L. - Perez-Elias, M. J. - Emery, S. - Gatell, J. - D'Arminio Monforte, A. - Torti, C. - Zazzi, M. - Lane, C.
Journal: Artif Intell Med

OBJECTIVE: HIV treatment failure is commonly associated with drug resistance and the selection of a new regimen is often guided by genotypic resistance testing. The interpretation of complex genotypic data poses a major challenge. We have developed artificial neural network (ANN) models that predict virological response to therapy from HIV genotype and other clinical information. Here we compare the accuracy of ANN with alternative modelling methodologies, random forests (RF) and support vector machines (SVM). METHODS: Data from 1204 treatment change episodes (TCEs) were identified from the HIV Resistance Response Database Initiative (RDI) database and partitioned at random into a training set of 1154 and a test set of 50. The training set was then partitioned using an L-cross (L=10 in this study) validation scheme for training individual computational models. Seventy six input variables were used for training the models: 55 baseline genotype mutations; the 14 potential drugs in the new treatment regimen; four treatment history variables; baseline viral load; CD4 count and time to follow-up viral load. The output variable was follow-up viral load. Performance was evaluated in terms of the correlations and absolute differences between the individual models' predictions and the actual DeltaVL values. RESULTS: The correlations (r(2)) between predicted and actual DeltaVL varied from 0.318 to 0.546 for ANN, 0.590 to 0.751 for RF and 0.300 to 0.720 for SVM. The mean absolute differences varied from 0.677 to 0.903 for ANN, 0.494 to 0.644 for RF and 0.500 to 0.790 for SVM. ANN models were significantly inferior to RF and SVM models. The predictions of the ANN, RF and SVM committees all correlated highly significantly with the actual DeltaVL of the independent test TCEs, producing r(2) values of 0.689, 0.707 and 0.620, respectively. The mean absolute differences were 0.543, 0.600 and 0.607log(10)copies/ml for ANN, RF and SVM, respectively. There were no statistically significant differences between the three committees. Combining the committees' outputs improved correlations between predicted and actual virological responses. The combination of all three committees gave a correlation of r(2)=0.728. The mean absolute differences followed a similar pattern. CONCLUSIONS: RF and SVM models can produce predictions of virological response to HIV treatment that are comparable in accuracy to a committee of ANN models. Combining the predictions of different models improves their accuracy somewhat. This approach has potential as a future clinical tool and a combination of ANN and RF models is being taken forward for clinical evaluation.

post to: CiteULike

Application of constrained independent component analysis algorithms in electrocardiogram arrhythmias.

Publication Date: 2009 Jun 8 PMID: 19515541
Authors: Llinares, R. - Igual, J.
Journal: Artif Intell Med

OBJECTIVES: The extraction of the atrial activity in atrial fibrillation episodes is a must for clinical purposes. During atrial fibrillation arrhythmia, the independent atrial and ventricular signals are superposed in the electrocardiogram, fulfilling the independent component analysis (ICA) model. We propose three new algorithms that constrain the classical ICA solution to fit the spectral content of the atrial component. This constraint allows the statement of the problem in terms of semiblind source extraction instead of blind source separation (BSS), in the sense that we only recover one source and we exploit the prior information about the sources in the extraction process. METHODS AND MATERIALS: The methods used are extensions of classical BSS methods based on second and higher order statistics. We exploit the prior assumption about the sources in order to obtain the source extraction algorithms that are focused on the extraction of the atrial component. The material corresponds to 10 synthetic recordings in order to measure and compare the quality of the different algorithms and 66 real recordings coming from two different databases, one public database from Physionet and one database from the Clinical University Hospital, Valencia, Spain. RESULTS: We have analyzed the performance of the three new algorithms and compared it with the performance of the traditional ICA algorithms. In the case of the synthetic data, it is possible to obtain the mean square error, so the comparison is easier. The new methods outperform the non-constrained versions in addition to simplifying the solution, since they do not need to recover all the components in order to estimate the atrial activity, i.e., the new methods are focused on the extraction of the atrial activity, so the extraction is stopped after the atrial signal is recovered. CONCLUSIONS: We have shown that the ICA only version of the algorithms can be improved and adapted to fulfill the prior information about the characteristics of the atrial activity. This modification allows us to obtain new algorithms that have the following advantages compared to ICA only based solutions: they exploit prior information during the extraction, not in the postprocessing identification of the atrial signal; they extract only the interesting clinical signal instead of all the components; they outperform the ICA only version of the algorithm, improving the estimation of the atrial signal.

post to: CiteULike

Posterior probability profiles for the automated assessment of the recovery of patients with stroke from activity of daily living tasks.

Publication Date: 2009 Jul PMID: 19409768
Authors: Van Dijck, G. - Van Vaerenbergh, J. - Van Hulle, M. M.
Journal: Artif Intell Med

OBJECTIVE: Assessing recovery after stroke has been so far a time consuming procedure in which trained clinicians are required. A demand for automated assessment techniques arises due to the increasing number of patients with stroke and the continuous growth of new treatment options. In this study, we investigate the applicability of isometric force and torque measurements in activity of daily living tasks to assess the functional recovery after stroke in an automated way. METHODS AND MATERIALS: A new hybrid filter-wrapper feature subset technology was developed for a new mechatronic platform with the aim to identify the most important features and sensors that can distinguish normal controls from patients with stroke. We compared 3 different classification algorithms to make the distinction: k-nearest neighbors, kernel density estimation and least-squares support vector machines. Based on isometric force and torque measurements obtained from 16 patients with a first-ever ischemic or haemorrhagic stroke within the middle cerebral artery territory, we computed for each subject the probability to belong to the class of normal subjects. These probabilities were computed during a period of 6 months post-stroke to quantify the level of recovery during this period. The posterior probabilities were validated by means of a correlation study with the Lindmark modified Fugl-Meyer assessment. RESULTS: Patients with stroke and normal controls could be distinguished with an accuracy of 98.25% by means of kernel density estimation. The posterior probability profiles had a correlation of 76.6% and 80.29% with the global score of the Lindmark modified Fugl-Meyer scale and 'part A', the upper extremity subscore, respectively. This degree of correlation was as high as obtained with supervised scoring techniques such as the Barthel index. CONCLUSION: This study shows that the assessment of recovery after stroke can be automated by means of posterior probability profiles due to their high correlation with the Fugl-Meyer assessment. The posterior probability profiles confirm the importance of a recovery within the first weeks after stroke to obtain a higher recovery plateau compared to later changes in recovery.

post to: CiteULike

Prediction of periventricular leukomalacia. Part II: Selection of hemodynamic features using computational intelligence.

Publication Date: 2009 Jul PMID: 19162456
Authors: Samanta, B. - Bird, G. L. - Kuijpers, M. - Zimmerman, R. A. - Jarvik, G. P. - Wernovsky, G. - Clancy, R. R. - Licht, D. J. - Gaynor, J. W. - Nataraj, C.
Journal: Artif Intell Med

OBJECTIVE: The objective of Part II is to analyze the dataset of extracted hemodynamic features (Case 3 of Part I) through computational intelligence (CI) techniques for identification of potential prognostic factors for periventricular leukomalacia (PVL) occurrence in neonates with congenital heart disease. METHODS: The extracted features (Case 3 dataset of Part I) were used as inputs to CI based classifiers, namely, multi-layer perceptron (MLP) and probabilistic neural network (PNN) in combination with genetic algorithms (GA) for selection of the most suitable features predicting the occurrence of PVL. The selected features were next used as inputs to a decision tree (DT) algorithm for generating easily interpretable rules of PVL prediction. RESULTS: Prediction performance for two CI based classifiers, MLP and PNN coupled with GA are presented for different number of selected features. The best prediction performances were achieved with 6 and 7 selected features. The prediction success was 100% in training and the best ranges of sensitivity (SN), specificity (SP) and accuracy (AC) in test were 60-73%, 74-84% and 71-74%, respectively. The identified features when used with the DT algorithm gave best SN, SP and AC in the ranges of 87-90% in training and 80-87%, 74-79% and 79-82% in test. Among the variables selected in CI, systolic and diastolic blood pressures, and pCO(2) figured prominently similar to Part I. Decision tree based rules for prediction of PVL occurrence were obtained using the CI selected features. CONCLUSIONS: The proposed approach combines the generalization capability of CI based feature selection approach and generation of easily interpretable classification rules of the decision tree. The combination of CI techniques with DT gave substantially better test prediction performance than using CI and DT separately.

post to: CiteULike

Prediction of periventricular leukomalacia. Part I: Selection of hemodynamic features using logistic regression and decision tree algorithms.

Publication Date: 2009 Jul PMID: 19162455
Authors: Samanta, B. - Bird, G. L. - Kuijpers, M. - Zimmerman, R. A. - Jarvik, G. P. - Wernovsky, G. - Clancy, R. R. - Licht, D. J. - Gaynor, J. W. - Nataraj, C.
Journal: Artif Intell Med

OBJECTIVE: Periventricular leukomalacia (PVL) is part of a spectrum of cerebral white matter injury which is associated with adverse neurodevelopmental outcome in preterm infants. While PVL is common in neonates with cardiac disease, both before and after surgery, it is less common in older infants with cardiac disease. Pre-, intra-, and postoperative risk factors for the occurrence of PVL are poorly understood. The main objective of the present work is to identify potential hemodynamic risk factors for PVL occurrence in neonates with complex heart disease using logistic regression analysis and decision tree algorithms. METHODS: The postoperative hemodynamic and arterial blood gas data (monitoring variables) collected in the cardiac intensive care unit of Children's Hospital of Philadelphia were used for predicting the occurrence of PVL. Three categories of datasets for 103 infants and neonates were used-(1) original data without any preprocessing, (2) partial data keeping the admission, the maximum and the minimum values of the monitoring variables, and (3) extracted dataset of statistical features. The datasets were used as inputs for forward stepwise logistic regression to select the most significant variables as predictors. The selected features were then used as inputs to the decision tree induction algorithm for generating easily interpretable rules for prediction of PVL. RESULTS: Three sets of data were analyzed in SPSS for identifying statistically significant predictors (p<0.05) of PVL through stepwise logistic regression and their correlations. The classification success of the Case 3 dataset of extracted statistical features was best with sensitivity (SN), specificity (SP) and accuracy (AC) of 87, 88 and 87%, respectively. The identified features, when used with decision tree algorithms, gave SN, SP and AC of 90, 97 and 94% in training and 73, 58 and 65% in test. The identified variables in Case 3 dataset mainly included blood pressure, both systolic and diastolic, partial pressures pO(2) and pCO(2), and their statistical features like average, variance, skewness (a measure of asymmetry) and kurtosis (a measure of abrupt changes). Rules for prediction of PVL were generated automatically through the decision tree algorithms. CONCLUSIONS: The proposed approach combines the advantages of statistical approach (regression analysis) and data mining techniques (decision tree) for generation of easily interpretable rules for PVL prediction. The present work extends an earlier research [Galli KK, Zimmerman RA, Jarvik GP, Wernovsky G, Kuijpers M, Clancy RR, et al. Periventricular leukomalacia is common after cardiac surgery. J Thorac Cardiovasc Surg 2004;127:692-704] in the form of expanding the feature set, identifying additional prognostic factors (namely pCO(2)) emphasizing the temporal variations in addition to upper or lower values, and generating decision rules. The Case 3 dataset was further investigated in Part II for feature selection through computational intelligence.

post to: CiteULike

Model of experts for decision support in the diagnosis of leukemia patients.

Publication Date: 2009 Jul PMID: 19135343
Authors: Corchado, J. M. - De Paz, J. F. - Rodriguez, S. - Bajo, J.
Journal: Artif Intell Med

OBJECTIVE: Recent advances in the field of biomedicine, specifically in the field of genomics, have led to an increase in the information available for conducting expression analysis. Expression analysis is a technique used in transcriptomics, a branch of genomics that deals with the study of messenger ribonucleic acid (mRNA) and the extraction of information contained in the genes. This increase in information is reflected in the exon arrays, which require the use of new techniques in order to extract the information. The purpose of this study is to provide a tool based on a mixture of experts model that allows the analysis of the information contained in the exon arrays, from which automatic classifications for decision support in diagnoses of leukemia patients can be made. The proposed model integrates several cooperative algorithms characterized for their efficiency for data processing, filtering, classification and knowledge extraction. The Cancer Institute of the University of Salamanca is making an effort to develop tools to automate the evaluation of data and to facilitate de analysis of information. This proposal is a step forward in this direction and the first step toward the development of a mixture of experts tool that integrates different cognitive and statistical approaches to deal with the analysis of exon arrays. The mixture of experts model presented within this work provides great capacities for learning and adaptation to the characteristics of the problem in consideration, using novel algorithms in each of the stages of the analysis process that can be easily configured and combined, and provides results that notably improve those provided by the existing methods for exon arrays analysis. MATERIAL AND METHODS: The material used consists of data from exon arrays provided by the Cancer Institute that contain samples from leukemia patients. The methodology used consists of a system based on a mixture of experts. Each one of the experts incorporates novel artificial intelligence techniques that improve the process of carrying out various tasks such as pre-processing, filtering, classification and extraction of knowledge. This article will detail the manner in which individual experts are combined so that together they generate a system capable of extracting knowledge, thus permitting patients to be classified in an automatic and efficient manner that is also comprehensible for medical personnel. RESULTS AND CONCLUSION: The system has been tested in a real setting and has been used for classifying patients who suffer from different forms of leukemia at various stages. Personnel from the Cancer Institute supervised and participated throughout the testing period. Preliminary results are promising, notably improving the results obtained with previously used tools. The medical staff from the Cancer Institute considers the tools that have been developed to be positive and very useful in a supporting capacity for carrying out their daily tasks. Additionally the mixture of experts supplies a tool for the extraction of necessary information in order to explain the associations that have been made in simple terms. That is, it permits the extraction of knowledge for each classification made and generalized in order to be used in subsequent classifications. This allows for a large amount of learning and adaptation within the proposed system.

post to: CiteULike

Modelling treatment effects in a clinical Bayesian network using Boolean threshold functions.

Publication Date: 2009 Jul PMID: 19111448
Authors: Visscher, S. - Lucas, P. J. - Schurink, C. A. - Bonten, M. J.
Journal: Artif Intell Med

OBJECTIVE: Appropriate antimicrobial treatment of infections in critically ill patients should be started as soon as possible, as delay in treatment may reduce a patient's prognostic outlook considerably. Ventilator-associated pneumonia (VAP) occurs in patients in intensive care units who are mechanically ventilated and is almost always preceded by colonisation of the respiratory tract by the causative microorganisms. It is very difficult to clinically diagnose VAP and, therefore, some form of computer-based decision support might be helpful for the clinician. MATERIALS AND METHODS: As diagnosing and treating VAP involves reasoning with uncertainty, we have used a Bayesian network as the primary tool for building a decision-support system. The effects of usage of antibiotics on the colonisation of the respiratory tract by various pathogens and the subsequent antibiotic choices in case of VAP were modelled using the notion of causal independence. In particular, the conditional probability distribution of the random variable that represents the overall coverage of pathogens by antibiotics was modelled in terms of the conjunctive effect of the seven different pathogens, usually referred to as the noisy-AND model. In this paper, we investigate different coverage models, as well as generalisations of the noisy-AND, called noisy-threshold models, and test them on clinical data of intensive care unit (ICU) patients who are mechanically ventilated. RESULTS: Some of the constructed noisy-threshold models offered further improvement of the performance of the Bayesian network in covering present causative pathogens by advising appropriate antimicrobial treatment. CONCLUSIONS: By reconsidering the modelling of interactions between the random variables in a Bayesian network using the theory of causal independence, it is possible to refine its performance. This was clearly shown for our Bayesian network concerning VAP, indicating that only specific noisy-threshold models might be appropriate for the modelling of the interaction between pathogens and antimicrobial treatment with respect to susceptibility. The results obtained also provide evidence that the noisy-OR and noisy-AND might not always be the best functions to model interactions among random variables.

post to: CiteULike

A model-free ensemble method for class prediction with application to biomedical decision making.

Publication Date: 2009 Jul PMID: 19081231
Authors: Kodell, R. L. - Pearce, B. A. - Baek, S. - Moon, H. - Ahn, H. - Young, J. F. - Chen, J. J.
Journal: Artif Intell Med

OBJECTIVE: A classification algorithm that utilizes two-dimensional convex hulls of training-set samples is presented. METHODS AND MATERIAL: For each pair of predictor variables, separate convex hulls of positive and negative samples in the training set are formed, and these convex hulls are used to classify test points according to a nearest-neighbor criterion. An ensemble of these two-dimensional convex-hull classifiers is formed by trimming the (m)C(2) possible classifiers derived from the m predictors to a set of classifiers comprised of only unique predictor variables. Because only two-dimensional spaces are required to be populated by training-set samples, the "curse of dimensionality" is not an issue. At the same time, the power of ensemble voting is exploited by combining the classifications of the unique two-dimensional classifiers to reach a final classification. RESULTS: The algorithm is illustrated by application to three publicly available biomedical data sets with genomic predictors and is shown to have prediction accuracy that is competitive with a number of published classification procedures. CONCLUSION: Because of its superior performance in terms of sensitivity and negative predictive value compared to its competitors, the convex-hull ensemble classifier demonstrates good potential for medical screening, where often the major emphasis is placed on having reliable negative predictions.

post to: CiteULike