Data mining framework for fatty liver disease classification in ultrasound: a hybrid feature extraction paradigm

PURPOSE: Fatty liver disease (FLD) is an increasing prevalent disease that can be reversed if detected early. Ultrasound is the safest and ubiquitous method for identifying FLD. Since expert sonographers are required to accurately interpret the liver ultrasound images, lack of the same will result in interobserver variability. For more objective interpretation, high accuracy, and quick second opinions, computer aided diagnostic (CAD) techniques may be exploited. The purpose of this work is to develop one such CAD technique for accurate classification of normal livers and abnormal livers affected by FLD. METHODS: In this paper, the authors present a CAD technique (called Symtosis) that uses a novel combination of significant features based on the texture, wavelet transform, and higher order spectra of the liver ultrasound images in various supervised learning-based classifiers in order to determine parameters that classify normal and FLD-affected abnormal livers. RESULTS: On evaluating the proposed technique on a database of 58 abnormal and 42 normal liver ultrasound images, the authors were able to achieve a high classification accuracy of 93.3% using the decision tree classifier. CONCLUSIONS: This high accuracy added to the completely automated classification procedure makes the authors' proposed technique highly suitable for clinical deployment and usage.

The usefulness of ultrasound in the classification of chronic liver disease

Chronic Liver Disease is a progressive, most of the time asymptomatic, and potentially fatal disease. In this paper, a semi-automatic procedure to stage this disease is proposed based on ultrasound liver images, clinical and laboratorial data. In the core of the algorithm two classifiers are used: a k nearest neighbor and a Support Vector Machine, with different kernels. The classifiers were trained with the proposed multi-modal feature set and the results obtained were compared with the laboratorial and clinical feature set. The results showed that using ultrasound based features, in association with laboratorial and clinical features, improve the classification accuracy. The support vector machine, polynomial kernel, outperformed the others classifiers in every class studied. For the Normal class we achieved 100% accuracy, for the chronic hepatitis with cirrhosis 73.08%, for compensated cirrhosis 59.26% and for decompensated cirrhosis 91.67%.

Chronic liver disease staging classification based on ultrasound, clinical and laboratorial data

In this work the identification and diagnosis of various stages of chronic liver disease is addressed. The classification results of a support vector machine, a decision tree and a k-nearest neighbor classifier are compared. Ultrasound image intensity and textural features are jointly used with clinical and laboratorial data in the staging process. The classifiers training is performed by using a population of 97 patients at six different stages of chronic liver disease and a leave-one-out cross-validation strategy. The best results are obtained using the support vector machine with a radial-basis kernel, with 73.20% of overall accuracy. The good performance of the method is a promising indicator that it can be used, in a non invasive way, to provide reliable information about the chronic liver disease staging.

Diffuse liver disease classification from ultrasound surface characterization, clinical and laboratorial data

In this work liver contour is semi-automatically segmented and quantified in order to help the identification and diagnosis of diffuse liver disease. The features extracted from the liver contour are jointly used with clinical and laboratorial data in the staging process. The classification results of a support vector machine, a Bayesian and a k-nearest neighbor classifier are compared. A population of 88 patients at five different stages of diffuse liver disease and a leave-one-out cross-validation strategy are used in the classification process. The best results are obtained using the k-nearest neighbor classifier, with an overall accuracy of 80.68%. The good performance of the proposed method shows a reliable indicator that can improve the information in the staging of diffuse liver disease.

Fatty liver characterization and classification by ultrasound

Steatosis, also known as fatty liver, corresponds to an abnormal retention of lipids within the hepatic cells and reflects an impairment of the normal processes of synthesis and elimination of fat. Several causes may lead to this condition, namely obesity, diabetes, or alcoholism. In this paper an automatic classification algorithm is proposed for the diagnosis of the liver steatosis from ultrasound images. The features are selected in order to catch the same characteristics used by the physicians in the diagnosis of the disease based on visual inspection of the ultrasound images. The algorithm, designed in a Bayesian framework, computes two images: i) a despeckled one, containing the anatomic and echogenic information of the liver, and ii) an image containing only the speckle used to compute the textural features. These images are computed from the estimated RF signal generated by the ultrasound probe where the dynamic range compression performed by the equipment is taken into account. A Bayes classifier, trained with data manually classified by expert clinicians and used as ground truth, reaches an overall accuracy of 95% and a 100% of sensitivity. The main novelties of the method are the estimations of the RF and speckle images which make it possible to accurately compute textural features of the liver parenchyma relevant for the diagnosis.

Comparing the content of instruments assessing environmental factors using the International Classification of Functioning, Disability and Health

Purpose: To describe and compare the content of instruments that assess environmental factors using the International Classification of Functioning, Disability and Health (ICF). Methods: A systematic search of PubMed, CINAHL and PEDro databases was conducted using a pre-determined search strategy. The identified instruments were screened independently by two investigators, and meaningful concepts were linked to the most precise ICF category according to published linking rules. Results: Six instruments were included, containing 526 meaningful concepts. Instruments had between 20% and 98% of items linked to categories in Chapter 1. The highest percentage of items from one instrument linked to categories in Chapters 2–5 varied between 9% and 50%. The presence or absence of environmental factors in a specific context is assessed in 3 instruments, while the other 3 assess the intensity of the impact of environmental factors. Discussion: Instruments differ in their content, type of assessment, and have several items linked to the same ICF category. Most instruments primarily assess products and technology (Chapter 1), highlighting the need to deepen the discussion on the theory that supports the measurement of environmental factors. This discussion should be thorough and lead to the development of methodologies and new tools that capture the underlying concepts of the ICF.

Comorbidity adjustment index for the international classification of diseases, 10

OBJECTIVE: To develop a Charlson-like comorbidity index based on clinical conditions and weights of the original Charlson comorbidity index. METHODS: Clinical conditions and weights were adapted from the International Classification of Diseases, 10th revision and applied to a single hospital admission diagnosis. The study included 3,733 patients over 18 years of age who were admitted to a public general hospital in the city of Rio de Janeiro, southeast Brazil, between Jan 2001 and Jan 2003. The index distribution was analyzed by gender, type of admission, blood transfusion, intensive care unit admission, age and length of hospital stay. Two logistic regression models were developed to predict in-hospital mortality including: a) the aforementioned variables and the risk-adjustment index (full model); and b) the risk-adjustment index and patient's age (reduced model). RESULTS: Of all patients analyzed, 22.3% had risk scores >1, and their mortality rate was 4.5% (66.0% of them had scores >1). Except for gender and type of admission, all variables were retained in the logistic regression. The models including the developed risk index had an area under the receiver operating characteristic curve of 0.86 (full model), and 0.76 (reduced model). Each unit increase in the risk score was associated with nearly 50% increase in the odds of in-hospital death. CONCLUSIONS: The risk index developed was able to effectively discriminate the odds of in-hospital death which can be useful when limited information is available from hospital databases.

Towards the detection of deception in interactive multimedia environments

A classical application of biosignal analysis has been the psychophysiological detection of deception, also known as the polygraph test, which is currently a part of standard practices of law enforcement agencies and several other institutions worldwide. Although its validity is far from gathering consensus, the underlying psychophysiological principles are still an interesting add-on for more informal applications. In this paper we present an experimental off-the-person hardware setup, propose a set of feature extraction criteria and provide a comparison of two classification approaches, targeting the detection of deception in the context of a role-playing interactive multimedia environment. Our work is primarily targeted at recreational use in the context of a science exhibition, where the main goal is to present basic concepts related with knowledge discovery, biosignal analysis and psychophysiology in an educational way, using techniques that are simple enough to be understood by children of different ages. Nonetheless, this setting will also allow us to build a significant data corpus, annotated with ground-truth information, and collected with non-intrusive sensors, enabling more advanced research on the topic. Experimental results have shown interesting findings and provided useful guidelines for future work. Pattern Recognition

Multi-criteria design optimization study of solvent extraction in mixer–settler units

Solvent extraction is considered as a multi-criteria optimization problem, since several chemical species with similar extraction kinetic properties are frequently present in the aqueous phase and the selective extraction is not practicable. This optimization, applied to mixer–settler units, considers the best parameters and operating conditions, as well as the best structure or process flow-sheet. Global process optimization is performed for a specific flow-sheet and a comparison of Pareto curves for different flow-sheets is made. The positive weight sum approach linked to the sequential quadratic programming method is used to obtain the Pareto set. In all investigated structures, recovery increases with hold-up, residence time and agitation speed, while the purity has an opposite behaviour. For the same treatment capacity, counter-current arrangements are shown to promote recovery without significant impairment in purity. Recycling the aqueous phase is shown to be irrelevant, but organic recycling with as many stages as economically feasible clearly improves the design criteria and reduces the most efficient organic flow-rate.

Espetroscopia (1H) por ressonância magnética do disco intervertebral lombar no adulto e sua aplicação na rotina imagiológica

Objetivos – Demonstrar o potencial da espetroscopia (1H) por ressonância magnética na doença degenerativa discal lombar e defender a integração desta técnica na rotina clínico‑imagiológica para a precisa classificação da involução vs degenerescência dos discos L4‑L5 e L5‑S1 em doentes com lombalgia não relacionável com causa mecânica. Material e métodos – O estudo incluiu 102 discos intervertebrais lombares de 123 doentes. Foram estudados 61 discos de L4‑L5, 41 discos de L5‑S1 e 34 discos de D12‑L1. Utilizou‑se um sistema de ressonância magnética de 1,5 T e técnica monovoxel. Obtiveram‑se os rácios [Lac/Nacetyl] e [Nacetyl/(Lac+Lípidos)] e aplicou‑se a ressonância de lípidos para avaliar a bioquímica do disco com o fim de conhecer o estado de involução vs degenerescência que o suscetibilizam para a instabilidade e sobrecarga. Avaliou‑se o comportamento dos rácios e do teor lipídico dos discos L4‑L5‑S1 e as diferenças apresentadas em relação a D12‑L1. Foi também realizada a comparação entre os discos L4‑L5, L5‑S1 e D12‑L1 na ponderação T2 (T2W), segundo a classificação ajustada (1‑4) de Pfirrmann. Resultados – Verificou‑se que os rácios e o valor dos lípidos dos discos L4‑L5‑S1 apresentaram diferenças estatisticamente significativas quando relacionados com os discos D12‑L1. O rácio [Lac/Nacetyl] em L4‑L5‑S1 mostrou‑se aumentado em relação a D12‑L1 (p=0,033 para os discos com grau de involução [1+2] e p=0,004 para os discos com grau [3+4]). Estes resultados sugerem que a involução vs degenerescência dos discos nos graus mais elevados condiciona um decréscimo do pico do Lactato. O rácio [Nacetyl/(Lac+Lip)] discrimina os graus de involução [1+2] do [3+4] no nível L4‑L5, apresentando os valores dos rácios (média 0,65 e 0,5 respetivamente com p=0,04). O rácio médio de [Nacetyl/(Lac+Lip)] dos discos L4‑L5 foi 1,8 vezes mais elevado do que em D12‑L1. O espetro lipídico em L4‑L5‑S1 nos graus mais elevados não mostrou ter uma prevalência constante quanto às frequências de ressonância. Conclusão – A espetroscopia (1H) dos discos intervertebrais poderá ter aplicação na discriminação dos graus de involução vs degenerescência e representar um contributo semiológico importante em suplemento à ponderação T2 convencional. As ressonâncias de lípidos dos discos L4‑L5 e L5‑S1, involuídos ou degenerados, devem ser avaliadas em relação a D12‑L1, utilizando este valor como referência, pois este último é o nível considerado estável e com baixa probabilidade de degenerescência.

Determining the number of clusters in categorical data

Cluster analysis for categorical data has been an active area of research. A well-known problem in this area is the determination of the number of clusters, which is unknown and must be inferred from the data. In order to estimate the number of clusters, one often resorts to information criteria, such as BIC (Bayesian information criterion), MML (minimum message length, proposed by Wallace and Boulton, 1968), and ICL (integrated classification likelihood). In this work, we adopt the approach developed by Figueiredo and Jain (2002) for clustering continuous data. They use an MML criterion to select the number of clusters and a variant of the EM algorithm to estimate the model parameters. This EM variant seamlessly integrates model estimation and selection in a single algorithm. For clustering categorical data, we assume a finite mixture of multinomial distributions and implement a new EM algorithm, following a previous version (Silvestre et al., 2008). Results obtained with synthetic datasets are encouraging. The main advantage of the proposed approach, when compared to the above referred criteria, is the speed of execution, which is especially relevant when dealing with large data sets.

Risk criteria in occupational environments: critical overview and discussion

The current work can be seen as a starting point for the discussion of the problematic on risk acceptance criteria in occupational environments. Some obstacles to the quantitative acceptance criteria formulation and use were analyzed. A look to the long tradition of major hazards accidents was also performed. This work shows that organizations can have several difficulties in acceptance criteria formulation and that the use of pre-defined acceptance criteria in risk assessment methodologies can be inadequate in some cases. It is urgent to define guidelines that can help organizations in the formulation of risk acceptance criteria for occupational environments.

Automatic learning for the classification of chemical reactions and in statistical thermodynamics

This Thesis describes the application of automatic learning methods for a) the classification of organic and metabolic reactions, and b) the mapping of Potential Energy Surfaces(PES). The classification of reactions was approached with two distinct methodologies: a representation of chemical reactions based on NMR data, and a representation of chemical reactions from the reaction equation based on the physico-chemical and topological features of chemical bonds. NMR-based classification of photochemical and enzymatic reactions. Photochemical and metabolic reactions were classified by Kohonen Self-Organizing Maps (Kohonen SOMs) and Random Forests (RFs) taking as input the difference between the 1H NMR spectra of the products and the reactants. The development of such a representation can be applied in automatic analysis of changes in the 1H NMR spectrum of a mixture and their interpretation in terms of the chemical reactions taking place. Examples of possible applications are the monitoring of reaction processes, evaluation of the stability of chemicals, or even the interpretation of metabonomic data. A Kohonen SOM trained with a data set of metabolic reactions catalysed by transferases was able to correctly classify 75% of an independent test set in terms of the EC number subclass. Random Forests improved the correct predictions to 79%. With photochemical reactions classified into 7 groups, an independent test set was classified with 86-93% accuracy. The data set of photochemical reactions was also used to simulate mixtures with two reactions occurring simultaneously. Kohonen SOMs and Feed-Forward Neural Networks (FFNNs) were trained to classify the reactions occurring in a mixture based on the 1H NMR spectra of the products and reactants. Kohonen SOMs allowed the correct assignment of 53-63% of the mixtures (in a test set). Counter-Propagation Neural Networks (CPNNs) gave origin to similar results. The use of supervised learning techniques allowed an improvement in the results. They were improved to 77% of correct assignments when an ensemble of ten FFNNs were used and to 80% when Random Forests were used. This study was performed with NMR data simulated from the molecular structure by the SPINUS program. In the design of one test set, simulated data was combined with experimental data. The results support the proposal of linking databases of chemical reactions to experimental or simulated NMR data for automatic classification of reactions and mixtures of reactions. Genome-scale classification of enzymatic reactions from their reaction equation. The MOLMAP descriptor relies on a Kohonen SOM that defines types of bonds on the basis of their physico-chemical and topological properties. The MOLMAP descriptor of a molecule represents the types of bonds available in that molecule. The MOLMAP descriptor of a reaction is defined as the difference between the MOLMAPs of the products and the reactants, and numerically encodes the pattern of bonds that are broken, changed, and made during a chemical reaction. The automatic perception of chemical similarities between metabolic reactions is required for a variety of applications ranging from the computer validation of classification systems, genome-scale reconstruction (or comparison) of metabolic pathways, to the classification of enzymatic mechanisms. Catalytic functions of proteins are generally described by the EC numbers that are simultaneously employed as identifiers of reactions, enzymes, and enzyme genes, thus linking metabolic and genomic information. Different methods should be available to automatically compare metabolic reactions and for the automatic assignment of EC numbers to reactions still not officially classified. In this study, the genome-scale data set of enzymatic reactions available in the KEGG database was encoded by the MOLMAP descriptors, and was submitted to Kohonen SOMs to compare the resulting map with the official EC number classification, to explore the possibility of predicting EC numbers from the reaction equation, and to assess the internal consistency of the EC classification at the class level. A general agreement with the EC classification was observed, i.e. a relationship between the similarity of MOLMAPs and the similarity of EC numbers. At the same time, MOLMAPs were able to discriminate between EC sub-subclasses. EC numbers could be assigned at the class, subclass, and sub-subclass levels with accuracies up to 92%, 80%, and 70% for independent test sets. The correspondence between chemical similarity of metabolic reactions and their MOLMAP descriptors was applied to the identification of a number of reactions mapped into the same neuron but belonging to different EC classes, which demonstrated the ability of the MOLMAP/SOM approach to verify the internal consistency of classifications in databases of metabolic reactions. RFs were also used to assign the four levels of the EC hierarchy from the reaction equation. EC numbers were correctly assigned in 95%, 90%, 85% and 86% of the cases (for independent test sets) at the class, subclass, sub-subclass and full EC number level,respectively. Experiments for the classification of reactions from the main reactants and products were performed with RFs - EC numbers were assigned at the class, subclass and sub-subclass level with accuracies of 78%, 74% and 63%, respectively. In the course of the experiments with metabolic reactions we suggested that the MOLMAP / SOM concept could be extended to the representation of other levels of metabolic information such as metabolic pathways. Following the MOLMAP idea, the pattern of neurons activated by the reactions of a metabolic pathway is a representation of the reactions involved in that pathway - a descriptor of the metabolic pathway. This reasoning enabled the comparison of different pathways, the automatic classification of pathways, and a classification of organisms based on their biochemical machinery. The three levels of classification (from bonds to metabolic pathways) allowed to map and perceive chemical similarities between metabolic pathways even for pathways of different types of metabolism and pathways that do not share similarities in terms of EC numbers. Mapping of PES by neural networks (NNs). In a first series of experiments, ensembles of Feed-Forward NNs (EnsFFNNs) and Associative Neural Networks (ASNNs) were trained to reproduce PES represented by the Lennard-Jones (LJ) analytical potential function. The accuracy of the method was assessed by comparing the results of molecular dynamics simulations (thermal, structural, and dynamic properties) obtained from the NNs-PES and from the LJ function. The results indicated that for LJ-type potentials, NNs can be trained to generate accurate PES to be used in molecular simulations. EnsFFNNs and ASNNs gave better results than single FFNNs. A remarkable ability of the NNs models to interpolate between distant curves and accurately reproduce potentials to be used in molecular simulations is shown. The purpose of the first study was to systematically analyse the accuracy of different NNs. Our main motivation, however, is reflected in the next study: the mapping of multidimensional PES by NNs to simulate, by Molecular Dynamics or Monte Carlo, the adsorption and self-assembly of solvated organic molecules on noble-metal electrodes. Indeed, for such complex and heterogeneous systems the development of suitable analytical functions that fit quantum mechanical interaction energies is a non-trivial or even impossible task. The data consisted of energy values, from Density Functional Theory (DFT) calculations, at different distances, for several molecular orientations and three electrode adsorption sites. The results indicate that NNs require a data set large enough to cover well the diversity of possible interaction sites, distances, and orientations. NNs trained with such data sets can perform equally well or even better than analytical functions. Therefore, they can be used in molecular simulations, particularly for the ethanol/Au (111) interface which is the case studied in the present Thesis. Once properly trained, the networks are able to produce, as output, any required number of energy points for accurate interpolations.

Radiographers practice concerning breast compression in mammography: gradual versus no gradual

Purpose: Evaluate the type of breast compression (gradual or no gradual) that provides less discomfort to the patient. Methods and Materials: The standard projections were simulated [craniocaudal/(CC) and mediolateral-oblique/(MLO)] with the two breast compressions in 90 volunteers women aged between 19 and 86. The women were organised in groups according to the breast density. The intensity of discomfort was evaluated using the scale that have represented several faces (0-10) proposed by Wong Baker in the end of each simulation. It was also applied an interview using focus group to debate the score that were attributed during pain evaluation and to identify the criteria that were considered to do the classification. Results: The women aged between 19-29y (with higher breast density) classified the pain during no gradual compression as 4 and the gradual compression as 2 for both projections. The MLO projection was considered the most uncomfortable. During the focus group interview applied to this group was highlighted that compression did not promoted pain but discomfort. They considered that the high expectations of pain did not correspond to the discomfort that they felt. Similar results were identified for the older women (30-50y; > 50y). Conclusion: The radiographers should considerer the technique for breast compression. The gradual compression was considered for the majority of the women as the most comfortable regardless of breast density. The MLO projection was considered as uncomfortable due to the positioning (axila and inclusion of pectoral muscle) and due to the higher breast compression compared to the CC projection.

Failure on clinical image quality criteria in digital mammography: how can radiographers do better?

Purpose: To assess image quality using PGMI (perfect, good, moderate, inadequate) scale in digital mammography examinations acquired in DR systems. Identify the main failures and propose corrective actions. Evaluate the most typical breast density. Methods and Materials: Clinical image quality criteria were evaluated considering mammograms acquired in 13 DR systems and classified according to PGMI scale using the criteria described in European Commission guidelines for radiographers. The breast density was assessed according to ACR recommendations. The data were collected on the acquisition system monitor to reproduce the daily practice of the radiographer. Results: The image quality criteria were evaluated in 3044 images. The criteria were fully achieved in 41% of the images that were classified as P (perfect), 31 % of the images were classified as M (moderate), 20% G (good) and 9% I (inadequate). The main cause of inadequate image quality was absence of all breast tissue in the image, skin folders in the pectoral muscle and in the infra-mammary angle. The higher number of failures occurred in MLO projections (809 out of 1022). The most represented (36%) breast type was type 2 (25-50% glandular tissue). Conclusion: Incorrect radiographic technique was frequently detected suggesting potential training needs and poor communication between the team members (radiographer and radiologists). Further correlations are necessary to identify the main causes for the failures, namely specific education and training in digital mammography and workload.