Medical information extraction in European Portuguese

The electronic storage of medical patient data is becoming a daily experience in most of the practices and hospitals worldwide. However, much of the data available is in free-form text, a convenient way of expressing concepts and events, but especially challenging if one wants to perform automatic searches, summarization or statistical analysis. Information Extraction can relieve some of these problems by offering a semantically informed interpretation and abstraction of the texts. MedInX, the Medical Information eXtraction system presented in this document, is the first information extraction system developed to process textual clinical discharge records written in Portuguese. The main goal of the system is to improve access to the information locked up in unstructured text, and, consequently, the efficiency of the health care process, by allowing faster and reliable access to quality information on health, for both patient and health professionals. MedInX components are based on Natural Language Processing principles, and provide several mechanisms to read, process and utilize external resources, such as terminologies and ontologies, in the process of automatic mapping of free text reports onto a structured representation. However, the flexible and scalable architecture of the system, also allowed its application to the task of Named Entity Recognition on a shared evaluation contest focused on Portuguese general domain free-form texts. The evaluation of the system on a set of authentic hospital discharge letters indicates that the system performs with 95% F-measure, on the task of entity recognition, and 95% precision on the task of relation extraction. Example applications, demonstrating the use of MedInX capabilities in real applications in the hospital setting, are also presented in this document. These applications were designed to answer common clinical problems related with the automatic coding of diagnoses and other health-related conditions described in the documents, according to the international classification systems ICD-9-CM and ICF. The automatic review of the content and completeness of the documents is an example of another developed application, denominated MedInX Clinical Audit system.

Pregnancy and newborns disorders followed by urine metabolomics

Chapter 1 introduces the scope of the work by identifying the clinically relevant prenatal disorders and presently available diagnostic methods. The methodology followed in this work is presented, along with a brief account of the principles of the analytical and statistical tools employed. A thorough description of the state of the art of metabolomics in prenatal research concludes the chapter, highlighting the merit of this novel strategy to identify robust disease biomarkers. The scarce use of maternal and newborn urine in previous reports enlightens the relevance of this work. Chapter 2 presents a description of all the experimental details involved in the work performed, comprising sampling, sample collection and preparation issues, data acquisition protocols and data analysis procedures. The proton Nuclear Magnetic Resonance (NMR) characterization of maternal urine composition in healthy pregnancies is presented in Chapter 3. The urinary metabolic profile characteristic of each pregnancy trimester was defined and a 21-metabolite signature found descriptive of the metabolic adaptations occurring throughout pregnancy. 8 metabolites were found, for the first time to our knowledge, to vary in connection to pregnancy, while known metabolic effects were confirmed. This chapter includes a study of the effects of non-fasting (used in this work) as a possible confounder. Chapter 4 describes the metabolomic study of 2nd trimester maternal urine for the diagnosis of fetal disorders and prediction of later-developing complications. This was achieved by applying a novel variable selection method developed in the context of this work. It was found that fetal malformations (FM) (and, specifically those of the central nervous system, CNS) and chromosomal disorders (CD) (and, specifically, trisomy 21, T21) are accompanied by changes in energy, amino acids, lipids and nucleotides metabolic pathways, with CD causing a further deregulation in sugars metabolism, urea cycle and/or creatinine biosynthesis. Multivariate analysis models´ validation revealed classification rates (CR) of 84% for FM (87%, CNS) and 85% for CD (94%, T21). For later-diagnosed preterm delivery (PTD), preeclampsia (PE) and intrauterine growth restriction (IUGR), it is found that urinary NMR profiles have early predictive value, with CRs ranging from 84% for PTD (11-20 gestational weeks, g.w., prior to diagnosis), 94% for PE (18-24 g.w. pre-diagnosis) and 94% for IUGR (2-22 g.w. pre-diagnosis). This chapter includes results obtained for an ultraperformance liquid chromatography-mass spectrometry (UPLC-MS) study of pre-PTD samples and correlation with NMR data. One possible marker was detected, although its identification was not possible. Chapter 5 relates to the NMR metabolomic study of gestational diabetes mellitus (GDM), establishing a potentially predictive urinary metabolic profile for GDM, 2-21 g.w. prior to diagnosis (CR 83%). Furthermore, the NMR spectrum was shown to carry information on individual phenotypes, able to predict future insulin treatment requirement (CR 94%). Chapter 6 describes results that demonstrate the impact of delivery mode (CR 88%) and gender (CR 76%) on newborn urinary profile. It was also found that newborn prematurity, respiratory depression, large for gestational age growth and malformations induce relevant metabolic perturbations (CR 82-92%), as well as maternal conditions, namely GDM (CR 82%) and maternal psychiatric disorders (CR 91%). Finally, the main conclusions of this thesis are presented in Chapter 7, highlighting the value of maternal or newborn urine metabolomics for pregnancy monitoring and disease prediction, towards the development of new early and non-invasive diagnostic methods.

Metabolic signature of lung cancer: a metabolomic study of human tissues and biofluids

This thesis reports the application of metabolomics to human tissues and biofluids (blood plasma and urine) to unveil the metabolic signature of primary lung cancer. In Chapter 1, a brief introduction on lung cancer epidemiology and pathogenesis, together with a review of the main metabolic dysregulations known to be associated with cancer, is presented. The metabolomics approach is also described, addressing the analytical and statistical methods employed, as well as the current state of the art on its application to clinical lung cancer studies. Chapter 2 provides the experimental details of this work, in regard to the subjects enrolled, sample collection and analysis, and data processing. In Chapter 3, the metabolic characterization of intact lung tissues (from 56 patients) by proton High Resolution Magic Angle Spinning (HRMAS) Nuclear Magnetic Resonance (NMR) spectroscopy is described. After careful assessment of acquisition conditions and thorough spectral assignment (over 50 metabolites identified), the metabolic profiles of tumour and adjacent control tissues were compared through multivariate analysis. The two tissue classes could be discriminated with 97% accuracy, with 13 metabolites significantly accounting for this discrimination: glucose and acetate (depleted in tumours), together with lactate, alanine, glutamate, GSH, taurine, creatine, phosphocholine, glycerophosphocholine, phosphoethanolamine, uracil nucleotides and peptides (increased in tumours). Some of these variations corroborated typical features of cancer metabolism (e.g., upregulated glycolysis and glutaminolysis), while others suggested less known pathways (e.g., antioxidant protection, protein degradation) to play important roles. Another major and novel finding described in this chapter was the dependence of this metabolic signature on tumour histological subtype. While main alterations in adenocarcinomas (AdC) related to phospholipid and protein metabolisms, squamous cell carcinomas (SqCC) were found to have stronger glycolytic and glutaminolytic profiles, making it possible to build a valid classification model to discriminate these two subtypes. Chapter 4 reports the NMR metabolomic study of blood plasma from over 100 patients and near 100 healthy controls, the multivariate model built having afforded a classification rate of 87%. The two groups were found to differ significantly in the levels of lactate, pyruvate, acetoacetate, LDL+VLDL lipoproteins and glycoproteins (increased in patients), together with glutamine, histidine, valine, methanol, HDL lipoproteins and two unassigned compounds (decreased in patients). Interestingly, these variations were detected from initial disease stages and the magnitude of some of them depended on the histological type, although not allowing AdC vs. SqCC discrimination. Moreover, it is shown in this chapter that age mismatch between control and cancer groups could not be ruled out as a possible confounding factor, and exploratory external validation afforded a classification rate of 85%. The NMR profiling of urine from lung cancer patients and healthy controls is presented in Chapter 5. Compared to plasma, the classification model built with urinary profiles resulted in a superior classification rate (97%). After careful assessment of possible bias from gender, age and smoking habits, a set of 19 metabolites was proposed to be cancer-related (out of which 3 were unknowns and 6 were partially identified as N-acetylated metabolites). As for plasma, these variations were detected regardless of disease stage and showed some dependency on histological subtype, the AdC vs. SqCC model built showing modest predictive power. In addition, preliminary external validation of the urine-based classification model afforded 100% sensitivity and 90% specificity, which are exciting results in terms of potential for future clinical application. Chapter 6 describes the analysis of urine from a subset of patients by a different profiling technique, namely, Ultra-Performance Liquid Chromatography coupled to Mass Spectrometry (UPLC-MS). Although the identification of discriminant metabolites was very limited, multivariate models showed high classification rate and predictive power, thus reinforcing the value of urine in the context of lung cancer diagnosis. Finally, the main conclusions of this thesis are presented in Chapter 7, highlighting the potential of integrated metabolomics of tissues and biofluids to improve current understanding of lung cancer altered metabolism and to reveal new marker profiles with diagnostic value.

A computational study of ionic liquids used in the production of fuels and biofuels

For the past decades it has been a worldwide concern to reduce the emission of harmful gases released during the combustion of fossil fuels. This goal has been addressed through the reduction of sulfur-containing compounds, and the replacement of fossil fuels by biofuels, such as bioethanol, produced in large scale from biomass. For this purpose, a new class of solvents, the Ionic Liquids (ILs), has been applied, aiming at developing new processes and replacing common organic solvents in the current processes. ILs can be composed by a large number of different combinations of cations and anions, which confer unique but desired properties to ILs. The ability of fine-tuning the properties of ILs to meet the requirements of a specific application range by mixing different cations and anions arises as the most relevant aspect for rendering ILs so attractive to researchers. Nonetheless, due to the huge number of possible combinations between the ions it is required the use of cheap predictive approaches for anticipating how they will act in a given situation. Molecular dynamics (MD) simulation is a statistical mechanics computational approach, based on Newton’s equations of motion, which can be used to study macroscopic systems at the atomic level, through the prediction of their properties, and other structural information. In the case of ILs, MD simulations have been extensively applied. The slow dynamics associated to ILs constitutes a challenge for their correct description that requires improvements and developments of existent force fields, as well as larger computational efforts (longer times of simulation). The present document reports studies based on MD simulations devoted to disclose the mechanisms of interaction established by ILs in systems representative of fuel and biofuels streams, and at biomass pre-treatment process. Hence, MD simulations were used to evaluate different systems composed of ILs and thiophene, benzene, water, ethanol and also glucose molecules. For the latter molecules, it was carried out a study aiming to ascertain the performance of a recently proposed force field (GROMOS 56ACARBO) to reproduce the dynamic behavior of such molecules in aqueous solution. The results here reported reveal that the interactions established by ILs are dependent on the individual characteristics of each IL. Generally, the polar character of ILs is deterministic in their propensity to interact with the other molecules. Although it is unquestionable the advantage of using MD simulations, it is necessary to recognize the need for improvements and developments of force fields, not only for a successful description of ILs, but also for other relevant compounds such as the carbohydrates.