Machine Learning and Clinical Text. Supporting Health Information Flow

Fluent health information flow is critical for clinical decision-making. However, a considerable part of this information is free-form text and inabilities to utilize it create risks to patient safety and cost-­effective hospital administration. Methods for automated processing of clinical text are emerging. The aim in this doctoral dissertation is to study machine learning and clinical text in order to support health information flow.First, by analyzing the content of authentic patient records, the aim is to specify clinical needs in order to guide the development of machine learning applications.The contributions are a model of the ideal information flow,a model of the problems and challenges in reality, and a road map for the technology development. Second, by developing applications for practical cases,the aim is to concretize ways to support health information flow. Altogether five machine learning applications for three practical cases are described: The first two applications are binary classification and regression related to the practical case of topic labeling and relevance ranking.The third and fourth application are supervised and unsupervised multi-class classification for the practical case of topic segmentation and labeling.These four applications are tested with Finnish intensive care patient records.The fifth application is multi-label classification for the practical task of diagnosis coding. It is tested with English radiology reports.The performance of all these applications is promising. Third, the aim is to study how the quality of machine learning applications can be reliably evaluated.The associations between performance evaluation measures and methods are addressed,and a new hold-out method is introduced.This method contributes not only to processing time but also to the evaluation diversity and quality. The main conclusion is that developing machine learning applications for text requires interdisciplinary, international collaboration. Practical cases are very different, and hence the development must begin from genuine user needs and domain expertise. The technological expertise must cover linguistics,machine learning, and information systems. Finally, the methods must be evaluated both statistically and through authentic user-feedback.

Color ad spectral image assessment using novel quality and fidelity techniques

The ongoing development of the digital media has brought a new set of challenges with it. As images containing more than three wavelength bands, often called spectral images, are becoming a more integral part of everyday life, problems in the quality of the RGB reproduction from the spectral images have turned into an important area of research. The notion of image quality is often thought to comprise two distinctive areas – image quality itself and image fidelity, both dealing with similar questions, image quality being the degree of excellence of the image, and image fidelity the measure of the match of the image under study to the original. In this thesis, both image fidelity and image quality are considered, with an emphasis on the influence of color and spectral image features on both. There are very few works dedicated to the quality and fidelity of spectral images. Several novel image fidelity measures were developed in this study, which include kernel similarity measures and 3D-SSIM (structural similarity index). The kernel measures incorporate the polynomial, Gaussian radial basis function (RBF) and sigmoid kernels. The 3D-SSIM is an extension of a traditional gray-scale SSIM measure developed to incorporate spectral data. The novel image quality model presented in this study is based on the assumption that the statistical parameters of the spectra of an image influence the overall appearance. The spectral image quality model comprises three parameters of quality: colorfulness, vividness and naturalness. The quality prediction is done by modeling the preference function expressed in JNDs (just noticeable difference). Both image fidelity measures and the image quality model have proven to be effective in the respective experiments.

Development of Color Difference Equations Matching to Human Vision System

Research on color difference evaluation has been active in recent thirty years. Several color difference formulas were developed for industrial applications. The aims of this thesis are to develop the color density which is denoted by comb g and to propose the color density based chromaticity difference formulas. Color density is derived from the discrimination ellipse parameters and color positions in the xy , xyY and CIELAB color spaces, and the color based chromaticity difference formulas are compared with the line element formulas and CIE 2000 color difference formulas. As a result of the thesis, color density represents the perceived color difference accurately, and it could be used to characterize a color by the attribute of perceived color difference from this color.

Design and evaluation of a continuous flow, integrated nebulizer-hydride generator for flame atomic absorption spectrometry

An evaluation of the performance of a continuous flow hydride generator-nebulizer for flame atomic absorption spectrometry was carried out. Optimization of nebulizer gas flow rate, sample acid concentration, sample and tetrahydroborate uptake rates and reductant concentration, on the As and Se absorbance signals was carried out. A hydrogen-argon flame was used. An improvement of the analytical sensitivity relative to the conventional bead nebulizer used in flame AA was obtained (2 (As) and 4.8 (Se) µg L-1). Detection limits (3σb) of 1 (As) and 1.3 (Se) µg L-1 were obtained. Accuracy of the method was checked by analyzing an oyster tissue reference material.

Flow analysis-hydride generation-gas phase derivative molecular absorption spectrophotometric determination of antimony in antileishmanial drugs

In the present work, the development of a method based on the coupling of flow analysis (FA), hydride generation (HG), and derivative molecular absorption spectrophotometry (D-EAM) in gas phase (GP), is described in order to determine total antimony in antileishmanial products. Second derivative order (D²224nm) of the absorption spectrum (190 - 300 nm) is utilized as measurement criterion. Each one of the parameters involved in the development of the proposed method was examined and optimized. The utilization of the EAM in GP as detection system in a continuous mode instead of atomic absorption spectrometry represents the great potential of the analytic proposal.

Flow injection green method for the quantitative analysis of ketoconazole in pharmaceutical preparations

A flow injection method for the quantitative analysis of ketoconazole in tablets, based on the reaction with iron (III) ions, is presented. Ketoconazole forms a red complex with iron ions in an acid medium, with maximum absorbance at 495 nm. The detection limit was estimated to be 1×10--4 mol L-1; the quantitation limit is about 3×10--4 mol L-1 and approximately 30 determinations can be performed in an hour. The results were compared with those obtained with a reference HPLC method. Statistical comparisons were done using the Student's t procedure and the F test. Complete agreement was found at the 0.95 significance level between the proposed flow injection and the HPLC procedures. The two methods present similar precision, i.e., for HPLC the mean relative standard deviation was ca. 1.2% and for FIA ca. 1.6%.

Determination of cinnamic acid in human urine by flow injection chemiluminescence

It was found that cinnamic acid can react with potassium permanganate in the acidic medium and produce chemiluminescence, which was greatly enhanced by glyoxal. Under the optimum conditions, the linear range for the determination of cinnamic acid was 1.0×10-8 to 1.0×10-4 mol L-1 with a detection limit of 8.0×10-9 mol L-1, the relative standard deviation was 1.7% for 2.0×10-6 mol L-1 cinnamic acid solution in nine repeated measurements. This method was found to be novel0simple0fast and sensitive, it was successfully applied to the determination of cinnamic acid in human urine. Furthermore, the possible reaction mechanism was also discussed.

Preliminary data on Brazilian monofloral honey from the northeast region using FT-IR ATR spectroscopic, palynological, and color analysis

Fourier transform infrared attenuated total reflectance (FT-IR ATR) spectroscopy was used to determine 14 different measurands in northeast Brazilian honey samples. Nine different honey samples (six monofloral and three polyfloral) from 2009 obtained from the company CEARAPI underwent FT-IR ATR, palynological, color, and sensorial analysis to obtain preliminary results for these types of honey. The results showed that there are five monofloral, three bifloral, and one extrafloral honey, and also that mid-infrared spectrometry can be used as a screening method for the routine analysis of Brazilian honey, with the advantages of being rapid, nondestructive, and accurate.

Determination of sulfate in the wet-process of phosphoric acid by reverse flow injection

An improved method based on reverse flow injection is proposed for determining sulfate concentration in the wet-process of phosphoric acid (WPA). The effect of reagent composition, flow rate, temperature, acid concentration, length of the reaction coil, and linear response range on the flow system is discussed in detail. Optimal conditions are established for determining sulfate in the WPA samples. Baseline drift is avoided by a periodic washing step with EDTA in an alkaline medium. A linear response is observed within a range of 20 - 360 mg L-1, given by the equation A = 0.0020C (mg L-1) + 0.0300, R² = 0.9991. The detection limit of the proposed method for sulfate analysis is 3 mg L-1, and the relative standard deviation (n = 12) of sulfate absorbance peak is less than 1.60%. This method has a rate of up to 29 samples per hour, and the results compare well with those obtained with gravimetric method.

LOW COST ANALYZER FOR THE DETERMINATION OF PHOSPHORUS BASED ON OPEN-SOURCE HARDWARE AND PULSED FLOWS

The need for automated analyzers for industrial and environmental samples has triggered the research for new and cost-effective strategies of automation and control of analytical systems. The widespread availability of open-source hardware together with novel analytical methods based on pulsed flows have opened the possibility of implementing standalone automated analytical systems at low cost. Among the areas that can benefit from this approach are the analysis of industrial products and effluents and environmental analysis. In this work, a multi-pumping flow system is proposed for the determination of phosphorus in effluents and polluted water samples. The system employs photometric detection based on the formation of molybdovanadophosphoric acid, and the fluidic circuit is built using three solenoid micropumps. The detection is implemented with a low cost LED-photodiode photometric detection system and the whole system is controlled by an open-source Arduino Uno microcontroller board. The optimization of the timing to ensure the color development and the pumping cycle is discussed for the proposed implementation. Experimental results to evaluate the system behavior are presented verifying a linear relationship between the relative absorbance and the phosphorus concentrations for levels as high as 50 mg L-1.

The Regulation of Skeletal and Cardiac Muscle Blood Flow in Humans. Studies by positron emission tomography with special reference to exercise, adenosine and nitric oxide

Virtually every cell and organ in the human body is dependent on a proper oxygen supply. This is taken care of by the cardiovascular system that supplies tissues with oxygen precisely according to their metabolic needs. Physical exercise is one of the most demanding challenges the human circulatory system can face. During exercise skeletal muscle blood flow can easily increase some 20-fold and its proper distribution to and within muscles is of importance for optimal oxygen delivery. The local regulation of skeletal muscle blood flow during exercise remains little understood, but adenosine and nitric oxide may take part in this process. In addition to acute exercise, long-term vigorous physical conditioning also induces changes in the cardiovasculature, which leads to improved maximal physical performance. The changes are largely central, such as structural and functional changes in the heart. The function and reserve of the heart’s own vasculature can be studied by adenosine infusion, which according to animal studies evokes vasodilation via it’s a_2A receptors. This has, however, never been addressed in humans in vivo and also studies in endurance athletes have shown inconsistent results regarding the effects of sport training on myocardial blood flow. This study was performed on healthy young adults and endurance athletes and local skeletal and cardiac muscle blod flow was measured by positron emission tomography. In the heart, myocardial blood flow reserve and adenosine A_2A receptor density, and in skeletal muscle, oxygen extraction and consumption was also measured. The role of adenosine in the control of skeletal muscle blood flow during exercise, and its vasodilator effects, were addressed by infusing competitive inhibitors and adenosine into the femoral artery. The formation of skeletal muscle nitric oxide was also inhibited by a drug, with and without prostanoid blockade. As a result and conclusion, it can be said that skeletal muscle blood flow heterogeneity decreases with increasing exercise intensity most likely due to increased vascular unit recruitment, but exercise hyperemia is a very complex phenomenon that cannot be mimicked by pharmacological infusions, and no single regulator factor (e.g. adenosine or nitric oxide) accounts for a significant part of exercise-induced muscle hyperemia. However, in the present study it was observed for the first time in humans that nitric oxide is not only important regulator of the basal level of muscle blood flow, but also oxygen consumption, and together with prostanoids affects muscle blood flow and oxygen consumption during exercise. Finally, even vigorous endurance training does not seem to lead to supranormal myocardial blood flow reserve, and also other receptors than A_2A mediate the vasodilator effects of adenosine. In respect to cardiac work, atheletes heart seems to be luxuriously perfused at rest, which may result from reduced oxygen extraction or impaired efficiency due to pronouncedly enhanced myocardial mass developed to excel in strenuous exercise.

Separation of Gas Bubbles from Circulation Waters and Pulp Suspensions in Open Channel Flow

The objective of this thesis was to study the removal of gases from paper mill circulation waters experimentally and to provide data for CFD modeling. Flow and bubble size measurements were carried out in a laboratory scale open gas separation channel. Particle Image Velocimetry (PIV) technique was used to measure the gas and liquid flow fields, while bubble size measurements were conducted using digital imaging technique with back light illumination. Samples of paper machine waters as well as a model solution were used for the experiments. The PIV results show that the gas bubbles near the feed position have the tendency to escape from the circulation channel at a faster rate than those bubbles which are further away from the feed position. This was due to an increased rate of bubble coalescence as a result of the relatively larger bubbles near the feed position. Moreover, a close similarity between the measured slip velocities of the paper mill waters and that of literature values was obtained. It was found that due to dilution of paper mill waters, the observed average bubble size was considerably large as compared to the average bubble sizes in real industrial pulp suspension and circulation waters. Among the studied solutions, the model solution has the highest average drag coefficient value due to its relatively high viscosity. The results were compared to a 2D steady sate CFD simulation model. A standard Euler-Euler k-ε turbulence model was used in the simulations. The channel free surface was modeled as a degassing boundary. From the drag models used in the simulations, the Grace drag model gave velocity fields closest to the experimental values. In general, the results obtained from experiments and CFD simulations are in good qualitative agreement.

The use of fluorescent probes to assess viability of the plant pathogenic bacterium Clavibacter michiganensis subsp. michiganensis by flow cytometry

Determination of the viability of bacteria by the conventional plating technique is a time-consuming process. Methods based on enzyme activity or membrane integrity are much faster and may be good alternatives. Assessment of the viability of suspensions of the plant pathogenic bacterium Clavibacter michiganensis subsp. michiganensis (Cmm) using the fluorescent probes Calcein acetoxy methyl ester (Calcein AM), carboxyfluorescein diacetate (cFDA), and propidium iodide (PI) in combination with flow cytometry was evaluated. Heat-treated and viable (non-treated) Cmm cells labeled with Calcein AM, cFDA, PI, or combinations of Calcein AM and cFDA with PI, could be distinguished based on their fluorescence intensity in flow cytometry analysis. Non-treated cells showed relatively high green fluorescence levels due to staining with either Calcein AM or cFDA, whereas damaged cells (heat-treated) showed high red fluorescence levels due to staining with PI. Flow cytometry also allowed a rapid quantification of viable Cmm cells labeled with Calcein AM or cFDA and heat-treated cells labeled with PI. Therefore, the application of flow cytometry in combination with fluorescent probes appears to be a promising technique for assessing viability of Cmm cells when cells are labeled with Calcein AM or the combination of Calcein AM with PI.

Flow injection spectrophotometric determination of adrenaline using a solid-phase reactor containing triiodide ions immobilized in an anion-exchange resin

A flow injection spectrophotometric procedure with on-line solid-phase reactor containing ion triiodide immobilized in an anion-exchange resin is proposed for the determination of adrenaline (epinephrine) in pharmaceutical products. Adrenaline is oxidized by triiodide ion immobilized in an anionic-exchange resin yielding adrenochrome which is transported by the carrier solution and detected at a wavelength of 488 nm. Adrenaline was determined in three pharmaceutical products in the 6.4 x 10-6 to 3.0 x 10-4 mol L-1 concentration range with a detection limit of 4.8 x 10-7 mol L-1. The recovery of this analyte in three samples ranged from 96.0 to 105 %. The analytical frequency was 80 determinations per hour and the RSDs were less than 1 % for adrenaline concentrations of 6.4 x 10-5 and 2.0 x 10-4 mol L-1 (n=10). A paired t-test showed that all results obtained for adrenaline in commercial formulations using the proposed flow injection procedure and a spectrophotometric batch procedure agree at the 95% confidence level.