Carbohydrate Estimation by a Mobile Phone-Based System Versus Self-Estimations of Individuals With Type 1 Diabetes Mellitus: A Comparative Study

Background: Diabetes mellitus is spreading throughout the world and diabetic individuals have been shown to often assess their food intake inaccurately; therefore, it is a matter of urgency to develop automated diet assessment tools. The recent availability of mobile phones with enhanced capabilities, together with the advances in computer vision, have permitted the development of image analysis apps for the automated assessment of meals. GoCARB is a mobile phone-based system designed to support individuals with type 1 diabetes during daily carbohydrate estimation. In a typical scenario, the user places a reference card next to the dish and acquires two images using a mobile phone. A series of computer vision modules detect the plate and automatically segment and recognize the different food items, while their 3D shape is reconstructed. Finally, the carbohydrate content is calculated by combining the volume of each food item with the nutritional information provided by the USDA Nutrient Database for Standard Reference. Objective: The main objective of this study is to assess the accuracy of the GoCARB prototype when used by individuals with type 1 diabetes and to compare it to their own performance in carbohydrate counting. In addition, the user experience and usability of the system is evaluated by questionnaires. Methods: The study was conducted at the Bern University Hospital, “Inselspital” (Bern, Switzerland) and involved 19 adult volunteers with type 1 diabetes, each participating once. Each study day, a total of six meals of broad diversity were taken from the hospital’s restaurant and presented to the participants. The food items were weighed on a standard balance and the true amount of carbohydrate was calculated from the USDA nutrient database. Participants were asked to count the carbohydrate content of each meal independently and then by using GoCARB. At the end of each session, a questionnaire was completed to assess the user’s experience with GoCARB. Results: The mean absolute error was 27.89 (SD 38.20) grams of carbohydrate for the estimation of participants, whereas the corresponding value for the GoCARB system was 12.28 (SD 9.56) grams of carbohydrate, which was a significantly better performance ( P=.001). In 75.4% (86/114) of the meals, the GoCARB automatic segmentation was successful and 85.1% (291/342) of individual food items were successfully recognized. Most participants found GoCARB easy to use. Conclusions: This study indicates that the system is able to estimate, on average, the carbohydrate content of meals with higher accuracy than individuals with type 1 diabetes can. The participants thought the app was useful and easy to use. GoCARB seems to be a well-accepted supportive mHealth tool for the assessment of served-on-a-plate meals.

Methodological and physiological test-retest reliability of (13) C-MRS glycogen measurements in liver and in skeletal muscle of patients with type 1 diabetes and matched healthy controls.

Glycogen is a major substrate in energy metabolism and particularly important to prevent hypoglycemia in pathologies of glucose homeostasis such as type 1 diabetes mellitus (T1DM). (13) C-MRS is increasingly used to determine glycogen in skeletal muscle and liver non-invasively; however, the low signal-to-noise ratio leads to long acquisition times, particularly when glycogen levels are determined before and after interventions. In order to ease the requirements for the subjects and to avoid systematic effects of the lengthy examination, we evaluated if a standardized preparation period would allow us to shift the baseline (pre-intervention) experiments to a preceding day. Based on natural abundance (13) C-MRS on a clinical 3 T MR system the present study investigated the test-retest reliability of glycogen measurements in patients with T1DM and matched controls (n = 10 each group) in quadriceps muscle and liver. Prior to the MR examination, participants followed a standardized diet and avoided strenuous exercise for two days. The average coefficient of variation (CV) of myocellular glycogen levels was 9.7% in patients with T1DM compared with 6.6% in controls after a 2 week period, while hepatic glycogen variability was 13.3% in patients with T1DM and 14.6% in controls. For comparison, a single-session test-retest variability in four healthy volunteers resulted in 9.5% for skeletal muscle and 14.3% for liver. Glycogen levels in muscle and liver were not statistically different between test and retest, except for hepatic glycogen, which decreased in T1DM patients in the retest examination, but without an increase of the group distribution. Since the CVs of glycogen levels determined in a "single session" versus "within weeks" are comparable, we conclude that the major source of uncertainty is the methodological error and that physiological variations can be minimized by a pre-study standardization. For hepatic glycogen examinations, familiarization sessions (MR and potentially strenuous interventions) are recommended. Copyright © 2016 John Wiley & Sons, Ltd.

Glucocorticoid alterations in the human type-1/type-2 cytokine balance and the role of CD28/B7 costimulation

We have recently reported that psychological stress is associated with a shift in the human type-1/type-2 cytokine balance toward a type-2 cytokine response. The mechanisms of these cytokine alterations are unknown, but likely involve glucocorticoid (GC) modulation of cytokine production. Therefore we sought to characterize the effects of GC on the in vitro human type-1/type-2 cytokine balance. We hypothesized that GC induce a type-2 cytokine shift through modulation of critical regulatory cytokines and alterations in the CD28/B7 costimulatory pathway. ^ We first sought to characterize the effect of the GC, dexamethasone (DEX), on type-1 (IFN-γ, IL-12) and type-2 (IL-4, IL-10) cytokine production by human peripheral blood mononuclear blood cells (pBMC) stimulated with a variety of T-lymphocyte and monocyte stimuli. DEX, at concentrations mimicking stress and supraphysiologic levels of cortisol, decreased IFN-γ and IL-12 production and increased IL-4 and IL-10 production, indicating a shift in the type-1/type-2 cytokine balance toward a type-2 response. Furthermore, both CD4+ and CD8+ T-lymphocytes were susceptible to the cytokine modulating effects of DEX. Furthermore, in the absence of the monocyte, the DEX-induced alterations in T-lymphocyte cytokine production were reduced, indicating that the interaction between the monocyte and T-lymphocyte plays a significant role. ^ We next determined the role of regulatory cytokines, known to modulate the type-1/type-2 cytokine balance, in the DEX-induced cytokine alterations. The addition of the recombinant IL-12p70 and IFN-γ, but not the neutralization of IL-4, IL-10 or IL-13 using monoclonal antibodies, attenuated the DEX-induced type-1/type-2 cytokine alterations. These data suggest that the DEX-induced cytokine alterations are mediated, at least in part, through the initial inhibition type-1 cytokines. Lastly, we investigated the role of the CD28/B7 costimulatory pathway in these cytokine alterations. DEX decreased the expression of CD80 and CD86 on THP-1 cells, a monocyte cell line, and the expression of CD28 and CTLA-4 on PHA-stimulated pBMC. The DEX-induced decrease in CD28 and CTLA-4 expression was attenuated by rhIL-12. Finally, CD28 activation attenuated the DEX-induced decrease in IFN-γ production, suggesting that modulation of the CD28/B7 costimulatory pathway may contribute to the DEX-induced type-1/type-2 cytokine alterations. ^

17-beta estradiol and progesterone-induced alterations in human type-1/type-2 cytokine balance and the role of costimulatory and apoptotic mechanisms

Evidence suggests that sex-based differences in immune function may predispose women to numerous hypersensitivity conditions such as Systemic lupus erythematosus (SLE), Hashimoto's thyroiditis and asthma. To date, the exact mechanisms of sexual dimorphism in immunity are not fully characterized but sex hormones such as 17-β estradiol (E2) and progesterone (PR) are believed to be involved. Since E2 and PR may modulate the production of critical regulatory cytokines, we sought to characterize their effects on the in vitro human type-1/type-2 cytokine balance. We hypothesized that E2 and/or PR vary cytokine production and influence costimulatory molecule expression and apoptosis. We first described the effect of E2 and/or PR on type-1 (IFN-γ and IL-12) and type-2 (IL-4 and IL-10) cytokine production by human peripheral blood mononuclear cells (PBMC) treated with various T-lymphocyte and monocyte stimuli. E2 and/or PR were each used at concentrations similar to those found at the maternal-fetal interface during pregnancy. At this dose, E2 increased IFN-γ and IL-12 production and PR decreased IFN-γ production and tended to increase IL-4 production. Furthermore, the combination of E2+PR decreased IL-12 production. This suggests that E2 shifts the type-1/type-2 cytokine balance towards a type-1 response and that PR and E2+PR shift the balance towards a type-2 response. Next, we used intracellular cytokine detection to demonstrate that E2 and/or PR are capable of altering cytokine production of CD3+ T-cells and the CD3+CD4+ and CD3+CD8+ subsets. In addition, we used the H9 T-lymphocyte cell line and the THP-1 monocyte cell line to show that E2 and/or PR can induce cytokine effects in both T-cells and monocytes independent of their interaction. Lastly, we determined the effect of E2 and/or PR on costimulatory molecule expression and apoptosis as potential mechanisms for the cytokine-induced alterations. E2 increased and PR decreased CD80 expression on THP-1 cells and PR and E2+PR decreased CD28 expression in PBMC and Jurkat cells. Furthermore, E2, PR and E2+PR increased Fas-mediated apoptosis in Jurkat cells and E2 increased FasL expression on THP-1 cells. Thus, E2 and/or PR may alter the cytokine balance by modulating the CD28/CD80 costimulatory pathway and apoptosis. ^

Age Related Differences in Individual Quality of Life Domains in Youth with Type 1 Diabetes.

BACKGROUND: Investigating individual, as opposed to predetermined, quality of life domains may yield important information about quality of life. This study investigated the individual quality of life domains nominated by youth with type 1 diabetes. METHODS: Eighty young people attending a diabetes summer camp completed the Schedule for the Evaluation of Individual Quality of Life-Direct Weighting interview, which allows respondents to nominate and evaluate their own quality of life domains. RESULTS: The most frequently nominated life domains were 'family', 'friends', 'diabetes', 'school', and 'health' respectively; ranked in terms of importance, domains were 'religion', 'family', 'diabetes', 'health', and 'the golden rule'; ranked in order of satisfaction, domains were 'camp', 'religion', 'pets', and 'family' and 'a special person' were tied for fifth. Respondent age was significantly positively associated with the importance of 'friends', and a significantly negatively associated with the importance of 'family'. Nearly all respondents nominated a quality of life domain relating to physical status, however, the specific physical status domain and the rationale for its nomination varied. Some respondents nominated 'diabetes' as a domain and emphasized diabetes 'self-care behaviors' in order to avoid negative health consequences such as hospitalization. Other respondents nominated 'health' and focused more generally on 'living well with diabetes'. In an ANOVA with physical status domain as the independent variable and age as the dependent variable, participants who nominated 'diabetes' were younger (M = 12.9 years) than those who nominated 'health' (M = 15.9 years). In a second ANOVA, with rationale for nomination the physical status domain as the independent variable, and age as the dependent variable, those who emphasized 'self care behaviors' were younger (M = 11.8 years) than those who emphasized 'living well with diabetes' (M = 14.6 years). These differences are discussed in terms of cognitive development and in relation to the decline in self-care and glycemic control often observed during adolescence. CONCLUSIONS: Respondents nominated many non-diabetes life domains, underscoring that QOL is multidimensional. Subtle changes in conceptualization of diabetes and health with increasing age may reflect cognitive development or disease adjustment, and speak to the need for special attention to adolescents. Understanding individual quality of life domains can help clinicians motivate their young patients with diabetes for self-care. Future research should employ a larger, more diverse sample, and use longitudinal designs.

The roles of costimulatory molecules in the cooperative effects of combination epinephrine and dexamethasone on type-1/type-2 cytokine production in tetanus-toxoid specific stimulated human peripheral blood mononuclear cells

A growing number of studies show strong associations between stress and altered immune function. In vivo studies of chronic and acute stress have demonstrated that cognitive stressors are strongly correlated with high circulating levels of catecholamines (CT) and corticosteroids (CS) that are associated with changes in type-1/type-2 cytokine expression. Although individual pharmacologic doses of CS and CT can inhibit the expression of T-helper 1 (Th1, type-1 like) and promote the production of T-helper 2 (Th2, type-2 like) cytokines in antigen-specific and mitogen stimulated human leukocyte cultures in vitro, little attention has been focused on the effects of combination physiologic-stress doses of CT and CS that may be more physiologically relevant. In addition, both in-vivo and in-vitro studies suggest that the differential expression of the B7 family of costimulatory molecules CD80 and CD86 may promote the expression of type-1 or type-2 cytokines, respectively. Furthermore, corticosteroids can influence the expression of β2-adrenergic receptors in various human tissues. We therefore investigated the combined effects of physiologic-stress doses of in vitro CT and CS upon the type-1/type-2 cytokine balance and expression of B7 costimulatory molecules of human peripheral blood mononuclear cells (PBMC) as a model to study the immunomodulatory effects of physiologic stress. Results demonstrated a significant decrease in type-1 cytokine expression and a significant increase in type-2 cytokine production in our CS+CT incubated cultures when compared to either CT or CS agents alone. In addition, we demonstrated the differential expression of CD80/CD86 in favor of CD86 at the cellular and population level as determined by flow cytometry in lipopolysaccharide stimulated human Monocytes. Furthermore, we developed flow cytometry based assays to detect total β2AR in human CD4+ T-lymphocytes that demonstrated decreased expression of β2AR in mitogen stimulated CD4+ T-lymphocytes in the presence of physiologic stress levels of CS and CT as single in vitro agents, however, when both CS and CT were combined, significantly higher expression of β2AR was observed. In summary, our in vitro data suggest that both CS and CT work cooperatively to shift immunity towards type-2 responses. ^

The impact of nutrition intervention on glycemic control in a pediatric type 1 diabetes mellitus population

Aim: The goal of this study was to evaluate the change in hemoglobin A1C and glycemic control after nutrition intervention among a population of type 1 diabetic pediatric patients. Methods: Data was collected from all type 1 diabetic patients who were scheduled for a consultation with the diabetes/endocrine RD from January 2006 through December 2006. Two groups were compared, those who kept their RD appointment and those who did not keep their appointment. The main outcome measure was HgbA1C. An independent samples t-test compared the two groups with respect to change in HbgA1C before and after the most recent scheduled appointment with the RD. Baseline characteristics were used as covariates and analyzed and controlled for using analysis of covariance (ANCOVA). Results: There was no difference in HgbA1c after either attending an RD appointment or not having attended an RD appointment. Those who arrived for and attended their RD appointment and those who did not arrive for and attend their RD appointment, had statistically different HgbA1C's before their scheduled appointment as well as after the RD appointment. However, the two groups were not equal at the beginning of the study period. Discussion: A study design with inclusion criteria of a specified range of HgbA1C values within which the study subjects needed to fall, would have potentially eliminated the difference between the two groups at the beginning of the study period. Conducting either another retrospective study that controlled for the initial HgbA1C value or conducting a prospective study that designated a range of HgbA1C values would be worth investigating to evaluate the impact of medical nutrition therapy intervention and the role of the RD in diabetes management. It is an interesting finding that there was a significant difference in the initial HgbA1c for those who came to the RD appointment compared to those who did not come. The fact that in this study those who did not arrive for their RD appointment had worse control of their diabetes suggests that this is a high-risk group. Targeting diabetes education toward this group of patients may prove to be beneficial. ^

Diabetic nephropathy in African-Americans with type 1 diabetes: The new Jersey 725 study

Diabetic nephropathy is the most common cause of end-stage renal disease (ESRD) in the United States. African-Americans and patients with type 1 diabetes (T1D) are at increased risk. We studied the rate and factors that influenced progression of glomerular filtration rate (GFR) in 401 African-American T1D patients who were followed for 6 years through the observational cohort New Jersey 725 study. Patients with ESRD and/or GFR<20 ml/min were excluded. The mean (SD) baseline GFR was 106.8 (27.04) ml/min and it decreased by 13.8 (mean, SD 32.2) ml/min during the 6-year period (2.3 ml/min/year). In patients with baseline macroproteinuria, GFR decreased by 31.8 (39.0) ml/min (5.3 ml/min/year) compared to 8.2 (mean, SD 27.6) ml/min (1.3 ml/min/year) in patients without it (p<0.00001). Six-year GFR fell to <20 ml/min in 5.25% of all patients, but in 16.8% of macroproteinuric patients.^ A model including baseline GFR, proteinuria category and hypertension category, explained 35% of the 6-year GFR variability (p<0.0001). After adjustment for other variables in the model, 6-year GFR was 24.9 ml/min lower in macroproteinuric patients than in those without proteinuria (p=0.0001), and 12.6 ml/min lower in patients with treated but uncontrolled hypertension compared to normotensive patients (p=0.003). In this sample of patients, with an elevated mean glycosylated hemoglobin of 12.4%, glycemic control did not independently influence GFR deterioration, nor did BMI, cholesterol, gender, age at diabetes onset or socioeconomic level.^ Taken together, our findings suggest that proteinuria and hypertension are the most important factors associated with GFR deterioration in African-American T1D patients.^

Binding of the HIV-1 gp120 -V3 to cellular glycosphingolipids is necessary for CXCR4 recruitment and infection

Infection by human immunodeficiency virus type 1 (HIV-1) is a multi-step process, and detailed analyses of the various events critical for productive infection are necessary to clearly understanding the infection process and identifying novel targets for therapeutic interventions. Evidence from this study reveals binding of the viral envelope protein to host cell glycosphingolipids (GSLs) as a novel event necessary for the orderly progression of the host cell-entry and productive infection by HIV-1. Data obtained from co-immunoprecipitation analyses and confocal microscopy showed that the ability of viral envelope to interact with the co-receptor CXCR4 and productive infection of HIV-1 were inhibited in cells rendered GSL-deficient, while both these activities were restored after reconstitution of the cells with specific GSLs like GM3. Furthermore, evidence was obtained using peptide-inhibitors of HIV-1 infection to show that binding of a specific region within the V3-loop of the envelope protein gp120 to the host cell GSLs is the trigger necessary for the CD4-bound gp120 to recruit the CXCR4 co-receptor. Infection-inhibitory activity of the V3 peptides was compromised in GSL-deficient cells, but could be restored by reconstitution of GSLs. Based on these findings, a revised model for HIV-1 infection is proposed that accounts for the established interactions between the viral envelope and host cell receptors while enumerating the importance of the new findings that fill the gap in the current knowledge of the sequential events for the HIV-1 entry. According to this model, post-CD4 binding of the HIV-1 envelope surface protein gp120 to host cell GSLs, mediated by the gp120-V3 region, enables formation of the gp120-CD4-GSL-CXCR4 immune-complex and productive infection. The identification of cellular GSLs as an additional class of co-factors necessary for HIV-1 infection is important for enhancing the basic knowledge of the HIV-1 entry that can be exploited for developing novel antiviral therapeutic strategies. ^

Ascorbic acid effects on the immune system and type -1 /type -2 cytokine balance in humans

Vitamin C (ascorbic acid--AA) can have a substantial impact on human health by reducing the incidence and/or severity of coryza. Studies also suggest it has immunomodulatory functions in humans. Immune function is controlled by cytokines, such as type-1 cytokines (IFNγ) that promote antiviral immunity and type-2 cytokines (IL-4, IL-10) that promote humoral immunity. Knowing the mechanisms responsible for both antiviral immunity and type-1/type-2 cytokine balance, we sought to identify AA-induced alterations of human peripheral blood mononuclear cells (PBMC) in vivo and in vitro . We hypothesized that AA modulates the immune system, altering both number and function of PBMC. We first described the effect of 14 days of oral (1 gram) AA in healthy subjects. AA increased circulating natural killer (NK) cells, CD25+ and HLA-DR+ T cells, and PMA/ionomycin-stimulated intracellular IFNγ. We subsequently developed models for in vitro use. We determined that AA was toxic in vitro to T cells when used at doses found intracellularly but doses found in plasma from individuals taking 1gm/day AA were nontoxic. The model that most fully reproduced our in vivo intracellular cytokine findings used dehydroascorbic acid and buffers to deliver AA intracellularly. This model generated the largest increase in IFNγ at physiologic plasma concentrations. Previous studies demonstrate that chronic psychological stress is associated with a type-2 cytokine response. We hypothesized that vitamin C could prevent the type-2 cytokine shift associated with stress. In a study of medical students taking 1 g AA or placebo, a significant increase in IFNγ was seen intracellularly in CD4+ and CD8+ cells and in tetanus-stimulated cultures in the AA group only. We also observed increases in IFNγ/IL-4 and IFNγ/IL-10 ratios with AA supplementation, indicating a type-1 shift. Furthermore, we noted increased numbers of NK cells and activated T cells in the peripheral blood in the AA treated group only. Lastly, we investigated the role of the CD40L/CD40 and CD28/B7 costimulatory pathway in these cytokine alterations. AA did not have any effect on either pathway studied. Thus costimulatory pathways are not contributing to AA induced modulation of the type-1/type-2 immune balance. ^

Multi-dimensional classification using Bayesian networks for stationary and evolving streaming data

Hoy en día, con la evolución continua y rápida de las tecnologías de la información y los dispositivos de computación, se recogen y almacenan continuamente grandes volúmenes de datos en distintos dominios y a través de diversas aplicaciones del mundo real. La extracción de conocimiento útil de una cantidad tan enorme de datos no se puede realizar habitualmente de forma manual, y requiere el uso de técnicas adecuadas de aprendizaje automático y de minería de datos. La clasificación es una de las técnicas más importantes que ha sido aplicada con éxito a varias áreas. En general, la clasificación se compone de dos pasos principales: en primer lugar, aprender un modelo de clasificación o clasificador a partir de un conjunto de datos de entrenamiento, y en segundo lugar, clasificar las nuevas instancias de datos utilizando el clasificador aprendido. La clasificación es supervisada cuando todas las etiquetas están presentes en los datos de entrenamiento (es decir, datos completamente etiquetados), semi-supervisada cuando sólo algunas etiquetas son conocidas (es decir, datos parcialmente etiquetados), y no supervisada cuando todas las etiquetas están ausentes en los datos de entrenamiento (es decir, datos no etiquetados). Además, aparte de esta taxonomía, el problema de clasificación se puede categorizar en unidimensional o multidimensional en función del número de variables clase, una o más, respectivamente; o también puede ser categorizado en estacionario o cambiante con el tiempo en función de las características de los datos y de la tasa de cambio subyacente. A lo largo de esta tesis, tratamos el problema de clasificación desde tres perspectivas diferentes, a saber, clasificación supervisada multidimensional estacionaria, clasificación semisupervisada unidimensional cambiante con el tiempo, y clasificación supervisada multidimensional cambiante con el tiempo. Para llevar a cabo esta tarea, hemos usado básicamente los clasificadores Bayesianos como modelos. La primera contribución, dirigiéndose al problema de clasificación supervisada multidimensional estacionaria, se compone de dos nuevos métodos de aprendizaje de clasificadores Bayesianos multidimensionales a partir de datos estacionarios. Los métodos se proponen desde dos puntos de vista diferentes. El primer método, denominado CB-MBC, se basa en una estrategia de envoltura de selección de variables que es voraz y hacia delante, mientras que el segundo, denominado MB-MBC, es una estrategia de filtrado de variables con una aproximación basada en restricciones y en el manto de Markov. Ambos métodos han sido aplicados a dos problemas reales importantes, a saber, la predicción de los inhibidores de la transcriptasa inversa y de la proteasa para el problema de infección por el virus de la inmunodeficiencia humana tipo 1 (HIV-1), y la predicción del European Quality of Life-5 Dimensions (EQ-5D) a partir de los cuestionarios de la enfermedad de Parkinson con 39 ítems (PDQ-39). El estudio experimental incluye comparaciones de CB-MBC y MB-MBC con los métodos del estado del arte de la clasificación multidimensional, así como con métodos comúnmente utilizados para resolver el problema de predicción de la enfermedad de Parkinson, a saber, la regresión logística multinomial, mínimos cuadrados ordinarios, y mínimas desviaciones absolutas censuradas. En ambas aplicaciones, los resultados han sido prometedores con respecto a la precisión de la clasificación, así como en relación al análisis de las estructuras gráficas que identifican interacciones conocidas y novedosas entre las variables. La segunda contribución, referida al problema de clasificación semi-supervisada unidimensional cambiante con el tiempo, consiste en un método nuevo (CPL-DS) para clasificar flujos de datos parcialmente etiquetados. Los flujos de datos difieren de los conjuntos de datos estacionarios en su proceso de generación muy rápido y en su aspecto de cambio de concepto. Es decir, los conceptos aprendidos y/o la distribución subyacente están probablemente cambiando y evolucionando en el tiempo, lo que hace que el modelo de clasificación actual sea obsoleto y deba ser actualizado. CPL-DS utiliza la divergencia de Kullback-Leibler y el método de bootstrapping para cuantificar y detectar tres tipos posibles de cambio: en las predictoras, en la a posteriori de la clase o en ambas. Después, si se detecta cualquier cambio, un nuevo modelo de clasificación se aprende usando el algoritmo EM; si no, el modelo de clasificación actual se mantiene sin modificaciones. CPL-DS es general, ya que puede ser aplicado a varios modelos de clasificación. Usando dos modelos diferentes, el clasificador naive Bayes y la regresión logística, CPL-DS se ha probado con flujos de datos sintéticos y también se ha aplicado al problema real de la detección de código malware, en el cual los nuevos ficheros recibidos deben ser continuamente clasificados en malware o goodware. Los resultados experimentales muestran que nuestro método es efectivo para la detección de diferentes tipos de cambio a partir de los flujos de datos parcialmente etiquetados y también tiene una buena precisión de la clasificación. Finalmente, la tercera contribución, sobre el problema de clasificación supervisada multidimensional cambiante con el tiempo, consiste en dos métodos adaptativos, a saber, Locally Adpative-MB-MBC (LA-MB-MBC) y Globally Adpative-MB-MBC (GA-MB-MBC). Ambos métodos monitorizan el cambio de concepto a lo largo del tiempo utilizando la log-verosimilitud media como métrica y el test de Page-Hinkley. Luego, si se detecta un cambio de concepto, LA-MB-MBC adapta el actual clasificador Bayesiano multidimensional localmente alrededor de cada nodo cambiado, mientras que GA-MB-MBC aprende un nuevo clasificador Bayesiano multidimensional. El estudio experimental realizado usando flujos de datos sintéticos multidimensionales indica los méritos de los métodos adaptativos propuestos. ABSTRACT Nowadays, with the ongoing and rapid evolution of information technology and computing devices, large volumes of data are continuously collected and stored in different domains and through various real-world applications. Extracting useful knowledge from such a huge amount of data usually cannot be performed manually, and requires the use of adequate machine learning and data mining techniques. Classification is one of the most important techniques that has been successfully applied to several areas. Roughly speaking, classification consists of two main steps: first, learn a classification model or classifier from an available training data, and secondly, classify the new incoming unseen data instances using the learned classifier. Classification is supervised when the whole class values are present in the training data (i.e., fully labeled data), semi-supervised when only some class values are known (i.e., partially labeled data), and unsupervised when the whole class values are missing in the training data (i.e., unlabeled data). In addition, besides this taxonomy, the classification problem can be categorized into uni-dimensional or multi-dimensional depending on the number of class variables, one or more, respectively; or can be also categorized into stationary or streaming depending on the characteristics of the data and the rate of change underlying it. Through this thesis, we deal with the classification problem under three different settings, namely, supervised multi-dimensional stationary classification, semi-supervised unidimensional streaming classification, and supervised multi-dimensional streaming classification. To accomplish this task, we basically used Bayesian network classifiers as models. The first contribution, addressing the supervised multi-dimensional stationary classification problem, consists of two new methods for learning multi-dimensional Bayesian network classifiers from stationary data. They are proposed from two different points of view. The first method, named CB-MBC, is based on a wrapper greedy forward selection approach, while the second one, named MB-MBC, is a filter constraint-based approach based on Markov blankets. Both methods are applied to two important real-world problems, namely, the prediction of the human immunodeficiency virus type 1 (HIV-1) reverse transcriptase and protease inhibitors, and the prediction of the European Quality of Life-5 Dimensions (EQ-5D) from 39-item Parkinson’s Disease Questionnaire (PDQ-39). The experimental study includes comparisons of CB-MBC and MB-MBC against state-of-the-art multi-dimensional classification methods, as well as against commonly used methods for solving the Parkinson’s disease prediction problem, namely, multinomial logistic regression, ordinary least squares, and censored least absolute deviations. For both considered case studies, results are promising in terms of classification accuracy as well as regarding the analysis of the learned MBC graphical structures identifying known and novel interactions among variables. The second contribution, addressing the semi-supervised uni-dimensional streaming classification problem, consists of a novel method (CPL-DS) for classifying partially labeled data streams. Data streams differ from the stationary data sets by their highly rapid generation process and their concept-drifting aspect. That is, the learned concepts and/or the underlying distribution are likely changing and evolving over time, which makes the current classification model out-of-date requiring to be updated. CPL-DS uses the Kullback-Leibler divergence and bootstrapping method to quantify and detect three possible kinds of drift: feature, conditional or dual. Then, if any occurs, a new classification model is learned using the expectation-maximization algorithm; otherwise, the current classification model is kept unchanged. CPL-DS is general as it can be applied to several classification models. Using two different models, namely, naive Bayes classifier and logistic regression, CPL-DS is tested with synthetic data streams and applied to the real-world problem of malware detection, where the new received files should be continuously classified into malware or goodware. Experimental results show that our approach is effective for detecting different kinds of drift from partially labeled data streams, as well as having a good classification performance. Finally, the third contribution, addressing the supervised multi-dimensional streaming classification problem, consists of two adaptive methods, namely, Locally Adaptive-MB-MBC (LA-MB-MBC) and Globally Adaptive-MB-MBC (GA-MB-MBC). Both methods monitor the concept drift over time using the average log-likelihood score and the Page-Hinkley test. Then, if a drift is detected, LA-MB-MBC adapts the current multi-dimensional Bayesian network classifier locally around each changed node, whereas GA-MB-MBC learns a new multi-dimensional Bayesian network classifier from scratch. Experimental study carried out using synthetic multi-dimensional data streams shows the merits of both proposed adaptive methods.

Artificial pancreas using a personalized rule-based controller achieves overnight normoglycemia in patients with type 1 diabetes

Objective: This study assessed the efficacy of a closed-loop (CL) system consisting of a predictive rule-based algorithm (pRBA) on achieving nocturnal and postprandial normoglycemia in patients with type 1 diabetes mellitus (T1DM). The algorithm is personalized for each patient’s data using two different strategies to control nocturnal and postprandial periods. Research Design and Methods: We performed a randomized crossover clinical study in which 10 T1DM patients treated with continuous subcutaneous insulin infusion (CSII) spent two nonconsecutive nights in the research facility: one with their usual CSII pattern (open-loop [OL]) and one controlled by the pRBA (CL). The CL period lasted from 10 p.m. to 10 a.m., including overnight control, and control of breakfast. Venous samples for blood glucose (BG) measurement were collected every 20 min. Results: Time spent in normoglycemia (BG, 3.9–8.0 mmol/L) during the nocturnal period (12 a.m.–8 a.m.), expressed as median (interquartile range), increased from 66.6% (8.3–75%) with OL to 95.8% (73–100%) using the CL algorithm (P<0.05). Median time in hypoglycemia (BG, <3.9 mmol/L) was reduced from 4.2% (0–21%) in the OL night to 0.0% (0.0–0.0%) in the CL night (P<0.05). Nine hypoglycemic events (<3.9 mmol/L) were recorded with OL compared with one using CL. The postprandial glycemic excursion was not lower when the CL system was used in comparison with conventional preprandial bolus: time in target (3.9–10.0 mmol/L) 58.3% (29.1–87.5%) versus 50.0% (50–100%). Conclusions: A highly precise personalized pRBA obtains nocturnal normoglycemia, without significant hypoglycemia, in T1DM patients. There appears to be no clear benefit of CL over prandial bolus on the postprandial glycemia

Glucose-Insulin regulator for type 1 diabetes using high order neural networks

In this paper a Glucose-Insulin regulator for Type 1 Diabetes using artificial neural networks (ANN) is proposed. This is done using a discrete recurrent high order neural network in order to identify and control a nonlinear dynamical system which represents the pancreas? beta-cells behavior of a virtual patient. The ANN which reproduces and identifies the dynamical behavior system, is configured as series parallel and trained on line using the extended Kalman filter algorithm to achieve a quickly convergence identification in silico. The control objective is to regulate the glucose-insulin level under different glucose inputs and is based on a nonlinear neural block control law. A safety block is included between the control output signal and the virtual patient with type 1 diabetes mellitus. Simulations include a period of three days. Simulation results are compared during the overnight fasting period in Open-Loop (OL) versus Closed- Loop (CL). Tests in Semi-Closed-Loop (SCL) are made feedforward in order to give information to the control algorithm. We conclude the controller is able to drive the glucose to target in overnight periods and the feedforward is necessary to control the postprandial period.