Search

Data discretization aims to transform a set of continuous features into discrete features, thus simplifying the representation of information and making it easier to understand, use, and explain. In practice, users can take advantage of the discretization process to improve knowledge discovery and data analysis on medical domain problem datasets containing continuous features. However, certain feature values were frequently missing. Many data-mining algorithms cannot handle incomplete datasets. In this study, we considered the use of both discretization and missing-value imputation to process incomplete medical datasets, examining how the order of discretization and missing-value imputation combined influenced performance. The experimental results were obtained using seven different medical domain problem datasets: two discretizers, including the minimum description length principle (MDLP) and ChiMerge; three imputation methods, including the mean/mode, classification and regression tree (CART), and k-nearest neighbor (KNN) methods; and two classifiers, including support vector machines (SVM) and the C4.5 decision tree. The results show that a better performance can be obtained by first performing discretization followed by imputation, rather than vice versa. Furthermore, the highest classification accuracy rate was achieved by combining ChiMerge and KNN with SVM. [ABSTRACT FROM AUTHOR]

Breast cancer is an all too common disease in women, making how to effectively predict it an active research problem. A number of statistical and machine learning techniques have been employed to develop various breast cancer prediction models. Among them, support vector machines (SVM) have been shown to outperform many related techniques. To construct the SVM classifier, it is first necessary to decide the kernel function, and different kernel functions can result in different prediction performance. However, there have been very few studies focused on examining the prediction performances of SVM based on different kernel functions. Moreover, it is unknown whether SVM classifier ensembles which have been proposed to improve the performance of single classifiers can outperform single SVM classifiers in terms of breast cancer prediction. Therefore, the aim of this paper is to fully assess the prediction performance of SVM and SVM ensembles over small and large scale breast cancer datasets. The classification accuracy, ROC, F-measure, and computational times of training SVM and SVM ensembles are compared. The experimental results show that linear kernel based SVM ensembles based on the bagging method and RBF kernel based SVM ensembles with the boosting method can be the better choices for a small scale dataset, where feature selection should be performed in the data pre-processing stage. For a large scale dataset, RBF kernel based SVM ensembles based on boosting perform better than the other classifiers. [ABSTRACT FROM AUTHOR]

Background: Previous research has suggested a link between antidepressants use and the development of cerebrovascular events, but there has never been any study investigating the risk of stroke in obsessive-compulsive disorder (OCD) patients treated with a selective serotonin reuptake inhibitor (SSRI). Methods: A retrospective observational cohort study was conducted using data from the National Health Insurance Database of Taiwan between the year of 2001 and 2009. A total of 527 OCD patients with 412 subjects in the SSRI use group and 115 in the non SSRI use group were included. Multivariable Cox proportional-hazards models were used to explore the associations between SSRI use and the occurrence of stroke, controlling for age, gender, concomitant medications, and comorbid medical illnesses. Results: A total of nineteen OCD patients were diagnosed with new onset of stroke during the follow-up period including six cases in the SSRI group and thirteen in the non SSRI use group. SSRI use was demonstrated to be associated with a decreased risk of stroke (hazard ratio [HR] = 0.30; 95% confidence interval [CI] = 0.10–0.86, P = 0.02). The increase in age-related risk of strokes was 2.55 per decade (HR = 2.55; 95% CI = 1.74–3.75, P<0.001). Alternatively, sex, concomitant use of aspirin and non-steroidal anti-inflammatory drugs, and comorbidities with angina pectoris, diabetes mellitus, hypertension, and hyperlipidemia were not found to be associated with an increased risk for stroke in OCD patients. Conclusions: Our study showed that SSRI use was associated with decreased risk of stroke in OCD patients. Further investigation into the possible biological mechanisms underlying the relationship between stroke and SSRI use in OCD patients is warranted. [ABSTRACT FROM AUTHOR]

Background: Irritable bowel syndrome (IBS) is the most common functional gastrointestinal (GI) disorder observed in patients who visit general practitioners for GI-related complaints. A high prevalence of psychiatric comorbidities, particularly anxiety and depressive disorders, has been reported in patients with IBS. However, a clear temporal relationship between IBS and psychiatric disorders has not been well established. Objective: We explored the relationship between IBS and the subsequent development of psychiatric disorders including schizophrenia, bipolar disorder, depressive disorder, anxiety disorder, and sleep disorder. Methods: We selected patients who were diagnosed with IBS caused by gastroenteritis, according to the data in the Taiwan National Health Insurance Research Database. A comparison cohort was formed of patients without IBS who were matched according to age and sex. The incidence rate and the hazard ratios (HRs) of subsequent new-onset psychiatric disorders were calculated for both cohorts, based on psychiatrist diagnoses. Results: The IBS cohort consisted of 4689 patients, and the comparison cohort comprised 18756 matched control patients without IBS. The risks of depressive disorder (HR = 2.71, 95% confidence interval [CI] = 2.30–3.19), anxiety disorder (HR = 2.89, 95% CI = 2.42–3.46), sleep disorder (HR = 2.47, 95% CI = 2.02–3.02), and bipolar disorder (HR = 2.44, 95% CI = 1.34–4.46) were higher in the IBS cohort than in the comparison cohort. In addition, the incidence of newly diagnosed depressive disorder, anxiety disorder, and sleep disorder remained significantly increased in all of the stratified follow-up durations (0–1, 1–5, ≥5 y). Conclusions: IBS may increase the risk of subsequent depressive disorder, anxiety disorder, sleep disorder, and bipolar disorder. The risk ratios are highest for these disorders within 1 year of IBS diagnosis, but the risk remains statistically significant for more than 5 years. Clinicians should pay particular attention to psychiatric comorbidities in IBS patients. [ABSTRACT FROM AUTHOR]

Background: Polycystic ovary syndrome (PCOS) is one of the most common endocrine disorders among women of reproductive age. A higher prevalence of psychiatric comorbidities, including depressive disorder, anxiety disorder, and bipolar disorder has been proved in patients with PCOS. However, a clear temporal causal relationship between PCOS and psychiatric disorders has not been well established. Objective: We explored the relationship between PCOS and the subsequent development of psychiatric disorders including schizophrenia, bipolar disorder, depressive disorder, anxiety disorder, and sleep disorder. Methods: We identified patients who were diagnosed with PCOS by an obstetrician-gynecologist in the Taiwan National Health Insurance Research Database. A comparison cohort was constructed of patients without PCOS who were matched according to age and sex. The occurrence of subsequent new-onset psychiatric disorders was evaluated in both cohorts based on diagnoses made by psychiatrists. Results: The PCOS cohort consisted of 5431 patients, and the comparison cohort consisted of 21,724 matched control patients without PCOS. The incidence of depressive disorder (hazard ratio [HR] 1.296, 95% confidence interval [CI] 1.084–.550), anxiety disorder (HR 1.392, 95% CI 1.121–1.729), and sleep disorder (HR 1.495, 95% CI 1.176–1.899) were higher among the PCOS patients than among the patients in the comparison cohort. In addition, a higher incidence of newly diagnosed depressive disorder, anxiety disorder, and sleep disorder remained significantly increased in all of the stratified follow-up durations (0–1, 1–5, ≥5 y). Conclusions: PCOS might increase the risk of subsequent newly diagnosed depressive disorder, anxiety disorder, and sleep disorder. The risk of newly diagnosed bipolar disorder, which has often been reported in the literature to be comorbid with PCOS, was not significantly elevated. [ABSTRACT FROM AUTHOR]

Background & Aims: To evaluate the risk of depressive disorders among non-alcoholic patients by using the Taiwan National Health Insurance Research Database (NHIRD). Methods: We conducted a retrospective study of a matched cohort of 52 725 participants (10 545 non-alcoholic cirrhotic patients and 42 180 control patients) who were selected from the NHIRD. Patients were observed for a maximum of 11 years to determine the rates of newly onset depressive disorders, and Cox regression was used to identify the risk factors associated with depressive disorders in cirrhotic patients. Results: During the 11-year follow-up period, 395 (3.75%) non-alcoholic cirrhotic patients and 1 183 (2.80%) control patients were diagnosed with depressive disorders. The incidence risk ratio of depressive disorders between non-alcoholic cirrhotic patients and control patients was 1.76 (95% CI, 1.57–1.98, P<.001). After adjusting for age, sex, and comorbidities, non-alcoholic cirrhotic patients were 1.75 times more likely to develop depressive disorders (95% CI, 1.56–1.96, P<.001) compared with the control patients. The hazard ratios for patients younger than 60 years old (1.31) and female (1.25) indicated that each is an independent risk factor for depressive disorders in non-alcoholic cirrhotic patients. Conclusions: The likelihood of developing depressive disorders is greater among non-alcoholic cirrhotic patients than among patients without cirrhosis. Symptoms of depression should be sought in patients with cirrhosis. [ABSTRACT FROM AUTHOR]

Objective: To evaluate the risk of cancer among patients with generalized anxiety disorder (GAD) in a nationwide population-based dataset. Methods: We recruited newly-diagnosed GAD patients aged 20 years or older without antecedent cancer from the Taiwan National Health Insurance Research database between 2000–2010. Standardized incidence ratios (SIRs) of cancers were calculated in GAD patients, and the subgroup of GAD patients diagnosed by psychiatric specialists. Results: A total of 559 cancers developed among 19,793 GAD patients with a follow-up of 89,485 person-years (median follow-up of 4.34 years), leading to a significantly increased SIR of 1.14 [95% confidence interval (CI) 1.05–1.24]. Male GAD patients had a significantly increased SIR overall (1.30, 95% CI 1.15–1.46) and for lung and prostate cancer (1.77, 95% CI 1.33–2.30 and 2.17, 95% CI 1.56–2.93, respectively). Patients over 80 years of age also had a significantly increased SIR (1.56, 95% CI 1.25–1.92), especially in males. However, psychiatrist-diagnosed GAD patients did not show increased cancer risk relative to the general population, perhaps due to having fewer physical comorbidities than non-psychiatrist-diagnosed GAD patients. Conclusion: This study found that overall cancer risk is elevated among patients with GAD. The risk of lung and prostate cancer also increased in male patients with GAD. This increased cancer risk may be due to physical comorbidities and surveillance bias. Further prospective study is necessary to confirm these findings. [ABSTRACT FROM AUTHOR]