Machine Learning-Based Prediction Models for Cognitive Decline Progression: A Comparative Study in Multilingual Settings Using Speech Analysis.

Background: Mild cognitive impairment (MCI) is a condition commonly associated with dementia. Therefore, early prediction of progression from MCI to dementia is essential for preventing or alleviating cognitive decline. Given that dementia affects cognitive functions like language and speech, detecting disease progression through speech analysis can provide a cost-effective solution for patients and caregivers. Design-Participants: In our study, we examined spontaneous speech (SS) and written Mini Mental Status Examination (MMSE) scores from a 60-patient dataset obtained from the Mugla University Dementia Outpatient Clinic (MUDC) and a 153-patient dataset from the Alzheimer's Dementia Recognition through Spontaneous Speech (ADRess) challenge. Our study, for the first time, analyzed the impact of audio features extracted from SS in distinguishing between different degrees of cognitive impairment using both an Indo-European language and a Turkic language, which exhibit distinct word order, agglutination, noun cases, and grammatical markers. Results: When each machine learning model was tested on its respective trained language, we attained a 95% accuracy using the random forest classifier on the ADRess dataset and a 94% accuracy on the MUDC dataset employing the multilayer perceptron (MLP) neural network algorithm. In our second experiment, we evaluated the effectiveness of each language-specific machine learning model on the dataset of the other language. We achieved accuracies of 72% for English and 76% for Turkish, respectively. Conclusion: These findings underscore the cross-language potential of audio features for automated tracking of cognitive impairment progression in MCI patients, offering a convenient and cost-effective option for clinicians or patients.

Epicardial fat thickness in patients with gestational diabetes mellitus and its association with N-terminal pro-brain natriuretic peptide.

OBJECTIVE: The female population with gestational diabetes mellitus (GDM) has a postpartum profile with increased cardiovascular (CV) risk factors and heightened prospective CV risk. Epicardial fat tissue was reported to be related to cardiometabolic diseases as metabolically active adipose tissue and natriuretic peptides (NPs) have been shown to have metabolic effects. This study's aim was to determine the relationship between epicardial fat thickness (EFT) and NPs in the female population diagnosed with GDM. PATIENTS AND METHODS: The study involved 161 pregnant women: 96 with GDM, and 65 healthy controls. GDM was diagnosed following the American Diabetes Association (2013) norms for diagnosing diabetes. All patients underwent echocardiography to measure EFT. N-terminal pro-brain natriuretic peptide (NT-proBNP) and other parameters were quantified in blood samples. The Independent Samples t-test, Pearson's correlation test, and a multivariable logistic regression analysis (LRA) were performed for statistical evaluation. A p-value <0.05 was considered statistically significant. RESULTS: Fasting (91.46±14.29 mg/dl vs. 82.18±8.21 mg/dl) (p<0.001), first-hour (202.30± 21.60 mg/dl vs. 161.57±16.21 mg/dl) (p<0.001), and second-hour (176.95±20.43 mg/dl vs. 130.93±16.95 mg/dl) glucose levels (p<0.001), fasting insulin level (14.54±3.50 mUL/mL vs. 11.51±2.04 mUL/mL; p<0.001), and Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) value (3.28±0.99 vs. 2.31±0.45; p<0.0001) in GDM Group were significantly higher than Control Group. The GDM Group also had significantly increased EFT values compared to the Control Group (4.74±0.65 mm vs. 3.77±0.66 mm; p<0.0001), whereas NT-proBNP levels of diabetic women were considerably lower compared to controls (21.59±19.86 pg/ml vs. 39.74±33.96 pg/ml; p<0.0001). In addition, EFT thickness and the NT-proBNP level were determined to be significantly negatively correlated (r=-32, p<0.0001). CONCLUSIONS: EFT evaluation might play a prognostic role in detecting cardiometabolic risk associated with potential disorders such as GDM. This study showed a potential interplay between epicardial adipose tissue and NPs secreted by cardiomyocytes and practical effects on fat metabolism in GDM subjects.

THE EFFECT OF EXENATIDE THERAPY IN PREVIOUSLY INSULIN-TREATED TYPE 2 DIABETIC PATIENTS.

OBJECTIVE: To evaluate the effect of multiple daily injection (MDI) treatment replaced by Exenatide BID as compared with continuation of MDI. PATIENTS AND METHODS: A total of 140 patients with type 2 diabetes, taking metformin and multiple daily insulin injections, were randomized to exenatide or insulin group that continued their insulin treatment. Patients were followed-up for 16 weeks. Blood glucose profiles, BMI, waist circumference, HbA1C, serum lipids and side effects were assesssed at weeks 0,12 and 16. RESULTS: There were no significant differences between the two groups with respect to baseline parameters. Glycemic control was similar between the two groups. The mean changes in HbA1C in exenatide group were -0.66±0.63% and in insulin group -0.74±0.92 % (p=0.594). In exenatide group, 59.6 % of patients and in insulin group 85.71 % of patients had maintained or improved glycemic control at the end of the study. In insulin group, insulin requirement increased 5.86 ± 4.46 units/day. Body weight and waist circumference decreased significantly in exenatide treatment group with respect to insulin group (p<0.001). CONCLUSIONS: Substituting exenatide for insulin might be an option in insulin-treated, type 2 diabetic patients having obesity, and poor glycemic control. However, patients with longer duration of diabetes and insulin treatment and with lower C-peptide levels might not benefit from exenatide therapy.

IS SUBCLINICAL HYPOTHYROIDISM IN PATIENTS WITH POLYCYSTIC OVARY SYNDROME ASSOCIATED WITH BMI?

OBJECTIVE: To investigate the impact of body weight on the subclinical hypothyroidism observed in patients with PCOS. METHODS: The study included 95 normal weight (Group-1) and 122 overweight or obese women (Group-2) with PCOS. The control group consisted of age and BMI matched healthy individuals and grouped as normal weight (n: 66, Group-3) and overweight or obese (n: 65, Group-4. Women with chronic disease such as overt thyroid dysfunction, late-onset adrenal hyperplasia, and diabetes were excluded from the study. Plasma glucose and lipid profile, thyroid hormones, insulin, FSH, LH, total testosterone, estradiol, progesterone and DHEA-S were measured. RESULTS: While fasting glucose was similar, insulin and HOMA-IR were higher in Group-2 and Group-4 (p: 0.001). The groups were similar with respect to FSH, Estradiol, prolactine, DHEAS. While total testosterone and LH levels were higher (ptestosterone: 0,009), progesterone was lower in both PCOS groups (pprogesterone: 0.041). Free T3, free T4, thyroid antibodies were similar between the groups, but the prevalence of subclinical hypothyroidism was greater in Group-2 and -4 than in Group-1 and -3 (p: 0.044). TSH was only correlated with BMI (r: 0.122, p: 0.02). CONCLUSION: The increased prevalence of subclinical hypothyroidism in women with PCOS might be the result of increased BMI.

Healthcare information technology infrastructures in Turkey.

OBJECTIVES: The objective of this paper is to describe some of the major healthcare information technology (IT) infrastructures in Turkey, namely, Saglik-Net (Turkish for "Health-Net"), the Centralized Hospital Appointment System, the Basic Health Statistics Module, the Core Resources Management System, and the e-prescription system of the Social Security Institution. International collaboration projects that are integrated with Saglik-Net are also briefly summarized. METHODS: The authors provide a survey of the some of the major healthcare IT infrastructures in Turkey. RESULTS: Saglik-Net has two main components: the National Health Information System (NHIS) and the Family Medicine Information System (FMIS). The NHIS is a nation-wide infrastructure for sharing patients' Electronic Health Records (EHRs). So far, EHRs of 78.9 million people have been created in the NHIS. Similarly, family medicine is operational in the whole country via FMIS. Centralized Hospital Appointment System enables the citizens to easily make appointments in healthcare providers. Basic Health Statistics Module is used for collecting information about the health status, risks and indicators across the country. Core Resources Management System speeds up the flow of information between the headquarters and Provincial Health Directorates. The e-prescription system is linked with Saglik-Net and seamlessly integrated with the healthcare provider information systems. Finally, Turkey is involved in several international projects for experience sharing and disseminating national developments. CONCLUSION: With the introduction of the "Health Transformation Program" in 2003, a number of successful healthcare IT infrastructures have been developed in Turkey. Currently, work is going on to enhance and further improve their functionality.

Polymorphic variants of MnSOD Val16Ala, CAT-262 C < T and GPx1 Pro198Leu genotypes and the risk of laryngeal cancer in a smoking population.

OBJECTIVE: To investigate the relationship between development of laryngeal cancer and the presence of polymorphisms of the MnSOD Val16Ala, CAT-262 C < T and GPx1 Pro198Leu genes in a smoking population. PATIENTS AND METHODS: Single nucleotide polymorphisms were determined in DNA from the peripheral blood erythrocytes of 48 heavy smokers (25 patients with laryngeal cancer and 23 cancer-free controls), using polymerase chain reaction. RESULTS: There were no significant differences in age, smoking duration or smoking intensity, comparing the two groups. The homozygous AA genotype of MnSOD Val16Ala was significantly more prevalent in the cancer group than the control group (92 vs 13 per cent, respectively), while the heterozygous AV genotype of MnSOD Val16Ala was more prevalent in the control group than the cancer group (87 vs 8 per cent, respectively) (p < 0.001). There were no significant differences between the cancer and control groups regarding GPx1 Pro198Leu or CAT-262 C < T polymorphisms. CONCLUSION: Polymorphism of the MnSOD Val16Ala gene may contribute to susceptibility to laryngeal cancer among smokers.

Electronic health record interoperability as realized in the Turkish health information system.

OBJECTIVES: The objective of this paper is to describe the techniques used in developing the National Health Information System of Turkey (NHIS-T), a nation-wide infrastructure for sharing electronic health records (EHRs). METHODS: The UN/CEFACT Core Components Technical Specification (CCTS) methodology was applied to design the logical EHR structure and to increase the reuse of common information blocks in EHRs. RESULTS: The NHIS-T became operational on January 15, 2009. By June 2010, 99% of the public hospitals and 71% of the private and university hospitals were connected to NHIS-T with daily feeds of their patients' EHRs. Out of the 72 million citizens of Turkey, electronic healthcare records of 43 million citizens have already been created in NHIS-T. Currently, only the general practitioners can access the EHRs of their patients. In the second phase of the implementation and once the legal framework is completed, the proper patient consent mechanisms will be available through the personal health record system that is under development. At this time authorized healthcare professionals in secondary and tertiary healthcare systems can access the patients' EHRs. CONCLUSIONS: A number of factors affected the successful implementation of NHIS-T. First, all stakeholders have to adopt the specified standards. Second, the UN/CEFACT CCTS approach was applied which facilitated the development and understanding of rather complex EHR schemas. Finally, the comprehensive testing of vendor-based hospital information systems for their conformance to and interoperability with NHIS-T through an automated testing platform enhanced substantially the fast integration of vendor-based solutions with the NHIS-T.

Cancer and tuberculosis: case series.

Tuberculosis is seen with an increased frequency in cancer patients. Possible reasons of reactivation are thought to be related to chemotherapy and insufficient nutrition together with compromised immune system. The diagnosis of tuberculosis may be missed in cancer patients and may be diagnosed with newly developed radiological and clinical findings during treatment. In this case, tuberculosis should be considered and related diagnostic work up should be completed. Also, PPD test should be applied to cancer patients and if needed isoniazid prophylaxis should be initiated. We present herein 4 cancer patients diagnosed with pulmonary tuberculosis. Two patients suffered from solid malignancies (lung cancer) and 2 from non-solid malignancies (acute myeloid leukemia).

IL-2 and IL-10 levels in induced sputum and serum samples of asthmatics.

BACKGROUND: There is consisting evidence that asthma is associated with airway inflammation. Originally IL-10 and IL-2 were described as lymphokines produced by T cells in mediating cellular infiltration into the airways and continue to be of interest in evaluating asthma pathogenesis. The aim of this study was to evaluate the serum and sputum levels of IL-2 and IL-10 in asthmatic subjects and healthy controls and to correlate disease activity and other clinical indices with concentrations of IL-2 and IL-10 in serum and sputum samples. METHODS: We evaluated cell profiles and IL-2 and IL-10 levels in induced sputum samples and in serum samples of 6 mild, 5 moderate, 7 severe asthmatic patients and 5 healthy controls by using ELISA. RESULTS: The mean IL-2 in sputum samples of asthmatics and controls were 35.3 +/- 13.2 pg/ml and 35.3 +/- 8.4 pg/ml, respectively. The mean IL-2 in serum samples of asthmatics and controls were 42.7 +/- 21.1 pg/ml and 30.3 +/- 2.4 pg/ml, respectively. Both levels did not result in any statistically significant difference between asthmatics and controls. There was no correlation between serum and sputum IL-2 levels, however sputum IL-2 levels correlated with percentage of sputum lymphocytes (p < 0.03, r = 0.51). The mean IL-10 levels in sputum samples of asthmatics and controls were 4.4 +/- 3.3 pg/ml and 3.9 +/- 5.9 pg/ml, respectively, the mean IL-10 level in serum of asthmatics and controls were 4.1 +/- 3.8 pg/ml and 2.3 +/- 2.5 pg/ml, respectively. We could not find statistically significant difference of serum or sputum IL-10 levels between asthmatics and controls. There was only correlation between serum and sputum IL-10 levels in asthmatics (p < 0.0008, r = 0.73). There was no difference between asthmatic subgroups regarding sputum and serum levels of IL-2 and IL-10. No correlation could be demonstrated between sputum or serum IL-2 and IL-10 levels and clinical severity. CONCLUSIONS: We have demonstrated the presence of detectable concentrations of the IL-2 and IL-10 in serums and induced sputum samples of asthmatics, however, they have no predictive value for asthma since their levels are not increased in asthmatic patients over controls. Moreover, IL-2 level positively correlated with lymphocyte percentage in induced sputum. The results suggest that measurement of IL-2 and IL-10 concentrations in serum and sputum will not be of diagnostic use in asthma and a reflection of the severity of asthmatic airway inflammation.