Chi Jane Wang

Perinatal mortality and prevalence of major congenital malformations of twins in Taipei City

In order to examine perinatal mortality and prevalence of major congenital malformations in twins, deliveries in four teaching hospitals in Taipei City were studied. Among a total of 73,264 deliveries from October 1985 to June 1989, there were 844 pairs of twins. The zygosity of the twin pairs was determined by sex, placentation and 12 red blood cell antigens. There were 482 MZ and 252 DZ twin pairs identified, but the zygosity of a further 110 twin pairs was indeterminable due to lack of information on plancentation and/or blood types. A total of 4,573 singletons delivered in one study hospital from July 1986 to June 1987 were also studied as controls. The perinatal mortality rate was 7.5% for MZ twins, 1.4% for DZ twins, and 0.7% for singletons. The concordance rate of perinatal death was significantly higher in MZ (60%) than in DZ (0%) twins. The prevalence of major congenital malformations was 2.7% for MZ twins, 1.0% for DZ twins, and 0.6% for singletons. The concordance rate of major congenital malformations was 18% for MZ twins, but no DZ pair was concordant in any major congenital malformation. The concordance rate of facial clefts was 29% for MZ twins. There were 2 sets of conjoined twins giving a prevalence rate of 2.7 per 100,000 deliveries. These findings showing the prevalence of perinatal mortality and major congenital malformation to be highest in MZ twins, intermediate in DZ twins and lowest in singletons, suggest the importance of intrauterine environments in the determination of perinatal mortality and congenital malformations.

Multiple risk factors of hepatocellular carcinoma: A cohort study of 13 737 male adults in Taiwan

In order to explore multiple risk factors of hepatocellular carcinoma (HCC), a total of 13 737 male adult residents in 12 townships were studied for an average follow‐up period of 5.2 years. Sociodemographic characteristics, history of cigarette smoking and alcohol drinking, dietary habits, as well as personal and familial history of chronic liver diseases were obtained through standardized interviews based on structured questionnaires at the recruitment. Blood samples were also collected from 9688 (71%) study subjects and examined for the hepatitis B surface antigen (HBsAg). A total of 60 new HCC cases occurred giving an incidence rate of 83.3 per 100 000 person‐years. Coxs proportional hazards models were used to analyse multiple risk factors of HCC. In addition the HBsAg carrier status which showed a multivariate‐adjusted relative risk of 17.0, cumulative cigarette smoking, alcohol drinking quantity, vegetarian habit and low vegetable consumption were associated with the development of HCC. The multivariate‐adjusted relative risk was 1.8 for those who smoked 26 or more pack‐years of cigarettes compared with non‐smokers, 3.1 for those who drank alcohol 50 mL or more per day compared with those who were non‐drinkers or drank less than 50 mL per day, 2.5 for vegetarians compared with non‐vegetarians, as well as 4.6 and 2.6, respectively, for those who had a weekly vegetable consumption frequency of less than two meals and two to five meals compared with those who had a frequency of six or more meals.

Emotion recognition with consideration of facial expression and physiological signals

An emotion recognition system with consideration of facial expression and physiological signals is proposed in this paper. A specific designed mood induction experiment is performed to collect facial expressing images and physiological signals of subjects. We detected 14 feature points and extracted 12 facial features from facial expression images. Meanwhile, we measure the skin conductivity, finger temperature and heart rate from the subject. Both facial and physiological features are adopted to train the classifiers. Two learning vector quantization (LVQ) neural networks were applied to classify four emotions: love, joy, surprise and fear. Experimental results show the proposed recognition system is able to identify four emotions by facial expressions, physiological signals, and both of them.

A music recommendation system with consideration of personal emotion

Emotions evoked by were analyzed. An estimation of the correlation coefficient was applied to determine features of music that evoke an emotion. These features were then used to train two support vector machines (SVMs) for an individual subject to classify music that evokes happiness, anger, sadness, and peacefulness. The proposed approach can be used to classify music that evokes an emotion and to build a personal emotion-cognitive music recommendation system for an individual subject. Experiment results show the effectiveness of the proposed approach.

Chronological Changes in Genetic Variance and Heritability of Anthropometric Characteristics Among Chinese Twin Infants

In order to examine the chronologic changes in genetic variance and heritability of anthropometric characteristics of Chinese infants in Taiwan, a total of 521 pairs of same-sexed twin neonates given birth in four major general teaching hospitals in Taipei City were studied. Based on the placental pattern and 12 red blood cell antigens, 428 MZ and 93 DZ twin pairs were identified and followed up to the age of one year. There was no significant genetic variance for all anthropometric characteristics adjusted for sex and gestational week before the age of six months. After adjusting for sex and gestational week, a significant genetic variance was observed at the age of six months, with heritability values of 0.51 (weight), 0.63 (head circumference), 0.77 (chest circumference), and 0.53 (arm circumference), as well as at one year, although with considerably lower heritability values. This implies that growth is dynamically determined by both genetic and environmental factors during infancy.

Based on Support Vector Regression for emotion recognition using physiological signals

Facial expression are widely used for emotion recognition. Facial expressions may be expressed differently by different people subjectively, inaccurate results are unavoidable. Nevertheless, physiological reactions are non-autonomic nerves in physiology. The physiological reactions and the corresponding signals are hardly to control while emotions are excited. Therefore, an emotion recognition system with consideration of physiological signals is proposed in this paper. A specific designed mood induction experiment is performed to collect physiological signals of subjects. Five biosensors including electrocardiogram, respiration, galvanic skin responses (GSR), blood volume pulse, and pulse are used. Then a Support Vector Regression (SVR) is used to train three regression curves of three emotions (sad, fear, and pleasure). Experimental results show that the proposed method based on SVR emotion recognition has a good performance in accuracy.

Secular Trend and Associated Factors of Twinning in Taiwan

Delivery records of public hospitals and local health centers were analyzed to investigate the secular trend of multiple births from 1955 to 1990 in Taiwan. Twinning rates decreased gradually from the late 1950s to 1973, slightly increased and remained stable from 1974 to 1980, and then increased strikingly from 1981 to 1990. During the study period, MZ rates were consistently higher than DZ rates except in 1986 when a higher DZ rate was observed. The highest MZ rate of 7.7 per 1,000 deliveries was observed in 1985, while the highest DZ rate of 4.6 per 1,000 deliveries was observed in 1986. The triplet incidence rate also declined from the early 1960s to 1973 and then increased markedly from 1974 to 1990. A case-control study including 482 MZ and 252 DZ twin pairs and 1,496 singletons was carried out in four teaching hospitals in Taipei City between October 1985 and June 1989 to examine factors related to twinning. The zygosity of twin pairs was determined by sex, placentation and red blood cell antigens. Multiple logistic regression analysis showed that the higher the maternal age, the lower the MZ and DZ rates, while the higher the parity, the higher the rates. There was no association of paternal age or of maternal educational level with twinning rates.

Technology-enhanced game-based team learning for improving intake of food groups and nutritional elements

Nutrition is a critical issue for educators, particularly given the unhealthy eating behaviors of many adolescents. While knowledge and self-reflection are important, learners must also be motivated and held accountable for their health behaviors. In order to foster healthy eating, a technology-enhanced approach, game-based team learning (GBTL), is proposed, based on social-interdependence theory. A cloud diet assessment system (CDAS) was designed for automatically providing feedback on the nutritional intake of learners through a meal analysis algorithm. Furthermore, a cloud server hosted a social competitive game which, in addition to in-class team learning activities, allowed teams to compete against each other on the basis of each groups dietary habits. A pre-test post-test quasi-experimental design evaluated the effectiveness of the GBTL group (E1) as compared to a group which received only metacognitive feedback from the CDAS (E2) and a comparison group (C). Female high school participants from three classes were randomly assigned to the three groups (C, n?=?31; E1, n?=?20; E2, n?=?37). Results demonstrate significant improvement for E2 in terms of most food groups (including Dairy, Meats and Protein, Vegetables, and Fruit), as well as for macronutrients, such as calories and dietary fiber, and micronutrients, including Calcium and Vitamin C and B2. Within- and between-group comparisons confirmed the advantage of the E2 group, suggesting that technology-supported GBTL can foster healthy eating habits among adolescents, improving most nutritional elements to nearly 100% the recommended daily intake. Female adolescents seldom receive sufficient intake of dairy, fruits, and vegetables.Social interdependence through technology was used to encourage healthy eating.A cloud-based approach utilized team experiments and a social competitive game.Nutritional intake significantly improved by food group, micro- and macronutrients.Within-group and between-group results support the effectiveness of the design.

The Efficacy of Using Self-Monitoring Diaries in a Weight Loss Program for Chronically Ill Obese Adults in a Rural Area

Background: Self-monitoring is part of many weight-loss programs and is widely accepted as effective. However, there is a lack of research related to the efficacy of various self-monitoring instruments in meeting the needs of individuals with limited mobility or access to healthcare providers, especially those with limited education living in rural settings. Purpose: This study examined the efficacy of using self-monitoring diaries in a weight loss program targeting chronically ill and obese rural-dwelling adults. Methods: A community-based intervention program using a pretest and posttest design examined the effect of using self-monitoring diaries on weight loss. Fifty participants were enrolled from the chronic disease clinic of a district health center with limited medical resources in a remote village in southwestern Taiwan. All participants were diagnosed with diabetes and/or hypertension, had body mass indices between 27 and 32 kg/m2, and had a minimum educational level of junior high school. Mean participant age was 43.7 years. Participants were randomly assigned to the intervention or control group. All attended a mandatory 12-week weight loss program. The intervention group received instructions on how to record diet and exercise details in a structured, graphics-based diary provided by the researchers. Body weight and percentage of body fat were measured before and after the program, and data were analyzed by chi-square and ANCOVA. Results: The intervention group significantly lost more weight than the control group (5.7 kg vs. 2.1 kg; p < .05). The participants of 88% in the intervention group lost 5% or more of their baseline weight greater than the 23% in the control group. Both groups achieved the mean of body fat reductions by comparing pretest and posttest. Conclusions/Implications for Practice: Self-monitoring diaries can have a significant impact on weight loss in individuals living in rural communities. Healthcare providers and health promotion agencies can use the suggested checklist method to improve weight loss promotion programs in isolated rural communities with limited medical resources.

Music emotion recognition with consideration of personal preference

In order to discover the relationship between music and the emotion that it may evoke, twenty-one features have been extracted to describe music. A feature selection algorithm called sequential floating forward selection (SFFS) is utilized to find discriminative features. An estimation of the correlation coefficient was applied to determine features of music that evoke an emotion. These features were then used to train two support vector machines (SVMs) for an individual subject to classify music that evokes happiness, anger, sadness, and peacefulness. Experimental results show that the proposed approach can be used to classify music that evokes an emotion for an individual subject with high classification accuracy.