Learning data-driven models of non-verbal behaviors for building rapport using an intelligent virtual agent

There is a growing societal need to address the increasing prevalence of behavioral health issues, such as obesity, alcohol or drug use, and general lack of treatment adherence for a variety of health problems. The statistics, worldwide and in the USA, are daunting. Excessive alcohol use is the third leading preventable cause of death in the United States (with 79,000 deaths annually), and is responsible for a wide range of health and social problems. On the positive side though, these behavioral health issues (and associated possible diseases) can often be prevented with relatively simple lifestyle changes, such as losing weight with a diet and/or physical exercise, or learning how to reduce alcohol consumption. Medicine has therefore started to move toward finding ways of preventively promoting wellness, rather than solely treating already established illness.^ Evidence-based patient-centered Brief Motivational Interviewing (BMI) interventions have been found particularly effective in helping people find intrinsic motivation to change problem behaviors after short counseling sessions, and to maintain healthy lifestyles over the long-term. Lack of locally available personnel well-trained in BMI, however, often limits access to successful interventions for people in need. To fill this accessibility gap, Computer-Based Interventions (CBIs) have started to emerge. Success of the CBIs, however, critically relies on insuring engagement and retention of CBI users so that they remain motivated to use these systems and come back to use them over the long term as necessary.^ Because of their text-only interfaces, current CBIs can therefore only express limited empathy and rapport, which are the most important factors of health interventions. Fortunately, in the last decade, computer science research has progressed in the design of simulated human characters with anthropomorphic communicative abilities. Virtual characters interact using humans’ innate communication modalities, such as facial expressions, body language, speech, and natural language understanding. By advancing research in Artificial Intelligence (AI), we can improve the ability of artificial agents to help us solve CBI problems.^ To facilitate successful communication and social interaction between artificial agents and human partners, it is essential that aspects of human social behavior, especially empathy and rapport, be considered when designing human-computer interfaces. Hence, the goal of the present dissertation is to provide a computational model of rapport to enhance an artificial agent’s social behavior, and to provide an experimental tool for the psychological theories shaping the model. Parts of this thesis were already published in [LYL+12, AYL12, AL13, ALYR13, LAYR13, YALR13, ALY14].^

Préférences musicales alternatives et ajustement psychosocial à la fin de l’adolescence : une approche développementale semi-paramétrique

Essai doctoral présenté à la Faculté des études supérieures en vue de l'obtention du grade de Docteur en psychologie (D. Psy,), option clinique

Oral cancer incidence and survival rates in the Republic of Ireland, 1994-2009

Background: Oral cancer is a significant public health problem world-wide and exerts high economic, social, psychological, and physical burdens on patients, their families, and on their primary care providers. We set out to describe the changing trends in incidence and survival rates of oral cancer in Ireland between 1994 and 2009. Methods: National data on incident oral cancers [ICD 10 codes C01-C06] were obtained from the National Cancer Registry Ireland from 1994 to 2009. We estimated annual percentage change (APC) in oral cancer incidence during 1994–2009 using joinpoint regression software (version 4.2.0.2). The lifetime risk of oral cancer to age 79 was estimated using Irish incidence and population data from 2007 to 2009. Survival rates were also examined using Kaplan-Meier curves and Cox proportional hazard models to explore the influence of several demographic/lifestyle covariates with follow-up to end 2012. Results: Data were obtained on 2,147 oral cancer incident cases. Men accounted for two-thirds of oral cancer cases (n = 1,430). Annual rates in men decreased significantly during 1994–2001 (APC = -4.8 %, 95 % CI: −8.7 to −0.7) and then increased moderately (APC = 2.3 %, 95 % CI: −0.9 to 5.6). In contrast, annual incidence increased significantly in women throughout the study period (APC = 3.2 %, 95 % CI: 1.9 to 4.6). There was an elevated risk of death among oral cancer patients who were: older than 60 years of age; smokers; unemployed or retired; those living in the most deprived areas; and those whose tumour was sited in the base of the tongue. Being married and diagnosed in more recent years were associated with reduced risk of death. Conclusion: Oral cancer increased significantly in both sexes between 1999 and 2009 in Ireland. Our analyses demonstrate the influence of measured factors such as smoking, time of diagnosis and age on observed trends. Unmeasured factors such as alcohol use, HPV and dietary factors may also be contributing to increased trends. Several of these are modifiable risk factors which are crucial for informing public health policies, and thus more research is needed.

Préférences musicales alternatives et ajustement psychosocial à la fin de l’adolescence : une approche développementale semi-paramétrique

Essai doctoral présenté à la Faculté des études supérieures en vue de l'obtention du grade de Docteur en psychologie (D. Psy,), option clinique

Uso de dexmedetomidina como tratamiento coadyuvante de síndrome de abstinencia alcohólica: revisión sistemática

El objetivo de este estudio es establecer si la dexmedetomidina (DEX) es segura y efectiva para el manejo coadyuvante de síndrome de abstinencia a alcohol (SAA) a través de la búsqueda de evidencia científica. Metodología: se realiza una revisión sistemática de literatura publicada y no publicada desde enero de 1989 hasta febrero 2016 en PubMed, Embase, Scopus, Bireme, Cochrane library y en otras bases de datos y portales. Los criterios de inclusión fueron ensayos clínicos aleatorizados y no aleatorizados, estudios cuasi-experimentales, estudios de cohorte, y estudios de casos y controles; que incluyeron pacientes mayores de 18 años hospitalizados con diagnóstico de SAA y donde se usó DEX como terapia coadyuvante. Resultados: 7 estudios, 477 pacientes, se incluyeron en el análisis final. Se encontraron dos ensayos clínicos aleatorizados, tres estudios de casos y controles y dos estudios de cohorte retrospectivo. Solo uno de los estudios fue doble ciego y utilizó placebo como comparador. Análisis y conclusiones: en los estudios experimentales se determinan que el uso de DEX como terapia coadyuvante en el manejo de SAA tiene significancia clínica y estadística para disminuir dosis de BZD en las primeras 24 horas de tratamiento; pero no demostraron tener otros beneficios clínicos. En los estudios no aleatorizados existe consenso que relaciona el uso de DEX con menores dosis de BZD de forma temprana. Recomendaciones: no se recomienda el uso de DEX en SAA de forma rutinaria. Se recomienda usar DEX solo en casos en el que exista evidencia fallo terapéutico a BZD.

Safeguarding rural Australia : addressing masculinity and violence in rural settings. Data report No. 4 - Risky behaviour - misuse of alcohol, illicit drugs, firearms use and abuse, other risky behaviour

This report focuses on our examination of extant data which have been sourced with respect to personally and socially risky behaviour associated with males living in regional and remote Australia . The AIHW (2008: PHE 97:89) defines personally risky behaviour, on the one hand, as working, swimming, boating, driving or operating hazardous machinery while intoxicated with alcohol or an illicit drug. Socially risky behaviour, on the other hand, is defined as creating a public disturbance, damaging property, stealing or verbally or physically abusing someone while intoxicated with alcohol or an illicit drug. Additional commentary resulting from exploration, examination and analyses of secondary data is published online in complementary reports in this series.

Screening for alcohol and drug use in pregnancy

Objective - this study examined the clinical utility and precision of routine screening for alcohol and other drug use among women attending a public antenatal service. Study design - a survey of clients and audit of clinical charts. Participants and setting - clients attending an antenatal clinic of a large tertiary hospital in Queensland, Australia, from October to December 2009. Measurements and findings - data were collected from two sources. First, 32 women who reported use of alcohol or other drugs during pregnancy at initial screening were then asked to complete a full substance use survey. Second, data were collected from charts of 349 new clients who attended the antenatal clinic during the study period. Both sensitivity (86%, 67%) and positive predictive value (100%, 92%) for alcohol and other drug use respectively, were high. Only 15% of surveyed women were uncomfortable about being screened for substance use in pregnancy, yet the chart audit revealed poor staff compliance. During the study period, 25% of clients were either not screened adequately or not at all. Key conclusions and implications for practise - despite recommended universal screening in pregnancy and the apparent acceptance by our participants, alcohol and other drug (A&OD) screening in the antenatal setting remains problematic. Investigation into the reasons behind, and ways to overcome, the low screening rate could improve health outcomes for mothers and children in this at-risk group. Targeted education and training for midwives may form part of the solution as these clinicians have a key role in implementing prevention and early intervention strategies.

Review of Diagnostic Screening Instruments for Alcohol and Other Drug Use and Other Psychiatric Disorders

In recent years there has been a growing recognition that many people with drug or alcohol problems are also experiencing a range of other psychiatric and psychological problems. The presence of concurrent psychiatric or psychological problems is likely to impact on the success of treatment services. These problems vary greatly, from undetected major psychiatric illnesses that meet internationally accepted diagnostic criteria such as those outlined in the Diagnostic and Statistical Manual (DSM-IV) of the American Psychiatric Association (1994), to less defined feelings of low mood and anxiety that do not meet diagnostic criteria but nevertheless impact on an individual’s sense of wellbeing and affect their quality of life. Similarly, the presence of a substance misuse problem among those suffering from a major psychiatric illness, often goes undetected. For example, the use of illicit drugs such as cannabis and amphetamine is higher among those individuals suffering from schizophrenia (Hall, 1992) and the misuse of alcohol in people suffering from schizophrenia is well documented (e.g., Gorelick et al., 1990; Searles et al., 1990; Soyka et al., 1993). High rates of alcohol misuse have also been reported in a number of groups including women presenting for treatment with a primary eating disorder (Holderness, Brooks Gunn, & Warren, 1994), individuals suffering from post-traumatic stress disorder (Seidel, Gusman and Aubueg, 1994), and those suffering from anxiety and depression. Despite considerable evidence of high levels of co-morbidity, drug and alcohol treatment agencies and mainstream psychiatric services often fail to identify and respond to concurrent psychiatric or drug and alcohol problems, respectively. The original review was conducted as a first step in providing clinicians with information on screening and diagnostic instruments that may be used to assess previously unidentified co-morbidity. The current revision was conducted to extend the original review by updating psychometric findings on measures in the original review, and incorporating other frequently used measures that were not previously included. The current revision has included information regarding special populations, specifically Indigenous Australians, older persons and adolescents. The objectives were to: ● update the original review of AOD and psychiatric screening/diagnostic instruments, ● recommend when these instruments should be used, by whom and how they should be interpreted, ● identify limitations and provide recommendations for further research, ● refer the reader to pertinent Internet sites for further information and/or purchasing of assessment instruments.

Use of a rotating disc reactor to investigate the heterogeneously catalysed oxidation of cinnamyl alcohol in toluene and ionic liquids

To evaluate the effect of mass transfer limitations in the three-phase oxidation of cinnamyl alcohol carried out in toluene and an ionic liquid (1-butyl-3-methyl-imidazolium bis(trifluoromethylsulphonyl)imide), studies have been performed in a rotating disc reactor and compared with those carried out in a stirred tank reactor where mass transfer effects are considered negligible. High catalyst efficiencies are found in the stirred tank reactor with the use of both ionic liquid and toluene, although there is a decrease in rate for the ionic liquid reactions. In contrast, internal pore diffusion limits the reaction in both solvents in the rotating disc reactor. This mass transfer resistance reduces the problem of overoxidation of the metal surface when the reaction is carried out in toluene, leading to significantly higher rates of reaction than expected, although at the cost of decreased selectivity.

The extent and nature of family alcohol and drug use: Findings from the Belfast Youth Development Study

Using data from an ongoing longitudinal study of adolescent drug use, this study examines the proportion of teenagers living with parents who are problem alcohol or drug users. Around two percent of parents report high levels of problem drinking and one per cent report problem drug use. If a broader definition of hazardous drinking is used, the proportion of teenagers exposed increases to over 15 per cent. When substance use is examined at a family level (taking account of alcohol and drug use amongst dependent children in addition to that of parents), the proportion of families experiencing some form of substance use is considerable. These findings add further support to the call for increased recognition of the needs of dependent children within adult treatment services when working with parents. Likewise, the reduction of harm to children as a result of parent substance use should be an increasingly important priority for family support services. This is likely to be achieved through the closer integration of addiction and family services.