Topic Break Detection in Interview Dialogues Using Sentence Embedding of Utterance and Speech Intention Based on Multitask Neural Networks.

Currently, task-oriented dialogue systems that perform specific tasks based on dialogue are widely used. Moreover, research and development of non-task-oriented dialogue systems are also actively conducted. One of the problems with these systems is that it is difficult to switch topics naturally. In this study, we focus on interview dialogue systems. In an interview dialogue, the dialogue system can take the initiative as an interviewer. The main task of an interview dialogue system is to obtain information about the interviewee via dialogue and to assist this individual in understanding his or her personality and strengths. In order to accomplish this task, the system needs to be flexible and appropriate for detecting topic switching and topic breaks. Given that topic switching tends to be more ambiguous in interview dialogues than in task-oriented dialogues, existing topic modeling methods that determine topic breaks based only on relationships and similarities between words are likely to fail. In this study, we propose a method for detecting topic breaks in dialogue to achieve flexible topic switching in interview dialogue systems. The proposed method is based on multi-task learning neural network that uses embedded representations of sentences to understand the context of the text and utilizes the intention of an utterance as a feature. In multi-task learning, not only topic breaks but also the intention associated with the utterance and the speaker are targets of prediction. The results of our evaluation experiments show that using utterance intentions as features improves the accuracy of topic separation estimation compared to the baseline model.

To guide or to follow? Teaching visual problem solving at the workplace.

Visual problem solving is essential to highly visual and knowledge-intensive professional domains such as clinical pathology, which trainees learn by participating in relevant tasks at the workplace (apprenticeship). Proper guidance of the visual problem solving of apprentices by the master is necessary. Interaction and adaptation to the expertise level of the learner are identified as key ingredients of this guidance. This study focuses on the effect of increased participation of the learner in the task on the interaction and adaptation of the guidance by masters. Thirteen unique dyads consisting of a clinical pathologist (master) and a resident (apprentice) discussed and diagnosed six microscope images. Their dialogues were analysed on their content. The dyads were divided in two groups according to the experience of the apprentice. For each dyad, master and apprentice both operated the microscope for half of the cases. Interaction was operationalised as the equal contribution of both master and apprentice to the dialogue. Adaptation was operationalised as the extent to which the content of the dialogues was adapted to the apprentice's level. The main hypothesis stated that the interaction and adaptation increase when apprentices operate the microscope. Most results confirmed this hypothesis: apprentices contributed more content when participating more and the content of these dialogues better reflected expertise differences of apprentices. Based on these results, it is argued that, for learning visual problem solving in a visual and knowledge-intensive domain, it is not only important to externalise master performance, but also that of the apprentice.

Communication about Sexual Health in Breast Cancer: What Can We Learn from Patients' Self-Report and Clinic Dialogue?

OBJECTIVE: Research assessing clinical communication about sexual health is limited. We compared clinical communication about sexual health across patients' self-reports and coded dialogue in breast cancer outpatients. METHODS: 134 patients had clinic visits audio-recorded and coded for sexual health communication and completed self-report questionnaires immediately after the visit. Associations between the self-report and dialogue were assessed using Phi coefficient. Agreements (present/absent) and discrepancies (omissions, commissions) about discussed topics were classified and discrepancies analyzed for themes. RESULTS: Sexual health was discussed in 61 of 134 patient visits (46%). Associations were significant (p < .01) but differed by topic (φ = .27-.76). 37 women (23%) had ≥ 1 self-report error. Discrepancies were common (19 omissions, 29 commissions). Patients often omitted communication about sexual concerns when such concerns were not problematic, and interpreted non-specific discussions as including specific topics of concern, even when not explicitly stated. Omissions were more common for women with lower education. CONCLUSIONS: Patients' reports of whether sexual health communication occurs does not always align with observed dialogue, and may vary by personal relevance of the topic. PRACTICE IMPLICATIONS: There are limitations in determining the prevalence of clinical communication about sexual health through patient self-report. Explaining sexual health terms might enhance shared understanding.

Protocol for a conversation-based analysis study: PREVENT-ED investigates dialogue features that may help predict dementia onset in later life.

INTRODUCTION: Decreasing the incidence of Alzheimer's disease (AD) is a global public health priority. Early detection of AD is an important requisite for the implementation of prevention strategies towards this goal. While it is plausible that patients at the early stages of AD may exhibit subtle behavioural signs of neurodegeneration, neuropsychological testing seems unable to detect these signs in preclinical AD. Recent studies indicate that spontaneous speech data, which can be collected frequently and naturally, provide good predictors for AD detection in cohorts with a clinical diagnosis. The potential of models based on such data for detecting preclinical AD remains unknown. METHODS AND ANALYSIS: The PREVENT-Elicitation of Dialogues (PREVENT-ED) study builds on the PREVENT Dementia project to investigate whether early behavioural signs of AD may be detected through dialogue interaction. Participants recruited through PREVENT, aged 40-59 at baseline, will be included in this study. We will use speech processing and machine learning methods to assess how well speech and visuospatial markers agree with neuropsychological, biomarker, clinical, lifestyle and genetic data from the PREVENT cohort. ETHICS AND DISSEMINATION: There are no expected risks or burdens to participants. The procedures are not invasive and do not raise significant ethical issues. We only approach healthy consenting adults and all participants will be informed that this is an exploratory study and therefore has no diagnostic aim. Confidentiality aspects such as data encryption and storage comply with the General Data Protection Regulation and with the requirements from sponsoring bodies and ethical committees. This study has been granted ethical approval by the London-Surrey Research Ethics Committee (REC reference No: 18/LO/0860), and by Caldicott and Information Governance (reference No: CRD18048). PREVENT-ED results will be published in peer-reviewed journals.

Patient-clinician communication about sexual health in breast cancer: A mixed-methods analysis of clinic dialogue.

OBJECTIVES: Breast cancer patients' sexual health needs are frequently unmet. We examined the prevalence and content of communication about sexual health between breast cancer patients and clinicians. METHODS: Female breast cancer patients being seen in oncology clinic follow-up had a clinic visit audio recorded and self-reported sexual problems after the visit. Transcripts were coded for sexual health communication; data were analyzed descriptively or using Chi-square tests. RESULTS: We recorded 67 patients (81% participation rate) interacting with one of 7 cancer clinicians (5 physicians; 2 advanced practice clinicians). Approximately 1/3 of women (n = 22) reported sexual problems; sexual health communication occurred with 10/22 of those women (45%). Across all 67 patients, 27 (40%) visits contained communication about sexual health. Seventy-percent of sexual health communication was clinician-initiated. When in-depth sexual health discussions occurred, the most common topics discussed were sexual inactivity (6), body image (5), vaginal dryness (4), and safety of sexual activity (2). CONCLUSION: Communication about sexual health was uncommon even for women reporting sexual problems and was largely initiated by clinicians. PRACTICE IMPLICATIONS: Because women with breast cancer often do not raise sexual concerns during clinic visits, breast cancer clinicians should raise the topic of sexual health for all patients.

Evaluating Surgical Coaching: A Mixed Methods Approach Reveals More Than Surveys Alone.

OBJECTIVE: Traditionally, surgical educators have relied upon participant survey data for the evaluation of educational interventions. However, the ability of such subjective data to completely evaluate an intervention is limited. Our objective was to compare resident and attending surgeons' self-assessments of coaching sessions from surveys with independent observations from analysis of intraoperative and postoperative coaching transcripts. DESIGN: Senior residents were video-recorded operating. Each was then coached by the operative attending in a 1:1 video review session. Teaching points made in the operating room (OR) and in post-OR coaching sessions were coded by independent observers using dialogue analysis then compared using t-tests. Participants were surveyed regarding the degree of teaching dedicated to specific topics and perceived changes in teaching level, resident comfort, educational assessments, and feedback provision between the OR and the post-OR coaching sessions. SETTING: A single, large, urban, tertiary-care academic institution. PARTICIPANTS: Ten PGY4 to 5 general surgery residents and 10 attending surgeons. RESULTS: Although the reported experiences of teaching and coaching sessions by residents and faculty were similar (Pearson correlation coefficient = 0.88), these differed significantly from independent observations. Observers found that residents initiated a greater proportion of teaching points and had more educational needs assessments during coaching, compared to the OR. However, neither residents nor attendings reported a change between the 2 environments with regard to needs assessments nor comfort with asking questions or making suggestions. The only metric on which residents, attendings, and observers agreed was the provision of feedback. CONCLUSIONS: Participants' perspectives, although considered highly reliable by traditional metrics, rarely aligned with analysis of the associated transcripts from independent observers. Independent observation showed a distinct benefit of coaching in terms of frequency and type of learning points. These findings highlight the importance of seeking different perspectives, data sources, and methodologies when evaluating clinical education interventions. Surgical education can benefit from increased use of dialogue analyses performed by independent observers, which may represent a viewpoint distinct from that obtained by survey methodology.

Korean anaphora recognition system to develop healthcare dialogue-type agent.

OBJECTIVES: Anaphora recognition is a process to identify exactly which noun has been used previously and relates to a pronoun that is included in a specific sentence later. Therefore, anaphora recognition is an essential element of a dialogue agent system. In the current study, all the merits of rule-based, machine learning-based, semantic-based anaphora recognition systems were combined to design and realize a new hybrid-type anaphora recognition system with an optimum capacity. METHODS: Anaphora recognition rules were encoded on the basis of the internal traits of referred expressions and adjacent contexts to realize a rule-based system and to serve as a baseline. A semantic database, related to predicate instances of sentences including referred expressions, was constructed to identify semantic co-relationships between the referent candidates (to which semantic tags were attached) and the semantic information of predicates. This approach would upgrade the anaphora recognition system by reducing the number of referent candidates. Additionally, to realize a machine learning-based system, an anaphora recognition model was developed on the basis of training data, which indicated referred expressions and referents. The three methods were further combined to develop a new single hybrid-based anaphora recognition system. RESULTS: The precision rate of the rule-based systems was 54.9%. However, the precision rate of the hybrid-based system was 63.7%, proving it to be the most efficient method. CONCLUSIONS: The hybrid-based method, developed by the combination of rule-based and machine learning-based methods, represents a new system with enhanced functional capabilities as compared to other pre-existing individual methods.

Korean Anaphora Recognition System to Develop Healthcare Dialogue-Type Agent / 대한의료정보학회지

OBJECTIVES: Anaphora recognition is a process to identify exactly which noun has been used previously and relates to a pronoun that is included in a specific sentence later. Therefore, anaphora recognition is an essential element of a dialogue agent system. In the current study, all the merits of rule-based, machine learning-based, semantic-based anaphora recognition systems were combined to design and realize a new hybrid-type anaphora recognition system with an optimum capacity. METHODS: Anaphora recognition rules were encoded on the basis of the internal traits of referred expressions and adjacent contexts to realize a rule-based system and to serve as a baseline. A semantic database, related to predicate instances of sentences including referred expressions, was constructed to identify semantic co-relationships between the referent candidates (to which semantic tags were attached) and the semantic information of predicates. This approach would upgrade the anaphora recognition system by reducing the number of referent candidates. Additionally, to realize a machine learning-based system, an anaphora recognition model was developed on the basis of training data, which indicated referred expressions and referents. The three methods were further combined to develop a new single hybrid-based anaphora recognition system. RESULTS: The precision rate of the rule-based systems was 54.9%. However, the precision rate of the hybrid-based system was 63.7%, proving it to be the most efficient method. CONCLUSIONS: The hybrid-based method, developed by the combination of rule-based and machine learning-based methods, represents a new system with enhanced functional capabilities as compared to other pre-existing individual methods.

Medical Jargon Used in Health Care Communication of Family Physician / 가정의학회지

BACKGROUND: When physicians use the medical terms difficult to understand the meaning, patients regard it as an authority of the physician. It has negative influence on physician-patient relationship and hinders delivering accurate meaning to the patient. The purpose of this research is to investigate what kind of medical terms which meaning is difficult to understand (medical jargon) are used in medical interviews and survey the patients' understanding of the medical terms. METHODS: In the preceding study 67 cases of physicians' interviewing with patient were videotaped in the family medicine clinic of a university hospital and they were transcribed from August, 2005 to January, 2007. For this study 60 cases of the transcribed conversations, which interviewing was completely recorded, were assessed for analyzing the content and frequency of medical jargon. The author selected 10 medical terms used in this study's interviews and surveyed the patients' understanding of the medical terms, who visited the family medicine clinic, by questionnaires. RESULTS: In 26 of 60 interviews one or more medical jargon were used by the physician. In 39 of 73 terms (53.4%) the physicians explained meaning of the medical terms to the patient. 213 patients responded to the questionnaires. More than half (median, 55.4%; range, 11.7 to 75.1%) of the respondents expressed that they 'never know' or 'hardly know' the meaning of the medical terms if it was used without explanation. The scores of level of patients' understanding of 10 medical terms significantly increased according to increasing level of income, but the trend is not statistically significant according to increasing age, difference of sex, and level of education. CONCLUSION: The family physicians used medical jargon in 43.3% of medical interviews, and they explained meaning of the medical terms to the patients for more than half of the terms. More than half of the patients in a university hospital family medicine clinic responded that they never knew or hardly knew meaning of the medical terms if they were used without explanation.