Digital health technologies: opportunities and challenges in rheumatology.

The past decade in rheumatology has seen tremendous innovation in digital health technologies, including the electronic health record, virtual visits, mobile health, wearable technology, digital therapeutics, artificial intelligence and machine learning. The increased availability of these technologies offers opportunities for improving important aspects of rheumatology, including access, outcomes, adherence and research. However, despite its growth in some areas, particularly with non-health-care consumers, digital health technology has not substantially changed the delivery of rheumatology care. This Review discusses key barriers and opportunities to improve application of digital health technologies in rheumatology. Key topics include smart design, voice enablement and the integration of electronic patient-reported outcomes. Smart design involves active engagement with the end users of the technologies, including patients and clinicians through focus groups, user testing sessions and prototype review. Voice enablement using voice assistants could be critical for enabling patients with hand arthritis to effectively use smartphone apps and might facilitate patient engagement with many technologies. Tracking many rheumatic diseases requires frequent monitoring of patient-reported outcomes. Current practice only collects this information sporadically, and rarely between visits. Digital health technology could enable patient-reported outcomes to inform appropriate timing of face-to-face visits and enable improved application of treat-to-target strategies. However, best practice standards for digital health technologies do not yet exist. To achieve the potential of digital health technology in rheumatology, rheumatology professionals will need to be more engaged upstream in the technology design process and provide leadership to effectively incorporate the new tools into clinical care.

Label-less Learning for Emotion Cognition.

In this paper, we propose a label-less learning for emotion cognition (LLEC) to achieve the utilization of a large amount of unlabeled data. We first inspect the unlabeled data from two perspectives, i.e., the feature layer and the decision layer. By utilizing the similarity model and the entropy model, this paper presents a hybrid label-less learning that can automatically label data without human intervention. Then, we design an enhanced hybrid label-less learning to purify the automatic labeled data. To further improve the accuracy of emotion detection model and increase the utilization of unlabeled data, we apply enhanced hybrid label-less learning for multimodal unlabeled emotion data. Finally, we build a real-world test bed to evaluate the LLEC algorithm. The experimental results show that the LLEC algorithm can improve the accuracy of emotion detection significantly.

Learning with Precise Spike Times: A New Decoding Algorithm for Liquid State Machines.

There is extensive evidence that biological neural networks encode information in the precise timing of the spikes generated and transmitted by neurons, which offers several advantages over rate-based codes. Here we adopt a vector space formulation of spike train sequences and introduce a new liquid state machine (LSM) network architecture and a new forward orthogonal regression algorithm to learn an input-output signal mapping or to decode the brain activity. The proposed algorithm uses precise spike timing to select the presynaptic neurons relevant to each learning task. We show that using precise spike timing to train the LSM and selecting the readout presynaptic neurons leads to a significant increase in performance on binary classification tasks, in decoding neural activity from multielectrode array recordings, as well as in a speech recognition task, compared with what is achieved using the standard architecture and training methods.

The Internet of Things Is Digitizing and Transforming Science.

Over the past decade, cloud software has transformed numerous industries-from finance to logistics, marketing to manufacturing. The simplified aggregation of data, enabled by cloud computing, empowers individuals to glean insights and make data-driven decisions rapidly. In science, however, such a transformation has yet to emerge. The domain lacks centralized, machine-readable repositories of scientific data; this absence inhibits analytics and expedient decision-making. Recently, the Internet of Things (IoT) has served as a catalyst for digitizing and automating science. IoT enables the centralized collection and analysis of scientific data (e.g., instruments, sensors, and environments). Here, we discuss this new technology trend, its applications in laboratories and promise as a platform for improved efficiency, more innovative capabilities, and machine learning/artificial intelligence.

Evaluation of the computational capabilities of a memristive random network (MN3) under the context of reservoir computing.

This work presents the simulation results of a novel recurrent, memristive neuromorphic architecture, the MN3 and explores its computational capabilities in the performance of a temporal pattern recognition task by considering the principles of the reservoir computing approach. A simple methodology based on the definitions of ordered and chaotic dynamical systems was used to determine the separation and fading memory properties of the architecture. The results show the potential use of this architecture as a reservoir for the on-line processing of time-varying inputs.

Development of sEMG sensors and algorithms for silent speech recognition.

OBJECTIVE: Speech is among the most natural forms of human communication, thereby offering an attractive modality for human-machine interaction through automatic speech recognition (ASR). However, the limitations of ASR-including degradation in the presence of ambient noise, limited privacy and poor accessibility for those with significant speech disorders-have motivated the need for alternative non-acoustic modalities of subvocal or silent speech recognition (SSR). APPROACH: We have developed a new system of face- and neck-worn sensors and signal processing algorithms that are capable of recognizing silently mouthed words and phrases entirely from the surface electromyographic (sEMG) signals recorded from muscles of the face and neck that are involved in the production of speech. The algorithms were strategically developed by evolving speech recognition models: first for recognizing isolated words by extracting speech-related features from sEMG signals, then for recognizing sequences of words from patterns of sEMG signals using grammar models, and finally for recognizing a vocabulary of previously untrained words using phoneme-based models. The final recognition algorithms were integrated with specially designed multi-point, miniaturized sensors that can be arranged in flexible geometries to record high-fidelity sEMG signal measurements from small articulator muscles of the face and neck. MAIN RESULTS: We tested the system of sensors and algorithms during a series of subvocal speech experiments involving more than 1200 phrases generated from a 2200-word vocabulary and achieved an 8.9%-word error rate (91.1% recognition rate), far surpassing previous attempts in the field. SIGNIFICANCE: These results demonstrate the viability of our system as an alternative modality of communication for a multitude of applications including: persons with speech impairments following a laryngectomy; military personnel requiring hands-free covert communication; or the consumer in need of privacy while speaking on a mobile phone in public.

Automatic speech recognition: A primer for speech-language pathology researchers.

Automatic speech recognition (ASR) is increasingly becoming an integral component of our daily lives. This trend is in large part due to recent advances in machine learning, and specifically in deep learning, that have led to accurate ASR across numerous tasks. This has led to renewed interest in providing technological support to populations whose speech patterns are atypical, including identifying the presence of a specific pathology and its severity, comparing speech characteristics before and after a surgery and enhancing the quality of life of individuals with speech pathologies. The purpose of this primer is to bring readers with relatively little technical background up to speed on fundamentals and recent advances in ASR. It presents a detailed account of the anatomy of modern ASR, with examples of how it has been used in speech-language pathology research.

Automated conversation system before pediatric primary care visits: a randomized trial.

BACKGROUND AND OBJECTIVES: Interactive voice response systems integrated with electronic health records have the potential to improve primary care by engaging parents outside clinical settings via spoken language. The objective of this study was to determine whether use of an interactive voice response system, the Personal Health Partner (PHP), before routine health care maintenance visits could improve the quality of primary care visits and be well accepted by parents and clinicians. METHODS: English-speaking parents of children aged 4 months to 11 years called PHP before routine visits and were randomly assigned to groups by the system at the time of the call. Parents' spoken responses were used to provide tailored counseling and support goal setting for the upcoming visit. Data were transferred to the electronic health records for review during visits. The study occurred in an urban hospital-based pediatric primary care center. Participants were called after the visit to assess (1) comprehensiveness of screening and counseling, (2) assessment of medications and their management, and (3) parent and clinician satisfaction. RESULTS: PHP was able to identify and counsel in multiple areas. A total of 9.7% of parents responded to the mailed invitation. Intervention parents were more likely to report discussing important issues such as depression (42.6% vs 25.4%; P < .01) and prescription medication use (85.7% vs 72.6%; P = .04) and to report being better prepared for visits. One hundred percent of clinicians reported that PHP improved the quality of their care. CONCLUSIONS: Systems like PHP have the potential to improve clinical screening, counseling, and medication management.

Exploration of an automated approach for receiving patient feedback after outpatient acute care visits.

BACKGROUND: To improve and learn from patient outcomes, particularly under new care models such as Accountable Care Organizations and Patient-Centered Medical Homes, requires establishing systems for follow-up and feedback. OBJECTIVE: To provide post-visit feedback to physicians on patient outcomes following acute care visits. DESIGN: A three-phase cross-sectional study [live follow-up call three weeks after acute care visits (baseline), one week post-visit live call, and one week post-visit interactive voice response system (IVRS) call] with three patient cohorts was conducted. A family medicine clinic and an HIV clinic participated in all three phases, and a cerebral palsy clinic participated in the first two phases. Patients answered questions about symptom improvement, medication problems, and interactions with the healthcare system. PATIENTS: A total of 616 patients were included: 142 from Phase 1, 352 from Phase 2 and 122 from Phase 3. MAIN MEASURES: Primary outcomes included: problem resolution, provider satisfaction with the system, and comparison of IVRS with live calls made by research staff. KEY RESULTS: During both live follow-up phases, at least 96% of patients who were reached completed the call compared to only 48% for the IVRS phase. At baseline, 98 of 113 (88%) patients reported improvement, as well as 167 of 196 (85%) in the live one-week follow-up. In the one-week IVRS phase, 25 of 39 (64%) reported improvement. In all phases, the majority of patients in both the improved and unimproved groups had not contacted their provider or another provider. While 63% of providers stated they wanted to receive patient feedback, they varied in the extent to which they used the feedback reports. CONCLUSIONS: Many patients who do not improve as expected do not take action to further address unresolved problems. Systematic follow-up/feedback mechanisms can potentially identify and connect such patients to needed care.

Alcohol, moods and male-female differences: daily interactive voice response over 6 months.

AIMS: The goal of this study was to better understand the predictive relationship in both directions between negative (anger, sadness) and positive (happiness) moods and alcohol consumption using daily process data among heavy drinkers. METHODS: Longitudinal daily reports of moods, alcohol use and other covariates such as level of stress were assessed over 180 days using interactive voice response telephone technology. Participants were heavy drinkers (majority meeting criteria for alcohol dependence at baseline) recruited through their primary care provider. The sample included 246 (166 men, 80 women) mostly Caucasian adults. Longitudinal statistical models were used to explore the varying associations between number of alcoholic drinks and mood scores the next day and vice versa with gender as a moderator. RESULTS: Increased alcohol use significantly predicted decreased happiness the next day (P < 0.005), more strongly for females than males. Increased anger predicted higher average alcohol use the next day for males only (P < 0.005). CONCLUSION: This daily process study challenges the notion that alcohol use enhances positive mood for both males and females. Our findings also suggest a strong association between anger and alcohol use that is specific to males. Thus, discussions about the effects of drinking on one's feeling of happiness may be beneficial for males and females as well as anger interventions may be especially beneficial for heavy-drinking males.

Improving older adults' experience with interactive voice response systems.

BACKGROUND: Interactive voice response (IVR) systems use computer-based voice recognition and software algorithms to conduct human/computer interactions. In recent years, there has been a proliferation of IVR applications in business and healthcare. The available evidence suggests that older people have negative attitudes towards IVR and experience significant difficulties using these systems. OBJECTIVE: The goal of this project was to identify areas of difficulties in IVR use by older people and propose strategies for improvement. MATERIALS AND METHODS: During two focus groups, we examined older people's perceptions of IVR systems and the most common difficulties experienced by seniors in interacting with these systems. We also recorded their suggestions for improvement of IVR. RESULTS: Frequency and chi square analyses were performed on the focus groups data. Some of the difficulties reported by participants in this study were congruent with previous findings, but we also uncovered some additional problems, such as frustration for not being able to reach an operator, being asked to wait too long on hold, being unable to recover from mistakes, and an absence of shortcuts in the systems. In addition, significant number of participants indicated that they prefer a system that adjusts to them automatically as opposed to a system that allows for adjustment. CONCLUSION: Generally, our findings suggest that the poor acceptability of IVR systems by older people could be improved by designing IVR algorithms that detect difficulties during an ongoing IVR exchange and direct people to different algorithms adapted for each person.

SACA: Software Assisted Call Analysis--an interactive tool supporting content exploration, online guidance and quality improvement of counseling dialogues.

OBJECTIVE: Nearly 30,000 individual inquiries are answered annually by the telephone cancer information service (CIS, KID) of the German Cancer Research Center (DKFZ). The aim was to develop a tool for evaluating these calls, and to support the complete counseling process interactively. METHODS: A novel software tool is introduced, based on a structure similar to a music score. Treating the interaction as a "duet", guided by the CIS counselor, the essential contents of the dialogue are extracted automatically. For this, "trained speech recognition" is applied to the (known) counselor's part, and "keyword spotting" is used on the (unknown) client's part to pick out specific items from the "word streams". The outcomes fill an abstract score representing the dialogue. RESULTS: Pilot tests performed on a prototype of SACA (Software Assisted Call Analysis) resulted in a basic proof of concept: Demographic data as well as information regarding the situation of the caller could be identified. CONCLUSION: The study encourages following up on the vision of an integrated SACA tool for supporting calls online and performing statistics on its knowledge database offline. PRACTICE IMPLICATIONS: Further research perspectives are to check SACA's potential in comparison with established interaction analysis systems like RIAS.

Comparação dos tempos de geração e digitação de laudos radiológicos entre um sistema eletrônico baseado em voz sobre IP (VoIP) e um sistema tradicional baseado em papel / Comparison between the time required to generate and type radiology reports by means of VoIP-based system and a traditional paper-based system

OBJETIVO: Comparar os tempos de geração e digitação de laudos radiológicos entre um sistema eletrônico baseado na tecnologia de voz sobre o protocolo de internet (VoIP) e o sistema tradicional, em que o radiologista escreve o laudo à mão. MATERIAIS E MÉTODOS: Foi necessário modelar, construir e implantar o sistema eletrônico proposto, capaz de gravar o laudo em formato de áudio digital, e compará-lo com o tradicional já existente. Por meio de formulários, radiologistas e digitadores anotaram os tempos de geração e digitação dos laudos nos dois sistemas. RESULTADOS: Comparadas as médias dos tempos entre os sistemas, o eletrônico apresentou redução de 20 por cento (p = 0,0410) do tempo médio de geração do laudo em comparação com o sistema tradicional. O tradicional foi mais eficiente em relação ao tempo de digitação, uma vez que a média de tempo do eletrônico foi três vezes maior (p < 0,0001). CONCLUSÃO: Os resultados mostraram diferença estatisticamente significante entre os sistemas comparados, sendo que o eletrônico foi mais eficiente do que o tradicional em relação ao tempo de geração dos laudos, porém, em relação ao tempo de digitação, o tradicional apresentou melhores resultados.

OBJECTIVE: To compare the time required for generation and typing of radiology reports by means of an electronic system based on the technology of voice over internet protocol (VoIP) and the traditional system, in which the report is handwritten by the radiologist. MATERIALS AND METHODS: It was necessary to model, build and deploy the proposed electronic system, capable of recording the reports in a digital audio format and comparing it with the traditional method. Radiologists and transcriptionists recorded the reports generation and typing times for both systems, using appropriate forms. RESULTS: When the mean times between both systems were compared, those from the electronic system presented a reduction of 20 percent (p = 0.0410) in the report generation time as compared with the traditional method. On the other hand, the traditional method was more efficient with respect to typing time, as the mean typing time with the electronic system was three times longer (p < 0.0001). CONCLUSION: The results demonstrated a statistically significant difference between the compared systems, with the electronic system being more efficient than the traditional one with respect to report generation time, while the traditional method presented better results with respect to typing time.