Adapting and Evaluating an AI-Based Chatbot Through Patient and Stakeholder Engagement to Provide Information for Different Health Conditions: Master Protocol for an Adaptive Platform Trial (the MARVIN Chatbots Study).

BACKGROUND: Artificial intelligence (AI)-based chatbots could help address some of the challenges patients face in acquiring information essential to their self-health management, including unreliable sources and overburdened health care professionals. Research to ensure the proper design, implementation, and uptake of chatbots is imperative. Inclusive digital health research and responsible AI integration into health care require active and sustained patient and stakeholder engagement, yet corresponding activities and guidance are limited for this purpose. OBJECTIVE: In response, this manuscript presents a master protocol for the development, testing, and implementation of a chatbot family in partnership with stakeholders. This protocol aims to help efficiently translate an initial chatbot intervention (MARVIN) to multiple health domains and populations. METHODS: The MARVIN chatbots study has an adaptive platform trial design consisting of multiple parallel individual chatbot substudies with four common objectives: (1) co-construct a tailored AI chatbot for a specific health care setting, (2) assess its usability with a small sample of participants, (3) measure implementation outcomes (usability, acceptability, appropriateness, adoption, and fidelity) within a large sample, and (4) evaluate the impact of patient and stakeholder partnerships on chatbot development. For objective 1, a needs assessment will be conducted within the setting, involving four 2-hour focus groups with 5 participants each. Then, a co-construction design committee will be formed with patient partners, health care professionals, and researchers who will participate in 6 workshops for chatbot development, testing, and improvement. For objective 2, a total of 30 participants will interact with the prototype for 3 weeks and assess its usability through a survey and 3 focus groups. Positive usability outcomes will lead to the initiation of objective 3, whereby the public will be able to access the chatbot for a 12-month real-world implementation study using web-based questionnaires to measure usability, acceptability, and appropriateness for 150 participants and meta-use data to inform adoption and fidelity. After each objective, for objective 4, focus groups will be conducted with the design committee to better understand their perspectives on the engagement process. RESULTS: From July 2022 to October 2023, this master protocol led to four substudies conducted at the McGill University Health Centre or the Centre hospitalier de l'Université de Montréal (both in Montreal, Quebec, Canada): (1) MARVIN for HIV (large-scale implementation expected in mid-2024), (2) MARVIN-Pharma for community pharmacists providing HIV care (usability study planned for mid-2024), (3) MARVINA for breast cancer, and (4) MARVIN-CHAMP for pediatric infectious conditions (both in preparation, with development to begin in early 2024). CONCLUSIONS: This master protocol offers an approach to chatbot development in partnership with patients and health care professionals that includes a comprehensive assessment of implementation outcomes. It also contributes to best practice recommendations for patient and stakeholder engagement in digital health research. TRIAL REGISTRATION: ClinicalTrials.gov NCT05789901; https://classic.clinicaltrials.gov/ct2/show/NCT05789901. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/54668.

Upper-Limb Rehabilitation of Patients with Neuromotor Deficits Using Impedance-Based Control of a 6-DOF Robot.

This paper presents a study on the use of impedance-based control of a 6-degree-of-freedom robot for upper-limb rehabilitation of patients with neuromotor deficits. The control strategy is based on impedance and does not require external force sensors at the end-effector for implementation. The experimental setup involved using the control algorithm to move the robot to a desired position, follow a desired trajectory while being moved out of the trajectory by the user, and reproduce three different rehabilitation exercises (passive, isometric, and ADL). The results suggest that the parameters of the control strategy can be adjusted to set the robot's compliance and support force according to the patient's needs. Ultimately, the study concluded that the proposed control strategy can serve as a foundation for rehabilitation robots, which could potentially improve access to continued upper-limb rehabilitation for stroke patients and reduce the number of healthcare professionals required per patient.Clinical Relevance- This paper proposes a human-robot interaction control scheme to be used in a robotic rehabilitation tool, with the goal of assisting both clinicians and patients during the rehabilitation exercises and ultimately reducing the load on healthcare system. This platform can be potentially used in telemedicine solutions for upper-limb rehabilitation of a wide range of patients with neuromotor deficits such as stroke.

A preliminary study in classification of the severity of spine deformation in adolescents with lumbar/thoracolumbar idiopathic scoliosis using machine learning algorithms based on lumbosacral joint efforts during gait.

To assess the severity and progression of adolescents with idiopathic scoliosis (AIS), radiography with X-rays is usually used. The methods based on statistical observations have been developed from 3D reconstruction of the trunk or topography. Machine learning has shown great potential to classify the severity of scoliosis on imaging data, generally on X-ray measurements. It is also known that AIS leads to the development of gait disorder. To our knowledge, machine learning has never been tested on spine intervertebral efforts during gait as a radiation-free method to classify the severity of spinal deformity in AIS. Develop automated machine learning algorithms in lumbar/thoracolumbar scoliosis to classify the severity of spinal deformity of AIS based on the lumbosacral joint (L5-S1) efforts during gait. The lumbosacral joint efforts of 30 individuals with lumbar/thoracolumbar AIS were used as distinctive features fed to the machine learning algorithms. Several tests were run using various classification algorithms. The labeling consisted of three classes reflecting the severity of scoliosis i.e. mild, moderate and severe. The ensemble classifier algorithm including k-nearest neighbors, support vector machine, random forest and multilayer perceptron achieved the most promising results, with accuracy scores of 91.4%. This preliminary study shows lumbosacral joint efforts can be used to classify the severity of spinal deformity in lumbar/thoracolumbar AIS. This method showed the potential of being used as an assessment tool to follow-up the progression of AIS as a radiation-free method, alternative to radiography. Future studies should be performed to test the method on other categories of AIS.

Implementation of an electronic patient-reported measure of barriers to antiretroviral therapy adherence with the Opal patient portal: Protocol for a mixed method type 3 hybrid pilot study at a large Montreal HIV clinic.

BACKGROUND: Adherence to antiretroviral therapy (ART) remains problematic. Regular monitoring of its barriers is clinically recommended, however, patient-provider communication around adherence is often inadequate. Our team thus decided to develop a new electronically administered patient-reported outcome measure (PROM) of barriers to ART adherence (the I-Score) to systematically capture this data for physician consideration in routine HIV care. To prepare for a controlled definitive trial to test the I-Score intervention, a pilot study was designed. Its primary objectives are to evaluate patient and physician perceptions of the I-Score intervention and its implementation strategy. METHODS: This one-arm, 6-month study will adopt a mixed method type 3 implementation-effectiveness hybrid design and be conducted at the Chronic Viral Illness Service of the McGill University Health Centre (Montreal, Canada). Four HIV physicians and 32 of their HIV patients with known or suspected adherence problems will participate. The intervention will involve having patients complete the I-Score through a smartphone application (Opal), before meeting with their physician. Both patients and physicians will have access to the I-Score results, for consideration during the clinic visits at Times 1, 2 (3 months), and 3 (6 months). The implementation strategy will focus on stakeholder involvement, education, and training; promoting the intervention's adaptability; and hiring an Application Manager to facilitate implementation. Implementation, patient, and service outcomes will be collected (Times 1-2-3). The primary outcome is the intervention's acceptability to patients and physicians. Qualitative data obtained, in part, through physician focus groups (Times 2-3) and patient interviews (Times 2-3) will help evaluate the implementation strategy and inform any methodological adaptations. DISCUSSION: This study will help plan a definitive trial to test the efficacy of the I-Score intervention. It will generate needed data on electronic PROM interventions in routine HIV care that will help improve understanding of conditions for their successful implementation. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov identifier: NCT04702412; https://clinicaltrials.gov/.

Current Trends and Challenges in Pediatric Access to Sensorless and Sensor-Based Upper Limb Exoskeletons.

Sensorless and sensor-based upper limb exoskeletons that enhance or support daily motor function are limited for children. This review presents the different needs in pediatrics and the latest trends when developing an upper limb exoskeleton and discusses future prospects to improve accessibility. First, the principal diagnoses in pediatrics and their respective challenge are presented. A total of 14 upper limb exoskeletons aimed for pediatric use were identified in the literature. The exoskeletons were then classified as sensorless or sensor-based, and categorized with respect to the application domain, the motorization solution, the targeted population(s), and the supported movement(s). The relative absence of upper limb exoskeleton in pediatrics is mainly due to the additional complexity required in order to adapt to children's growth and answer their specific needs and usage. This review highlights that research should focus on sensor-based exoskeletons, which would benefit the majority of children by allowing easier adjustment to the children's needs. Sensor-based exoskeletons are often the best solution for children to improve their participation in activities of daily living and limit cognitive, social, and motor impairments during their development.

Second-order simultaneous components model for the overshoot and "slow component" in V̇O2 kinetics.

The human oxygen uptake responses to exercise step on-transients present different shapes depending on the overshoot and/or the "slow component" manifestations. The conventional First-Order Multi-Exponential (FOME) model incorporates delayed add-on terms to comprise these phenomena, increasing parameter quantity, requiring a delayed recruitment of type II fibers to explain the "slow component," and not offering a unified structure for different individuals and intensity domains. We hypothesized that a model composed of two Second-Order Simultaneous Components (SOSC) would present a better overall fitting performance than the FOME. Fourteen well-trained male cyclists performed repeated step on-transitions to moderate, heavy, and severe cycling intensities, whose responses were fitted with FOME and SOSC models. The SOSC presented significantly smaller (p < 0.05) root mean squared errors for moderate, supra-moderate, and all intensities combined. Along with conceptual analyses, these findings suggest the SOSC as a comprehensive alternative to the FOME model, explaining all oxygen uptake step responses with as many parameters and without delayed add-on components.

Proof of Concept of an Assistive Robotic Arm Control Using Artificial Stereovision and Eye-Tracking.

Assistive robotic arms have become popular to help users with upper limb disabilities achieve autonomy in their daily tasks, such as drinking and grasping objects in general. Usually, these robotic arms are controlled with an adapted joystick. Joysticks are user-friendly when it comes to a general approach to an object. However, they are not as intuitive when having to accurately approach an object, especially when obstacles are present. Alternatively, the combined use of artificial stereovision and eye-tracking seems to be a promising solution, as the user's vision is usually dissociated from their upper limb disability. Hence, the objective of this study was to develop a proof of concept for the control of an assistive robotic arm using a low-cost combination of stereovision and eye-tracking. Using the developed control system, a typically developed person was able to control the robotic arm successfully reaching and grasping an object for 92% of the trials without obstacles with an average time of 13.8 seconds. Then, another set of trials with one obstacle had a success rate of 91% with an average time of 17.3 seconds. Finally, the last set of trials with two obstacles had a success rate of 98% with an average time of 18.4 seconds. Furthermore, the cost of an eye-tracker and stereovision remains below 400$.

Development of a multibody model to assess efforts along the spine for the rehabilitation of adolescents with idiopathic scoliosis.

INTRODUCTION: Gait analysis has often been recognized as helpful for the therapeutic follow-up of adolescents with idiopathic scoliosis (IS). METHODS: A multibody model of the human body was developed to display the intervertebral efforts along the spine of each adolescent with IS, and highlight the efforts that significantly differ from typical age-matched healthy adolescents. The intervertebral efforts of one adolescent with IS and an age-matched adolescent during a complete gait cycle were computed and compared. RESULTS: All intervertebral efforts are larger in the adolescent with IS compared to the healthy adolescent, except for the vertical torque. The average medio-lateral torque and force for the participant with IS are respectively 200% and 114% higher. CONCLUSION: This study revealed that the pathological efforts are not concentrated around critical points but distributed along the spine. Thus, higher average efforts along the spine in adolescent with IS may influence the spine deformity due to mechanical modulations according to the Hueter-Volkmann Law. The potential of this model is promising for the therapeutic follow-up of adolescents with IS because it provides real-time efforts along the spine, as well as the corresponding information about the asymmetrical behavior of the spine during gait.

Second-order modeling for the pulmonary oxygen uptake on-kinetics: a comprehensive solution for overshooting and nonovershooting responses to exercise.

The human oxygen uptake (V̇o2) response to step-like increases in work rate is currently modeled by a First Order System Multi-Exponential (FOME) arrangement. Because of their first-order nature, none of FOME model's exponentials is able to model an overshoot in the oxygen uptake kinetics (OV̇o2K). Nevertheless, OV̇o2K phenomena are observed in the fundamental component of trained individuals' step responses. We hypothesized that a Mixed Multi-Exponential (MiME) model, where the fundamental component is modeled with a second- instead of a first-order system, would present a better overall performance than that of the traditional FOME model in fitting V̇o2 on-kinetics at all work rates, either presenting or not OV̇o2K. Fourteen well-trained male cyclists performed three step on-transitions at each of three work rates below their individual lactate thresholds' work rate (WRLT), and two step on-transitions at each of two exercise intensities above WRLT. Averaged responses for each work rate were fitted with MiME and FOME models. Root mean standard errors were used for comparisons between fitting performances. Additionally, a methodology for detecting and quantifying OV̇o2K phenomena is proposed. Second order solutions performed better (P < 0.000) than the first-order exponential when the OV̇o2K was present, and did not differ statistically (P = 0.973) in its absence. OV̇o2K occurrences were observed below and, for the first time, above WRLT (88 and 7%, respectively). We concluded that the MiME model is more adequate and comprehensive than the FOME model in explaining V̇o2 step on-transient responses, considering cases with or without OV̇o2K altogether.NEW & NOTEWORTHY To our knowledge, this is the first study applying second-order system equations to model V̇o2 on-kinetics, which is useful for both mathematical representation and physiological understanding of the overshoot phenomenon manifesting in the fundamental components of some step responses. Moreover, an objective methodology for detecting and quantifying this overshoot that considers data from the whole response is proposed. Finally, this is the first work detecting overshoot occurrences outside the moderate domain of exercise.

Socioeconomic disparities and difficulties to access to healthcare services among Canadian children with neurodevelopmental disorders and disabilities.

OBJECTIVES: The aims of this study were to identify the associations of levels of severity of neurodevelopmental disorders and disabilities (NDD/D) in children with their household socioeconomic status (SES) and their frequency of visits to a healthcare provider, and to examine how the severity of disability varied with these determinants among NDD/D subgroups, in order to inform possible social policy changes and to improve access to the healthcare system. METHODS: Data from the 2006 Participation and Activity Limitation Survey on children aged 5-14 years, collected by Statistics Canada, were analyzed (n=7,072 and weighted n=340,340). Children with NDD/D constituted those with impairments in motor, speech, neurosensory, and psychological functioning, as well as those who had issues with learning/cognition and social interactions. The weighted sample size for this group was n=111,630 (total sample size for children with limitations: n=174,810). We used logistic regression to assess the associations of household SES and frequency of visits to a healthcare provider with disability level. We included NDD/D subgroups as interaction terms in the model. Multiple correspondence analysis (MCA) was conducted to develop a profile of disability level. RESULTS: After-tax low income, family assistance, out-of-pocket expenses, needing but not receiving health services from a social worker, condition of the dwelling, and residential location were associated with the severity of NDD/D. Using MCA, 2 disability profiles could be identified based on access to healthcare, household income status, and condition of the dwelling. CONCLUSIONS: More social interventions are needed to reduce difficulties in accessing healthcare and to diminish the socially determined health inequalities faced by children with NDD/D.

Classification of upper limb disability levels of children with spastic unilateral cerebral palsy using K-means algorithm.

Treatment for cerebral palsy depends upon the severity of the child's condition and requires knowledge about upper limb disability. The aim of this study was to develop a systematic quantitative classification method of the upper limb disability levels for children with spastic unilateral cerebral palsy based on upper limb movements and muscle activation. Thirteen children with spastic unilateral cerebral palsy and six typically developing children participated in this study. Patients were matched on age and manual ability classification system levels I to III. Twenty-three kinematic and electromyographic variables were collected from two tasks. Discriminative analysis and K-means clustering algorithm were applied using 23 kinematic and EMG variables of each participant. Among the 23 kinematic and electromyographic variables, only two variables containing the most relevant information for the prediction of the four levels of severity of spastic unilateral cerebral palsy, which are fixed by manual ability classification system, were identified by discriminant analysis: (1) the Falconer index (CAI E ) which represents the ratio of biceps to triceps brachii activity during extension and (2) the maximal angle extension (θ Extension,max). A good correlation (Kendall Rank correlation coefficient = -0.53, p = 0.01) was found between levels fixed by manual ability classification system and the obtained classes. These findings suggest that the cost and effort needed to assess and characterize the disability level of a child can be further reduced.

Intra- and Intersession Reliability of Surface Electromyography on Muscles Actuating the Forearm During Maximum Voluntary Contractions.

The aim of this study is to determine the intra- and intersession reliability of nonnormalized surface electromyography (sEMG) on the muscles actuating the forearm during maximum voluntary isometric contractions (MVIC). A subobjective of this study is to determine the intra- and intersession reliability of forearm MVIC force or torque, which is a prerequisite to assess sEMG reliability. Eighteen healthy adults participated at 4 different times: baseline, 1-h post, 6-h post, and 24-h post. They performed 3 MVIC trials of forearm flexion, extension, pronation, and supination. sEMG of the biceps brachii short head, brachialis, brachioradialis, triceps brachii long head, pronator teres, and pronator quadratus were measured. The intraclass correlation coefficient (ICC) on MVIC ranged from 0.36 to 0.99. Reliability was excellent for flexion, extension, and supination MVIC for both intra- and intersession. The ICC on sEMG ranged from 0.58 to 0.99. sEMG reliability was excellent for brachialis, brachioradialis, and pronator quadratus, and good to excellent for triceps brachii, biceps brachii, and pronator teres. This study shows that performing 3 MVICs is sufficient to obtain highly reliable maximal sEMG over 24 h for the main muscles actuating the forearm. These results confirm the potential of sEMG for muscle motor functional monitoring.