Can an app help identify psychomotor function impairments during drinking occasions in the real world? A mixed-method pilot study

Leveraging Mobile Phone Sensors, Machine Learning and Explainable Artificial Intelligence to Predict Imminent Same-Day Binge Drinking Events to Support Just-In-Time Adaptive Interventions: A Feasibility Study

BACKGROUND

Digital Just-In-Time Adaptive Interventions (JITAIs) can reduce binge drinking events (BDEs: consuming 4+/5+ drinks per occasion for women/men) in young adults, but need to be optimized for timing and content. Delivering just-in-time support messages in the hours prior to BDEs could improve intervention impact.

OBJECTIVE

We determined the feasibility of developing a machine learning model to accurately predict future, that is, same-day, 1 to 6-hours prior BDEs using smartphone sensor data. We aimed to identify the most informative phone sensor features associated with BDEs on weekend and weekdays, respectively, to determine the key features that explain prediction model performance.

METHODS

We collected phone sensor data from 75 young adults (ages 21-25; mean =22.4, SD=1.9) with risky drinking behavior who reported drinking behavior over 14 weeks. Participants in this secondary analysis were enrolled in a clinical trial. We developed machine learning models testing different algorithms (e.g., XGBoost, decision tree) to predict same-day BDEs (versus low-risk drinking events and non-drinking periods) using smartphone sensor data (e.g., accelerometer, GPS). We tested various "prediction distance" time windows (more proximal: 1-hour; to distant: 6-hour) from drinking onset. We also tested various analysis time windows (i.e., amount of data to be analyzed), ranging from 1 to 12 hours prior to drinking onset, because this determines the amount of data that needs to be stored on the phone to compute the model. Explainable AI (XAI) was used to explore interactions between the most informative phone sensor features contributing to BDEs.

RESULTS

The XGBoost model performed best in predicting imminent same-day BDE, with 95.0% accuracy on weekends and 94.3% accuracy on weekdays (F1 score = 0.95 and 0.94, respectively). This XGBoost model needed 12- and 9-hours of phone sensor data at 3- and 6- hours prediction distance from the onset of drinking, on weekends and weekdays, respectively, prior to predicting same-day BDEs. The most informative phone sensor features for BDE prediction were time (e.g., time of day) and GPS-derived, such as radius of gyration (an indicator of travel). Interactions among key features (e.g., time of day, GPS-derived features) contributed to prediction of same-day BDE.

CONCLUSIONS

We demonstrated the feasibility and potential use of smartphone sensor data and machine learning to accurately predict imminent (same-day) BDEs in young adults. The prediction model provides "windows of opportunity" and with the adoption of XAI, we identified "key contributing features" to trigger JITAI prior to the onset of BDEs, with the potential to reduce the likelihood of BDEs in young adults.

CLINICALTRIAL

Digital assessment of working memory and processing speed in everyday life: Feasibility, validation, and lessons-learned

OBJECTIVES

Cognitive functioning is often impaired in mental and neurological conditions and might fluctuate throughout the day. An existing experience-sampling tool was upgraded to assess individual's cognition in everyday life. The objectives were to test the feasibility and validity of two momentary cognition tasks.

METHODS

The momentary Visuospatial Working Memory Task (mVSWMT) and momentary Digit Symbol Substitution Task (mDSST) were add-ons to an experience sampling method (ESM) smartphone app. Healthy adults (n = 49) between 19 and 73 years of age performed the tasks within an ESM questionnaire 8 times a day, over 6 consecutive days. Feasibility was determined through completion rate and participant experience. Validity was assessed through contextualization of cognitive performance within intrapersonal and situational factors in everyday life.

FINDINGS

Participants experienced the tasks as pleasant, felt motivated, and the completion rate was high (71%). Social context, age, and distraction influenced cognitive performance in everyday life. The mVSWMT was too difficult as only 37% of recalls were correct and thus requires adjustments (i.e. fixed time between encoding and recall; more trials per moment). The mDSST speed outcome seems the most sensitive outcome measure to capture between- and within-person variance.

CONCLUSIONS

Short momentary cognition tasks for repeated assessment are feasible and hold promise, but more research is needed to improve validity and applicability in different samples. Recommendations for teams engaging in the field include matching task design with traditional neuropsychological tests and involving a multidisciplinary team as well as users. Special attention for individual needs can improve motivation and prevent frustration. Finally, tests should be attractive and competitive to stimulate engagement, but still reflect actual cognitive functioning.

Measuring within-day cognitive performance using the experience sampling method: A pilot study in a healthy population

Introduction

People with depression, anxiety, or psychosis often complain of confusion, problems concentrating or difficulties cognitively appraising contextual cues. The same applies to people with neurodegenerative diseases or brain damage such as dementia or stroke. Assessments of those cognitive difficulties often occurs in cross-sectional and controlled clinical settings. Information on daily moment-to-moment cognitive fluctuations and its relation to affect and context is lacking. The development and evaluation of a digital cognition task is presented. It enables the fine-grained mapping of cognition and its relation to mood, intrapersonal factors and context.

Methods

The momentary Digit Symbol Substitution Task is a modified digital version of the original paper-and-pencil task, with a duration of 30 seconds and implemented in an experience sampling protocol (8 semi-random assessments a day on 6 consecutive days). It was tested in the healthy population (N = 40). Descriptive statistics and multilevel regression analyses were used to determine initial feasibility and assess cognitive patterns in everyday life. Cognition outcome measures were the number of trials within the 30-second sessions and the percentage of correct trials.

Results

Subjects reported the task to be easy, pleasant and do-able. On average, participants completed 11 trials with 97% accuracy per 30-second session. Cognitive variation was related to mood, with an interaction between positive and negative affect for accuracy (% correct) (p = .001) and an association between positive affect and speed (number of trials) (p = .01). Specifically, cheerful, irritated and anxious seem to covary with cognition. Distraction and location are relevant contextual factors. The number of trials showed a learning effect (p < .001) and was sensitive to age (p < .001).

Conclusion

Implementing a digital cognition task within an experience-sampling paradigm shows promise. Fine-tuning in further research and in clinical samples is needed. Gaining insight into cognitive functioning could help patients navigate and adjust the demands of daily life.

Training Cognitive Functions Using Mobile Apps in Breast Cancer Patients: Systematic Review

Background

Breast cancer is an invalidating disease and its treatment can bring serious side effects that have a physical and psychological impact. Specifically, cancer treatment generally has a strong impact on cognitive function. In recent years, new technologies and eHealth have had a growing influence on health care and innovative mobile apps can be useful tools to deliver cognitive exercise in the patient’s home.

Objective

This systematic review gives an overview of the state-of-the-art mobile apps aimed at training cognitive functions to better understand whether these apps could be useful tools to counteract cognitive impairment in breast cancer patients.

Methods

We searched in a systematic way all the full-text articles from the PubMed and Embase databases.

Results

We found eleven studies using mobile apps to deliver cognitive training. They included a total of 819 participants. App and study characteristics are presented and discussed, including cognitive domains trained (attention, problem solving, memory, cognitive control, executive function, visuospatial function, and language). None of the apps were specifically developed for breast cancer patients. They were generally developed for a specific clinical population. Only 2 apps deal with more than 1 cognitive domain, and only 3 studies focus on the efficacy of the app training intervention.

Conclusions

These results highlight the lack of empirical evidence on the efficacy of currently available apps to train cognitive function. Cognitive domains are not well defined across studies. It is noteworthy that no apps are specifically developed for cancer patients, and their applicability to breast cancer should not be taken for granted. Future studies should test the feasibility, usability, and effectiveness of available cognitive training apps in women with breast cancer. Due to the complexity and multidimensionality of cognitive difficulties in this cancer population, it may be useful to design, develop, and implement an ad hoc app targeting cognitive impairment in breast cancer patients.

Mobile phone sensors and supervised machine learning to identify alcohol use events in young adults: Implications for just-in-time adaptive interventions

BACKGROUND

Real-time detection of drinking could improve timely delivery of interventions aimed at reducing alcohol consumption and alcohol-related injury, but existing detection methods are burdensome or impractical.

OBJECTIVE

To evaluate whether phone sensor data and machine learning models are useful to detect alcohol use events, and to discuss implications of these results for just-in-time mobile interventions.

METHODS

38 non-treatment seeking young adult heavy drinkers downloaded AWARE app (which continuously collected mobile phone sensor data), and reported alcohol consumption (number of drinks, start/end time of prior day's drinking) for 28days. We tested various machine learning models using the 20 most informative sensor features to classify time periods as non-drinking, low-risk (1 to 3/4 drinks per occasion for women/men), and high-risk drinking (>4/5 drinks per occasion for women/men).

RESULTS

Among 30 participants in the analyses, 207 non-drinking, 41 low-risk, and 45 high-risk drinking episodes were reported. A Random Forest model using 30-min windows with 1day of historical data performed best for detecting high-risk drinking, correctly classifying high-risk drinking windows 90.9% of the time. The most informative sensor features were related to time (i.e., day of week, time of day), movement (e.g., change in activities), device usage (e.g., screen duration), and communication (e.g., call duration, typing speed).

CONCLUSIONS

Preliminary evidence suggests that sensor data captured from mobile phones of young adults is useful in building accurate models to detect periods of high-risk drinking. Interventions using mobile phone sensor features could trigger delivery of a range of interventions to potentially improve effectiveness.

An Artificial Neural Network for Movement Pattern Analysis to Estimate Blood Alcohol Content Level

Impairments in gait occur after alcohol consumption, and, if detected in real-time, could guide the delivery of "just-in-time" injury prevention interventions. We aimed to identify the salient features of gait that could be used for estimating blood alcohol content (BAC) level in a typical drinking environment. We recruited 10 young adults with a history of heavy drinking to test our research app. During four consecutive Fridays and Saturdays, every hour from 8 p.m. to 12 a.m., they were prompted to use the app to report alcohol consumption and complete a 5-step straight-line walking task, during which 3-axis acceleration and angular velocity data was sampled at a frequency of 100 Hz. BAC for each subject was calculated. From sensor signals, 24 features were calculated using a sliding window technique, including energy, mean, and standard deviation. Using an artificial neural network (ANN), we performed regression analysis to define a model determining association between gait features and BACs. Part (70%) of the data was then used as a training dataset, and the results tested and validated using the rest of the samples. We evaluated different training algorithms for the neural network and the result showed that a Bayesian regularization neural network (BRNN) was the most efficient and accurate. Analyses support the use of the tandem gait task paired with our approach to reliably estimate BAC based on gait features. Results from this work could be useful in designing effective prevention interventions to reduce risky behaviors during periods of alcohol consumption.