Kimmo Kärkkäinen

Background: Sexual and gender minority (SGM) individuals are at heightened risk for substance use and sexually transmitted infections than their non-SGM peers. Collecting mobile phone usage data passively may open new opportunities for personalizing interventions, as behavioral risks could be identified without user input. Objective: This study aimed to determine (1) whether passively sensed mobile phone data can be used to identify substance use and sexual risk behaviors for sexually transmitted infection (STI) and HIV transmission among young SGM who have sex with men, (2) which outcomes can be predicted with a high level of accuracy, and (3) which passive data sources are most predictive of these outcomes. Methods: We developed a mobile phone app to collect participants’ messaging, location, and app use data and trained a machine learning model to predict risk behaviors for STI and HIV transmission. We used Scikit-learn to train logistic regression and gradient boosting classification models with simple linear model specification to predict participants' substance use and sexual behaviors (ie, condomless anal sex, number of sexual partners, and methamphetamine use), which were validated using self-report questionnaires. F₁-scores were used to quantify prediction accuracy of the model using different data sources (and combinations of these sources) for prediction. Differences between text, location, app use, and Linguistic Inquiry and Word Count (LIWC) domains by outcome were investigated using independent t tests where associations were considered significant at P<.05. Results: Among participants (n=82) who identified as SGM, were sexually active, and reported recent substance use, our model was highly predictive of methamphetamine use and having ≥6 sexual partners (F₁-scores as high as 0.83 and 0.69, respectively). The model was less predictive of condomless anal sex (highest F₁-score 0.38). Overall, text-based features were found to be most predictive, but app use and location data improved predictive accuracy, particularly for detecting ≥6 sexual partners. Methamphetamine use was significantly associated with dating app use (P=.01) and use of sex-related words (P=.002). Having ≥6 sex partners was associated with dating app use (0.02), use of sex-related words (P=.001), and traveling a further distance from home (P=.03), on average, compared to participants with fewer sex partners. Methamphetamine users were more likely to use social (P=.002) and affect words (P=.003) and less likely to use drive-related words (P=.02). People having 6 or more partners were more likely to use social, affect words, and cognitive process-related words (P=.003 and .004 respectively). Conclusions: Our results show that passively collected mobile phone data may be useful in detecting sexual risk behaviors. Expanding data collection may improve the results further, as certain behaviors, such as injection drug use, were quite rare in the study sample. These models may be used to personalize STI and HIV prevention as well as substance use harm reduction interventions.

Background: Sexual and gender minority (SGM) young people are disproportionately affected by HIV in the United States, and substance use is a major driver of new infections. People who use web-based venues to meet sex partners are more likely to report substance use, sexual risk behaviors, and sexually transmitted infections. To our knowledge, no machine learning (ML) interventions have been developed that use web-based and digital technologies to inform and personalize HIV and substance use prevention efforts for SGM young people. Objective: This study aims to test the acceptability, appropriateness, and feasibility of the uTECH intervention, a SMS text messaging intervention using an ML algorithm to promote HIV prevention and substance use harm reduction among SGM people aged 18 to 29 years who have sex with men. This intervention will be compared to the Young Men’s Health Project (YMHP) alone, an existing Centers for Disease Control and Prevention best evidence intervention for young SGM people, which consists of 4 motivational interviewing–based counseling sessions. The YMHP condition will receive YMHP sessions and will be compared to the uTECH+YMHP condition, which includes YMHP sessions as well as uTECH SMS text messages. Methods: In a study funded by the National Institutes of Health, we will recruit and enroll SGM participants (aged 18-29 years) in the United States (N=330) to participate in a 12-month, 2-arm randomized comparison trial. All participants will receive 4 counseling sessions conducted over Zoom (Zoom Video Communications, Inc) with a master’s-level social worker. Participants in the uTECH+YMHP condition will receive curated SMS text messages informed by an ML algorithm that seek to promote HIV and substance use risk reduction strategies as well as undergoing YMHP counseling. We hypothesize that the uTECH+YMHP intervention will be considered acceptable, appropriate, and feasible to most participants. We also hypothesize that participants in the combined condition will experience enhanced and more durable reductions in substance use and sexual risk behaviors compared to participants receiving YMHP alone. Appropriate statistical methods, models, and procedures will be selected to evaluate primary hypotheses and behavioral health outcomes in both intervention conditions using an α<.05 significance level, including comparison tests, tests of fixed effects, and growth curve modeling. Results: This study was funded in August 2019. As of June 2024, all participants have been enrolled. Data analysis has commenced, and expected results will be published in the fall of 2025. Conclusions: This study aims to develop and test the acceptability, appropriateness, and feasibility of uTECH, a novel approach to reduce HIV risk and substance use among SGM young adults. Trial Registration: ClinicalTrials.gov NCT04710901; https://clinicaltrials.gov/study/NCT04710901 International Registered Report Identifier (IRRID): DERR1-10.2196/58448

Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: Sexual and gender minority (SGM) young people are disproportionately affected by HIV in the United States, and substance use is a major driver of new infections. People who use online venues to meet sexual partners are more likely to report substance use, sexual risk behaviors, and sexually transmitted infections (STIs). To our knowledge, no machine learning interventions have been developed that utilize online and digital technologies to inform and personalize HIV and substance use prevention efforts with sexual and gender minorities. Objective: This study is a two-arm randomized comparison trial that tests the acceptability, appropriateness, and feasibility of the uTECH intervention, a text message intervention using a machine learning algorithm to promote substance use harm reduction and safer behaviors among 18-29 year old sexual and gender minority people who have sex with men. This intervention will be compared to YMHP (Young Men’s Health Program) alone, an existing CDC Best Evidence intervention for young SGM men, that consists of 4 motivational interviewing-based counseling sessions. The YMHP condition will receive YMHP sessions and will be compared to the uTECH+YMHP condition, which includes YMHP sessions as well as uTECH text messages. Methods: In a study funded by the National Institutes of Health, we will recruit and enroll sexual and gender minority participants (aged 18 to 29) in the United States (N=330) to participate in a 12-month study. All participants will receive 4 counseling sessions conducted over Zoom with a Masters-level social worker. Participants in the uTECH+YMHP condition will receive curated text messages informed by a machine learning algorithm that seeks to promote HIV and substance use risk reduction strategies as well as undergoing YMHP counseling. Over the course of the study, participants will receive between $350 to $450 in incentives, depending on condition assignment. We hypothesize the uTECH+YMHP intervention will be considered acceptable, appropriate, and feasible to most participants. We also hypothesize that participants in the combined condition will experience enhanced and more durable reductions in substance use and sexual risk behaviors compared to participants receiving YMHP alone. Results: This study will test the acceptability, appropriateness, and feasibility of uTECH for sexual and gender minority young adults in the United States (N=330). uTECH seeks to reduce HIV transmission by sending personalized and curated messages promoting harm reduction and safer sex behaviors based on a machine learning algorithm. Conclusions: This study aims develop and test the acceptability, appropriateness, and feasibility of uTECH, a novel approach to reduce HIV risk and substance use among sexual and gender minority young adults. Clinical Trial: Clinical Trials ID: NCT04710901

Purpose: We sought to understand technology-based communication regarding mpox (monkeypox) among gay, bisexual, and other men who have sex with men (GBMSM) during the global outbreak in 2022. Methods: Forty-four GBMSM (M _age = 25.3 years, 68.2% cisgender, 43.2% non-White) living in the United States participated. From May 2022 to August 2022, all text data related to mpox (174 instances) were downloaded from the smartphones of GBMSM. Text data and smartphone app usage were analyzed. Results: Content analysis revealed 10 text-based themes and 7 app categories. GBMSM primarily used search and browser, texting, and gay dating apps to share vaccine updates, seek mpox vaccination, find general mpox information, share mpox information with other GBMSM, and discuss links between mpox and gay culture. Data visualizations revealed that changes in communication themes and app usage were responsive to major milestones in the mpox outbreak. Conclusion: GBMSM used apps to facilitate a community-driven mpox response.

In many machine learning applications, we are faced with incomplete datasets. In the literature, missing data imputation techniques have been mostly concerned with filling missing values. However, the existence of missing values is synonymous with uncertainties not only over the distribution of missing values but also over target class assignments that require careful consideration. In this paper, we propose a simple and effective method for imputing missing features and estimating the distribution of target assignments given incomplete data. In order to make imputations, we train a simple and effective generator network to generate imputations that a discriminator network is tasked to distinguish. Following this, a predictor network is trained using the imputed samples from the generator network to capture the classification uncertainties and make predictions accordingly. The proposed method is evaluated on CIFAR-10 and MNIST image datasets as well as five real-world tabular classification datasets, under different missingness rates and structures. Our experimental results show the effectiveness of the proposed method in generating imputations as well as providing estimates for the class uncertainties in a classification task when faced with missing values.

In many real-world machine learning problems, feature values are not readily available. To make predictions, some of the missing features have to be acquired, which can incur a cost in money, computational time, or human time, depending on the problem domain. This leads us to the problem of choosing which features to use at the prediction time. The chosen features should increase the prediction accuracy for a low cost, but determining which features will do that is challenging. The choice should take into account the previously acquired feature values as well as the feature costs. This paper proposes a novel approach to address this problem. The proposed approach chooses the most useful features adaptively based on how relevant they are for the prediction task as well as what the corresponding feature costs are. Our approach uses a generic neural network architecture, which is suitable for a wide range of problems. We evaluate our approach on three cost-sensitive datasets, including Yahoo! Learning to Rank Competition dataset as well as two health datasets. We show that our approach achieves high accuracy with a lower cost than the current state-of-the-art approaches.

In many real-world scenarios where data is high dimensional, test time acquisition of features is a non-trivial task due to costs associated with feature acquisition and evaluating feature value. The need for highly confident models with an extremely frugal acquisition of features can be addressed by allowing a feature selection method to become target aware. We introduce an approach to feature selection that is based on Bayesian learning, allowing us to report target-specific levels of uncertainty, false positive, and false negative rates. In addition, measuring uncertainty lifts the restriction on feature selection being target agnostic, allowing for feature acquisition based on a single target of focus out of many. We show that acquiring features for a specific target is at least as good as common linear feature selection approaches for small non-sparse datasets, and surpasses these when faced with real-world healthcare data that is larger in scale and in sparseness.