ABSTRACT
Ecological Momentary Assessments (EMA) deliver insights on how patients perceive tinnitus at different times and how they are affected by it. Moving to the next level, an mHealth app can support users more directly by predicting a user's next EMA and recommending personalized services based on these predictions. In this study, we analyzed the data of 21 users who were exposed to an mHealth app with non-personalized recommendations, and we investigate ways of predicting the next vector of EMA answers. We studied the potential of entity-centric predictors that learn for each user separately and neighborhood-based predictors that learn for each user separately but take also similar users into account, and we compared them to a predictor that learns from all past EMA indiscriminately, without considering which user delivered which data, i.e., to a "global model." Since users were exposed to two versions of the non-personalized recommendations app, we employed a Contextual Multi-Armed Bandit (CMAB), which chooses the best predictor for each user at each time point, taking each user's group into account. Our analysis showed that the combination of predictors into a CMAB achieves good performance throughout, since the global model was chosen at early time points and for users with few data, while the entity-centric, i.e., user-specific, predictors were used whenever the user had delivered enough data-the CMAB chose itself when the data were "enough." This flexible setting delivered insights on how user behavior can be predicted for personalization, as well as insights on the specific mHealth data. Our main findings are that for EMA prediction the entity-centric predictors should be preferred over a user-insensitive global model and that the choice of EMA items should be further investigated because some items are answered more rarely than others. Albeit our CMAB-based prediction workflow is robust to differences in exposition and interaction intensity, experimentators that design studies with mHealth apps should be prepared to quantify and closely monitor differences in the intensity of user-app interaction, since users with many interactions may have a disproportionate influence on global models.
ABSTRACT
OBJECTIVES: Young people are often exposed to high music levels which make them more at risk to develop noise-induced symptoms such as hearing loss, hyperacusis, and tinnitus of which the latter is the symptom perceived the most by young adults. Although, subclinical neural damage was demonstrated in animal experiments, the human correlate remains under debate. Controversy exists on the underlying condition of young adults with normal hearing thresholds and noise-induced tinnitus (NIT) due to leisure noise. The present study aimed to assess differences in audiological characteristics between noise-exposed adolescents with and without NIT. METHODS: A group of 87 young adults with a history of recreational noise exposure was investigated by use of the following tests: otoscopy, impedance measurements, pure-tone audiometry including high-frequencies, transient and distortion product otoacoustic emissions, speech-in-noise testing with continuous and modulated noise (amplitude-modulated by 15 Hz), auditory brainstem responses (ABR) and questionnaires.Nineteen students reported NIT due to recreational noise exposure, and their measures were compared to the non-tinnitus subjects. RESULTS: No significant differences between tinnitus and non-tinnitus subjects could be found for hearing thresholds, otoacoustic emissions, and ABR results.Tinnitus subjects had significantly worse speech reception in noise compared to non-tinnitus subjects for sentences embedded in steady-state noise (mean speech reception threshold (SRT) scores, respectively -5.77 and -6.90 dB SNR; p = 0.025) as well as for sentences embedded in 15 Hz AM-noise (mean SRT scores, respectively -13.04 and -15.17 dB SNR; p = 0.013). In both groups speech reception was significantly improved during AM-15 Hz noise compared to the steady-state noise condition (p < 0.001). However, the modulation masking release was not affected by the presence of NIT. CONCLUSIONS: Young adults with and without NIT did not differ regarding audiometry, OAE, and ABR.However, tinnitus patients showed decreased speech-in-noise reception. The results are discussed in the light of previous findings suggestion NIT may occur in the absence of measurable peripheral damage as reflected in speech-in-noise deficits in tinnitus subjects.
