JMIR Publications

ABSTRACT

Background:

Tinnitus is often described as the phantom perception of a sound and it is experienced by 5.1% to 42.7% of the population worldwide at least once during lifetime. The symptoms often reduce the patient’s quality of life. The TrackYourTinnitus (TYT) mHealth Crowdsensing Platform was developed to help patients in demystifying the daily moment-to-moment variations of their tinnitus symptoms.

Objective:

In the current study, we explored whether the mobile operating system (Android; iOS) of the users can be predicted by the dynamic daily life TYT data.

Methods:

TYT mainly applies the paradigms Ecological Momentary Assessment (EMA) and mobile crowdsensing to collect dynamic daily life data. The dynamic EMA (EMA-D) data that was analyzed included 8 questions that were assessed by the users at least at 10 times. In the current study, we could include 352 users, thereby 131 were iOS users with 8130 EMA-D and 221 Android users with 24681 EMA-D. Machine learning methods (i.e., Feedforward Neural Network (FNN), Decision Tree (DT), and a Support Vector Machine (SVM)) were applied to address the research question.

Results:

Machine learning was able to predict the used mobile operating system of a TYT user with an accuracy up to 91.49% based on the provided EMA-D on assessment level. In this context, the daily measurements on the subjective mood level and the perceived arousal of the momentary tinnitus were particularly suitable for the prediction of the used mobile operating system.

Conclusions:

In the work at hand, two particular aspects have been revealed. First, machine learning can contribute to dynamic EMA-data in the medical context. Second, based on the EMA-data of TYT, a high accuracy was found to predict the used mobile operating system. The observed relationship between the collected data and the used operating system has many implications. Among these implications is the need to consider the operating system as a potential confounder for the assessed data.