Can et al. (2019b) developed an automatic stress detection system that employs machine learning tools and concluded that physiological modalities are more accurate than self-reported perceived stress when analyzing real-world data. With the development of wearable devices and sensors, the assessment of physiological responses to the treatments could become routine (Can et al., 2019a), providing real-world data from patients in their own environment instead of measurements made in the artificial clinical laboratory setting.