KENTECH Intelligent Mobile Computiong lab (IMC Lab)

Digital Healthcare

1. Stress Detection in the Wild

Stress is known to stimulate physiological responses such as heart activity changes and sweat, and such events can be detected using wearables such as smartwatches and chest sensors. Moreover, smartphones give access to rich situational contexts and allow stress detection research to move towards uncontrolled environments. Our methods incorporate collection of situational contexts and physiological sensing to accurately predict daily life stress, which is further used for Digital Therapeutics (DTx) such as timely intervention suggestions to prevent health complications by managing stress levels. In the future, alongside the health benefits, this research direction can help individuals optimize their daily stress levels and enhance productivity by “controlling” their stress levels.

2. Short-term Depression Detection

Depression is a major public health issue, often with a chronic course and poor prognosis. Traditional methods of estimating depression are time-consuming and costly, and may require professional involvement. Passive sensing using mobile and wearable devices can track daily activities and routines, providing digital behavior markers related to depression. Our work provides a systematic data processing pipeline that leverages mobile passive data for the clinical detection and forecasting of depression and depressive moods. We analyze a wide range of digital behavior markers related to depression, including physical movement, social interactions, and daily activities. By integrating these digital behavior markers, we perform comprehensive and accurate assessment of depression, allowing for personalized interventions and long-term symptom management.

3. Large Language Model for Mental Health

Traditional in-person counseling encounters limitations in terms of accessibility, flexibility, and social stigma. Additionally, low mental health literacy and embarrassment among individuals hinder help-seeking behavior. Meanwhile, the introduction of more sophisticated sensors embedded in ubiquitous devices such as smartphones and smartwatches and the release of powerful large language models create new opportunities to address the existing limitations of traditional counseling services. In that regard, we plan to develop a system that offers round-the-clock mental health counseling services. By leveraging large language models and continuous analysis of user context and digital phenotype, the system needs to deliver personalized counseling support. Through 24/7 passive monitoring, the system will continuously assess individuals’ mental states, initiate conversations on their behalf, and potentially trigger counseling services. Such specialized counseling services will be facilitated by a fine-tuned large language model such as chat-GPT, LLaMa, PaLM, and Bard.

NRF 차세대바이오(21.07~24.12)


  • IEEE International Conference on Big Data and Smart Computing (IEEE BigComp)
    WildStress: exploring rich situational contexts for stress detection in the wild
    Kobiljon Toshnazarov, Tae Hong Lee, Youngtae Noh
  • IEEE International Conference on Big Data and Smart Computing (IEEE BigComp)
    Data Processing Pipeline of Short-Term Depression Detection with Large-Scale Dataset
    Yonggeon Lee, Youngtae Noh, Uichin Lee
  • Sensors
    STDD: Short-Term Depression Detection with Passive Sensing
    Nematjon Narziev, Hwarang Goh, Kobiljon Toshnazarov, Seung Ah Lee, Kyong-Mee Chung, Youngtae Noh
  • Adjunct Proceedings of the 2023 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2023 ACM International Symposium on Wearable Computers
    Poster: WMGPT: Towards 24/7 Online Prime Counseling with ChatGPT
    Lismer Andres Caceres Najarro, Yonggeon Lee, Kobiljon E. Toshnazarov, Yoonhyung Jang, Hyungsook Kim, Youngtae Noh

Research Participants

  • Research Professor
    Caceres-Najarro Lismer Andres
    Research Interest
    • Evolutionary algorithms
    • Wireless sensor networks
    • Localization and tracking
    • Smart Grid
    • Smart healthcare
  • PhD Student
    Kobiljon Toshnazarov
    Research Interest
    • Stress Detection
    • mHealth Data Collection
    • Digital Therapeutics (DTx)
  • Integrated PhD Student
    Lee Yong geon
    Research Interest
    • Sensor Data Science
    • Energy Behavior Modeling
    • Energy Management System
    • Digital Therapeutics