Project
| # | Title | Team Members | TA | Documents | Sponsor |
|---|---|---|---|---|---|
| 1 | Sound Asleep |
Adam Tsouchlos Ambika Mohapatra Shub Pereira |
Weiman Yan | design_document1.pdf final_paper1.pdf photo1.png photo2.jpg presentation2.pdf presentation1.pdf proposal1.pdf |
|
| # **Sound Asleep** **Team Members:** - Adam Tsouchlos (adamtt2) - Ambika Mohapatra (ambikam2) - Shub Pereira (shubbp2) # **Problem** Poor sleep can have serious effects on your health, increasing chances of conditions like poor mental health, kidney failure, diabetes, and more. It was found that slow wave sleep declines with age and that it is considered the most restorative stage of sleep. It is important for improving immune function, memory consolidation, and emotional regulation. Recent literature discusses using auditory stimulation during sleep to increase longevity of slow wave sleep for better overall physical and mental health. There are other devices that use EEG technology, but most have no auditory stimulation and the others were said to be very uncomfortable. # Solution **Sound Asleep**: a non-invasive wearable that transmits EEG data to a companion app. This then interacts with the user’s Bluetooth device to deliver precisely timed auditory stimulation. The user can choose their own bluetooth device for increased comfort during sleep. # Solution Components # Subsystem 1 – EEG Acquisition and Wearable Hardware This subsystem is responsible for acquiring the EEG signals. - EEG leads optimized for overnight use. - Wearable headband or soft cap to keep electrodes in place throughout the night. - Low-noise amplification and filtering circuitry to ensure signals are usable for real-time processing. - Small rechargeable battery to power sensors and wireless transmission. # Subsystem 2 – Wireless Transmission and Power This subsystem ensures EEG data can be reliably sent to the processing unit. - Bluetooth Low Energy (BLE) or Wi-Fi module for continuous data transfer. - Onboard microcontroller to digitize EEG signals and handle wireless protocols. - Battery management system for safe charging and overnight operation. # Subsystem 3 – Sleep Stage Classification and Signal Processing This subsystem processes EEG data in real-time to detect sleep stages and identify slow wave activity. - Algorithms for sleep staging (NREM, REM, wake) using EEG features. - Slow wave detection algorithms trained/tested on pre-labeled EEG datasets. - Closed-loop timing logic to sync auditory stimulation with ongoing slow waves. - Possible algorithms to be used: **YASA Slow-waves detection.** https://github.com/raphaelvallat/yasa/blob/master/notebooks/05_sw_detection.ipynb **CoSleep GitHub project.** https://github.com/Frederik-D-Weber/cosleep # Subsystem 4 – Auditory Stimulation Delivery (and App User Interface) This subsystem delivers pink noise bursts at intervals during SWS. - Mobile (or desktop) app triggers sound output through the user’s paired Bluetooth device (primary option as of now). - Sound customization features via app for intensity, duration, frequency, and comfort. - Sleep session dashboard showing nightly summaries (total sleep, time in slow wave sleep, stimulation events delivered). # Criterion for Success # ****Hardware**** - Wearing the EEG device is considered comfortable by users. - EEG device stays attached during full night of sleep - EEG readings are accurately transmitted to the software. # Software - EEG readings are correctly detected and processed by the app. - Slow wave sleep stage is accurately identified. - Auditory stimulation is transmitted to user’s bluetooth device. # Outcomes - User has increased slow wave sleep duration and amplitude. - Improvement in memory test after sleeping with the device compared to without it. # References - Ngo et al. (2013). Auditory closed-loop stimulation of the sleep slow oscillation enhances memory. https://pubmed.ncbi.nlm.nih.gov/23583623/ - Bo-Lin Su et al. (2015). Detecting slow wave sleep using a single EEG signal channel. https://pubmed.ncbi.nlm.nih.gov/25637866/ |
|||||