의료기기

These are the user-facing interface (UI) and functional components.

• •

  Login: Authentication for user identification and protection of personal health data

• •

  Pulse wave measurement: An interface that links the mobile camera for real-time on-device pulse wave signal extraction

• •

  View results: Visualization of the analyzed real-time pulse wave graph

• •

  Health records: Daily/weekly/monthly pulse wave trend graphs and history management

• •

  Settings: General settings such as profile management, notification settings, and app version

• ▸

  On-device pulse wave extraction: After capturing the user's facial image via the smartphone camera, the device analyzes subtle color changes and reflected-light information on the facial surface in real time, directly on the mobile device, to extract the pulse wave signal.

• ▸

  HSI-based image processing (first-stage noise removal): To minimize distortion from facial curvature, slight rotation, or lighting changes, the RGB color space is converted into the HSI (Hue, Saturation, Intensity) space to separate the intensity component and secure a stable skin-region signal.

• ▸

  RLS algorithm application (second-stage noise removal): To effectively cancel motion artifacts caused by facial expression changes or movement, an RLS (Recursive Least Squares adaptive filter) algorithm is applied to correct distortion in the pulse wave signal and maximize the signal-to-noise ratio (SNR).

It uses the smartphone's camera and built-in sensors (accelerometer, gyroscope) to measure the user's biosignals (PPG, seismocardiogram), and analyzes and estimates blood pressure based on the pulse transit time (PTT) between the two signals to monitor health management.

• ▸

  Composite biosignal measurement:

  • •

    PPG (photoplethysmography) signal acquisition: Using the smartphone camera (and flash), peripheral blood flow changes at the fingertip and similar sites are detected to extract the PPG signal.

  • •

    Seismocardiogram (SCG) signal acquisition: By placing the smartphone against the chest, the subtle physical vibrations generated by the heartbeat are measured with the built-in accelerometer and gyroscope.

• ▸

  PTT (Pulse Transit Time) computation: The time delay (PTT) is calculated between the vibration generated when blood is ejected from the heart (a specific peak in the SCG signal) and the moment that blood flow reaches the peripheral vessels (a specific peak in the PPG signal).

• ▸

  Blood pressure estimation algorithm: PTT has a strong inverse correlation with vascular elasticity and blood pressure. The server's algorithm interprets this PTT value to derive trends in systolic and diastolic blood pressure.

• ▸

  Cuff-based blood pressure monitor calibration: To remove errors caused by individual vascular characteristics (vessel stiffness, length, etc.), a reference blood pressure value measured with an approved cuff-type monitor is entered to optimize the algorithm parameters (personalized mapping).

• ▸

  Signal extraction (rPPG): The smartphone camera captures the surface of the subject's facial skin. Subtle skin-color changes (variations in light absorption and reflectance) caused by changing blood volume with each heartbeat are captured through image signal processing to extract the remote photoplethysmography (rPPG) signal in a contactless manner.

• ▸

  AI-based blood pressure analysis: The extracted rPPG signal data is transmitted to the cloud server, where an LSTM-CNN-based artificial intelligence (AI) algorithm combines and analyzes its time-series and spatial features.

• ▸

  Personalized calibration: To maximize measurement accuracy, when the user enters an initial calibration value measured with a cuff-type blood pressure monitor, the AI server maps it to the rPPG analysis results to derive a personalized blood pressure estimate.

Deepcheck is a contactless health management solution that easily measures blood pressure, heart rate, stress, fatigue, and more with a smartphone camera.

It detects abnormal health signs in real time and notifies administrators, and continuously manages changes in health status—helping prevent safety accidents at industrial sites and supporting worker health management.