/Spindle-Detection

Implementation of 'Continuous Wavelet Transformation' with a morlet basis for spindle detection.

Primary LanguageJupyter Notebook

Snore Detection

The current Repository demonstrates how to detect Snoring using Ballistocardiography readings.

Workflow

1. Exploratory Data Analysis:

  • Data analysis for Trends and Patterns from a Descriptive perspective.

2. Statistical Analysis:

  • 5 Point Summary: Stating the Mean, Median, Min, Mix and Standard Deviation of the Sample.
  • Feature Engineering: Created several features for future analysis based on stated Assumptions.
  • Descriptive Analysis:
    • I study the Data Visually, in order to generate Hypothesis.
  • Hypothesis Testing:
    • Based on the Hypothesis generated, I perform several statistical tests to reject/do not reject the Hypothesis.

3. Snore Detection:

  • Spindle Analysis:
    • Spindle detection is performed from an 'Information' perspective.
  • Loss Based Approach:
    • Detection is performed with respect to pre-set threshold.

References:

EEG sleep spindle detection using continuous wavelet transformation
Automatic sleep spindle detection
Heart Rate Measurement
Numerical Python Book