TY - GEN
T1 - Transforming Polysomnography: Time-Frequency Transforms to Visualise and Classify Polysomnography Data
AU - McCausland, Christopher
AU - Bond, Raymond
AU - Finlay, Dewar
AU - Kennedy, Alan
AU - Biglarbeigi, Pardis
PY - 2023/10/6
Y1 - 2023/10/6
N2 - Data transforms simplify signals while preserving essential information. Polysomnography (PSG) is an example of several complex signals recorded simultaneously, which require classification into sleep stages and arousal events per epoch. This study outlines the initial investigation of a novel transform to visualise PSG data in a more human intuitive manner, considering the American Academy of Sleep Medicines (AASM) scoring guidelines. A machine learning classifier was employed to ensure epochs were accurately classified. Average sensitivity, specificity and Cohen’s kappa were 0.79, 0.95 and 0.62 respectively.
AB - Data transforms simplify signals while preserving essential information. Polysomnography (PSG) is an example of several complex signals recorded simultaneously, which require classification into sleep stages and arousal events per epoch. This study outlines the initial investigation of a novel transform to visualise PSG data in a more human intuitive manner, considering the American Academy of Sleep Medicines (AASM) scoring guidelines. A machine learning classifier was employed to ensure epochs were accurately classified. Average sensitivity, specificity and Cohen’s kappa were 0.79, 0.95 and 0.62 respectively.
UR - https://esleepeurope.eu/wp-content/uploads/2023/09/Transforming-Polysomnography-Time-Frequency-Transforms-to-Visualise-and-Classify-Polysomnography-Data.pdf
M3 - Conference contribution
BT - eSleep2023
ER -
