# Conversion of airflow signal into 1-D as a input for CNN [closed]

Trying to detect apnea from airflow signal. I have 100 files of the patient each file contains at least 8 hours of data and sampling rate is 32 Hz. How to prepare my dataset which is used as an input channel. I am new in this domain please help me.

• I don't understand your question – air flow (a scalar value) over time is already a 1D signal. – Marcus Müller Aug 21 '18 at 20:59

The airflow signal is usually respiratory input in a polysonogram. If your input files are .EDF files, then you need to extract the airflow signal only, which is a 1-D signal. I think there are many implementations of edfread.m that are available online.