biosignal processing
Budget: £10 – £20 GBP
Develop an algorithm in MATLAB with integrated custom functions to read PPG signals and calculate
all the PRV variables and graphs printed in the PRV calculations list below
The algorithm should detect each peak and foot of every PPG pulse and calculate the time
delay to the next (or previous) peak or foot to produce a time series for every recorded PPG
signal. All the metrics you calculate should be stored in the same structure. You will need one
structure per PPG file.
• DO NOT store the original PPG or any resampled PPG signal in any structures you create.
• Save one array element per heartbeat (Note: this is ‘unevenly sampled’ data as the time
interval between elements is not constant).
• Make sure your beat count is correct (i.e. approximately 50-120 beats per minute!).
• Remove any outlying NN intervals (i.e. those caused by artefacts etc) using a custom function
named “myremovenns”.
• Work through the list on the following page to produce your PRV analysis. You will need 3
custom functions:
1. “myprvstats” – This function should produce items 1 to 5 from the list below
2. “mypoincareplot” – This function will produce the data and plot the graph for
the Poincare plot, item number 6 in the list below
3. “myprvfdomain” – This function calculates all the frequency domain statistics,
items 7 to 10 in the list below.
• Note that for the frequency domain measurements you will need to resample the pulse-to-pulse
all the PRV variables and graphs printed in the PRV calculations list below
The algorithm should detect each peak and foot of every PPG pulse and calculate the time
delay to the next (or previous) peak or foot to produce a time series for every recorded PPG
signal. All the metrics you calculate should be stored in the same structure. You will need one
structure per PPG file.
• DO NOT store the original PPG or any resampled PPG signal in any structures you create.
• Save one array element per heartbeat (Note: this is ‘unevenly sampled’ data as the time
interval between elements is not constant).
• Make sure your beat count is correct (i.e. approximately 50-120 beats per minute!).
• Remove any outlying NN intervals (i.e. those caused by artefacts etc) using a custom function
named “myremovenns”.
• Work through the list on the following page to produce your PRV analysis. You will need 3
custom functions:
1. “myprvstats” – This function should produce items 1 to 5 from the list below
2. “mypoincareplot” – This function will produce the data and plot the graph for
the Poincare plot, item number 6 in the list below
3. “myprvfdomain” – This function calculates all the frequency domain statistics,
items 7 to 10 in the list below.
• Note that for the frequency domain measurements you will need to resample the pulse-to-pulse