A Python script that synthesizes audio from an spectral Doppler image generated by an ultrasound machine.
Some ultrasound machines are capable of measuring the flow of blood and can generate a graphical and audible representation of that flow. These machines are often used to measure fetal heartbeats and give expecting mothers a way to see and hear the heartbeat of an unborn child. Doctors can provide a printout from these machines to take home, but audio recordings aren’t usually provided.
A friend posted on Facebook with a picture of the printout she received from one such ultrasound machine. She didn’t take a recording with her phone at the time so all she had was a piece of paper that showed a spectrogram image. I did some research on ultrasound machines and hacked together a Python script that can process a scan of a printout and recreate the audio portion.
Here is the image I received to work with.
The script processes just the portion of the image above the axis line so I rotated, cropped, and cleaned it up using GIMP.
The cleaned image was then processed and the following audio was generated.
Full instructions on installation and usage of the script are included in the link below.
Note: the first 2 logs were written long before they were published. The dates on them vary slightly so they sort correctly.
- Ultrasound Synthesizer Script Python script that generates audio from a spectral Doppler image