PYNQ: PYTHON PRODUCTIVITY

Zynq for Ultrasound Image AI Processing

Support
#1

Looking for initial part selection / dev board guidance. I am veterinary ultrasound designer starting on 5th generation low power design.

For our next design we were considering a smaller 32 channel Zynq 7020 design as we cannot have fans (bovine use on farms), i.e. passive cooling only.

Is there any IP or Papers on using this for ML of image pixel patterns to distinguish anatomy? If we can make power and budget work, I would consider Ultrascale series.

I appreciate your guidance in advance.- Patrick

#2

The PYNQ support forum probably isn’t the best place to ask this. Your application sounds quite specialist. I’m not sure how different your ML application is to any other ML application. .

You could try searching for academic papers on the usual sites, or asking on the main Xilinx forum. I’m not sure if you will get a better answer there. You would probably be better doing your own research/google/search the usual places for academic papers/journals etc.

On the Ultrasound part, I do remember one (non-PYNQ) design from the Open Hardware design competition:
“Inexpensive 5 Watt, 1024-channel 2D & 3D Ultrasound Imager”
Second project on this page:

Cathal