Closed WOLVS closed 2 years ago
jgroehl
commented
2 years ago We have integrated a baseline delay and sum back-projection for arbitrary detection geometries now. We are still missing typical pre- and postprocessing steps throughout the literature. Maybe we can use this task to collect a list of pre-and post-processing steps/algorithms that are frequently used with PAI?
WOLVS
commented
2 years ago @jgroehl sorry for not being able to join regular meetings for a while. Do you mean collect pre-and post-processing algorithms for PAI ? What rules will be applied for collecting those algorithms. Post-processing will be mainly supervised machine learning methods. I have done some work using a github page . Are you aiming for something similar with this website https://quality-diversity.github.io/ ? Alternatively I can start to think about how to build a deep learning based benchmark. Unfortunately, I can not make tomorrow's meeting 15:00 to 15:30. But I am free from 16:00 to 17:00. If you are free, I am happy to arrange a chat with you.
jgroehl
commented
2 years ago Since backprojection is integrated into main, I think we can close this issue.
Image reconstruction task to reproduce universal back-project algorithm Universal back-projection algorithm for photoacoustic computed tomography Journal: Physical Review (1167 citations) DOI: https://doi.org/10.1103/PhysRevE.71.016706
The paper describes the principle and mathematical foundation of the back-projection algorithm (also referred to as “delay-and-sum” throughout the literature for the discrete implementation of the algorithm). The algorithm has been implemented numerous times in different variations and with different additional features and considerations, some of which are available open-source.