From imaging algorithms to quantum methods Seminar

Name: From imaging algorithms to quantum methods Seminar
Start: 2025-02-17T10:00:00+01:00
End: 2025-02-17T11:30:00+01:00
Location: Zoom

Monday 17 Feb 2025, 10:00 → 11:30 Europe/Warsaw

https://cern.zoom.us/j/66151941204?pwd=n7upvvZYibexBhbtyn5kvTpy36L0Wo.1 (Zoom)

https://cern.zoom.us/j/66151941204?pwd=n7upvvZYibexBhbtyn5kvTpy36L0Wo.1

Zoom

Konrad Klimaszewski (NCBJ), Wojciech Krzemien (NCBJ)

Description

Zoom link: https://cern.zoom.us/j/66151941204?pwd=n7upvvZYibexBhbtyn5kvTpy36L0Wo.1

Hide

## Title: Statistical reconstruction algorithm based on the method of moments in positronium imaging
## Presenter: Roman Y. Shopa
## Date: 17.02 2024

## Participants:

Wojciech Krzemień (WK)
Konrad Klimaszewski (KK)
Roman Shopa (RS)
Lech Raczynski (LR)
Mateusz Bała (MB)
Maciej Zajkowski (MZ)
Dominik Strzelecki (DS)
Aleksander Ogonowski (AO)

## Questions/Remarks:
WK: At what stage are the wj weights calculated?
RS: The answer is given on p.9, the weights are calculated with the algorithm analogical to MLEM algorithm. As input the estimate time per event is provided.

KK: Final result should be given by m?
RS: m is the positronium mean lifetime estimated for all decay mechanisms combined, m = w/activity.
The ortho-positronium lifetime is extracted using the method of moments.

WK:Do the obtained values of the lifetimes are comparable to the introduced ones in the MC?
RS: There is a bias. The results are systematically lower. It might be due to low statistics and/or median post-filter applied.

LR: Does the bias occur both in the background and signal regions?
RS: Yes, some negative bias observed, yet less pronounced as for the hot phantom spheres. Maybe there is still some problem in the MC scripts

KK: The last slide represent already orto-lifetime, after applying the method of moments?
RS: Yes.

MB: What is MPF?
RS: Median Post-Filter - is a convolution filter similar to Gaussian, applied to the final reconstructed image. One takes a voxel and its neighbors and calculates the median from them. Median filter works well for gradient and edge preservation, while taking the mean blurs them.

MB: What shape of the function E(\nu) is assumed for the methods of moments in the article given at p.13?
RS: Apparently, no shape is defined in the article, but maybe I had missed it. In the original article, moments are calculated numerically from the measured E(\nu).

MB: The assumption of intensity ratio between ortho and para =3 is for a vacuum?
RS: Good question. They assume it as a constant, but maybe the ratio may be violated inside the matter.
WK: For vacuum it is just a combinatorial relation: PS has one possible spin state while OPS has 3 possible spin states.

There are minutes attached to this event. Show them.

- 10:00 → 11:00
  
  Statistical reconstruction algorithm based on the method of moments in positronium imaging 1h
  
  Positronium imaging (PI) is the new technique, complementary to the positron emission tomography (PET), that aims to
  reconstruct 3D images of positronium lifetime in humans and animals. The existing PI algorithms are computationally expensive, require long scans and struggle with poor spatial resolution due to lower sensitivity caused by a need for an additional prompt gamma.
  
  In this talk, we shall review the SIMPLE-Moment (Statistical IMage reconstruction of Positron annihilation LifetimE by Moment weighting) -- a novel solution in PI that does not use curve-fitting and exhibit performance comparable to conventional PET reconstruction. In addition, a test result for our own data will be revealed.
  
  Speaker: Roman Shopa (National Centre for Nuclear Research)
  
  SIMPLE_moment_17-02-2025.pdf
- 11:00 → 11:30
  
  Discussion 30m

From imaging algorithms to quantum methods Seminar

https://cern.zoom.us/j/66151941204?pwd=n7upvvZYibexBhbtyn5kvTpy36L0Wo.1

Zoom

Your browser is out of date!