BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:From Signal Acquisition to Image Reconstruction: Potential Applica
 tions of Machine Learning in Positron Emission Tomography
DTSTART;VALUE=DATE-TIME:20240606T070000Z
DTEND;VALUE=DATE-TIME:20240606T074000Z
DTSTAMP;VALUE=DATE-TIME:20260909T224708Z
UID:indico-contribution-164-1567@events.ncbj.gov.pl
DESCRIPTION:Speakers: Georg Schramm ()\nPositron Emission Tomography (PET)
  is a functional medical imaging technique that allows for the visualizati
 on and measurement of metabolic processes in the body by detecting pairs o
 f 511-keV gamma rays originating from a tracer molecule\nlabeled with a po
 sitron emitter.\n\nDespite its advanced capabilities\, PET imaging faces s
 ignificant challenges\, including high noise levels and limited spatial re
 solution of the acquired data\, which severely hampers the diagnostic qual
 ity of the reconstructed images.\n\nIn addition to classical algorithms tr
 aditionally used for signal processing\, image reconstruction\, and image 
 post-processing\, machine learning (ML) based algorithms are now being exp
 lored to enhance the quality of PET raw data and the quality of reconstruc
 ted PET images.\n\nThis talk provides an overview of the current applicati
 ons of ML in PET imaging\,\nencompassing various stages from signal acquis
 ition to image reconstruction and post-processing.\n\nAdditionally\, the p
 resentation addresses the current challenges in the field and explores fut
 ure needs and directions for a sustainable and successful integration of M
 L in PET imaging.\n\nhttps://events.ncbj.gov.pl/event/314/contributions/15
 67/
LOCATION:
URL:https://events.ncbj.gov.pl/event/314/contributions/1567/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Machine Learning-based Scatter Correction for a Dual-Panel Positro
 n Emission Mammography Scanner
DTSTART;VALUE=DATE-TIME:20240606T080500Z
DTEND;VALUE=DATE-TIME:20240606T083000Z
DTSTAMP;VALUE=DATE-TIME:20260909T224708Z
UID:indico-contribution-164-1496@events.ncbj.gov.pl
DESCRIPTION:Speakers: Fernando Moncada-Gutiérrez (Instituto de Física\, 
 UNAM)\nPositron Emission Mammography (PEM) is a Nuclear Medicine technique
  for breast imaging based on a dedicated scanner assembled with parallel d
 ual-panel detector arrays. Patient positioning in close contact with the s
 canner enhances spatial resolution and sensitivity in comparison with ring
 -based scanners\, but this geometry hinders the adaptation of conventional
  attenuation and scatter correction methods\, which affects the quantitati
 ve assessment of studies. In this work we trained several machine learning
  algorithms for scatter correction with list-mode data from a Monte Carlo 
 simulation of a PEM prototype being built in our lab. The features for thi
 s binary classification problem were energy and position of detection\, wh
 ere energy had the higher feature importance in agreement with traditional
  methods. The best results were found with a Random Forest of 38 estimator
 s and a maximum depth of 7\, which reduced the scatter fraction of a study
  of 1 million events from 11% to 4% in 2 seconds.\n\nhttps://events.ncbj.g
 ov.pl/event/314/contributions/1496/
LOCATION:
URL:https://events.ncbj.gov.pl/event/314/contributions/1496/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Using 3D CNNs for distortion corrections in PET imaging
DTSTART;VALUE=DATE-TIME:20240606T074000Z
DTEND;VALUE=DATE-TIME:20240606T080500Z
DTSTAMP;VALUE=DATE-TIME:20260909T224708Z
UID:indico-contribution-164-1494@events.ncbj.gov.pl
DESCRIPTION:Speakers: Konrad Klimaszewski (National Centre for Nuclear Res
 earch)\nIn Positron Emission Tomography the problem of image distortion du
 e to scattered photons or accidental coincidences becomes more pronounced 
 for large field-of-view scanners capable of measuring the whole patient in
  one scan. We propose a novel method of encoding coincidence event informa
 tion to enhance the efficiency of noise filtration classification. The pro
 posed encoding enables the usage of Convolutional Neural Networks as featu
 re extractors in the classification task. We take advantage of the voxel n
 ature of underlying data and evaluate the performance of the 3-D CNN netwo
 rk to classify true\, scattered and accidental coincidences for imaging qu
 ality improvement with large field-of-view PET scanners.\n\nhttps://events
 .ncbj.gov.pl/event/314/contributions/1494/
LOCATION:
URL:https://events.ncbj.gov.pl/event/314/contributions/1494/
END:VEVENT
END:VCALENDAR