BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Introduction of non-image PET data transformation to image-form fo
 r classification using Convolutional Neural Networks
DTSTART;VALUE=DATE-TIME:20211011T080000Z
DTEND;VALUE=DATE-TIME:20211011T082000Z
DTSTAMP;VALUE=DATE-TIME:20261008T191450Z
UID:indico-contribution-249@indico.koza.if.uj.edu.pl
DESCRIPTION:Speakers: Lech Raczyński (Department of Complex Systems\, Nat
 ional Centre for Nuclear Research)\nRecently\,  Convolutional  Neural  Net
 works  (CNNs)  [1]  have  achieved  state-of-the-art performance in many a
 reas including medical sciences\, and are the method of choice commonly us
 ed for data recognition or classification. CNNs have proven to work most 
 efficiently on 2-dimensional data that are in form of images.  \n\nIn case
  of Positron Emission Tomography (PET) [2\,3] studies\, CNN may be applied
  directly to the  reconstructed  distribution  of  radioactive  tracer  in
 jected  to  the  patient's  body\,  as  for  example  a pattern recognitio
 n tool. Nonetheless\, much PET data still exists in non-image format and t
 herefore \nopens challenging research questions on whether they can be eff
 ectively trained using CNN. Examples of such tasks are estimation of time-
 of-flight from signals registered in scintillators [4] or classification 
 of coincidence events acquired by PET scanner [5].  \n\nThe goal of this p
 resentation is the introduction of scheme of non-image data transformation
  into 2-dimensional matrices\, as a preparation stage for classification b
 ased on CNNs. The first work to apply CNN on different kinds of non-image 
 datasets\, e.g.\, gene expression or text information\, was \nproposed  in
   [6].  Here\,  we  will  focus  mainly  on  the  problem  of  processing 
  of  vectors  with  small number of features  in  comparison  to the  numb
 er  of pixels  in the  output  images. As  an  example\,  a discussion of 
 application of the proposed methodology to classification of PET coinciden
 ce events will provided [7]. \n\n**References** \n \n[1]  Lecun Y\, Bengio
  Y and Hinton G. Deep learning\, *Nature* vol. 521\, pp. 436\, 2015.  \n\n
 [2]  Humm J L\, Rosenfeld A\, Del Guerra A. From PET detectors to PET scan
 ners\, \n*European J. of Nucl. Med. & Mol. Imag.* vol. 30\, pp. 1574\, 200
 3. \n  \n[3]  Bailey D L. *Positron Emission Tomography: Basic Sciences*\,
  Springer-Verlag\,  New York\, 2005. \n\n[4]  Berg E and Cherry S R. Using
  convolutional neural networks to estimate time-of-flight  from PET detect
 or waveforms\, *Phys. Med. Biol.*\, vol. 63\, pp. 02LT01\, 2018.  \n\n[5] 
  Bielecki J. Application of the machine learning methods to the multi-phot
 on event classification in the J-PET scanner\, *Msc thesis*\, *Warsaw Uni
 versity of Technology*\, 2019.   \n\n[6]  Sharma A\, Vans E\, Shigemizu D\
 , Boroevich K A and Tsunoda T. Deepinsight: A methodology  transform a non
 -image data to an image for convolution neural network architecture. *Scie
 ntific Report*s vol. 9\, pp. 11399\, 2019. \n  \n[7]  Konieczka P. Convolu
 tional Neural Networks in classification of multi-photon coincidences in J
 -PET scanner\, *1st Symposium on Theranostics*\, 9-11 October 2021.\n\nhtt
 ps://indico.koza.if.uj.edu.pl/event/4/contributions/249/
LOCATION:Theranostics Center / on-line
URL:https://indico.koza.if.uj.edu.pl/event/4/contributions/249/
END:VEVENT
END:VCALENDAR