BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Foundation Models in Imaging: From Pneumonia Risk Stratification t
 o Tomographic Reconstruction
DTSTART:20260521T093000Z
DTEND:20260521T110000Z
DTSTAMP:20260515T005700Z
UID:indico-event-18901@indico.psi.ch
CONTACT:benjamin.bejar@psi.ch
DESCRIPTION:Speakers: Amir Ehsan Khorashadizadeh\n\nAbstract: Recent adva
 nces in foundation models have shown how large-scale pretraining can signi
 ficantly improve medical image analysis\, especially in low-data settings.
  In this talk\, I will first present a pneumonia risk stratification study
  based on 3D chest CT imaging in collaboration with Kantonsspital Aarau. W
 e show that features extracted from a pretrained segmentation foundation m
 odel combined with a simple linear classifier significantly outperform str
 ong convolutional neural networks trained from scratch. This demonstrates 
 the strength of transferable representations learned from large and divers
 e datasets. Motivated by these findings\, the talk then explores a broader
  question relevant to computational and scientific imaging: can foundation
  models also transform tomographic image reconstruction? I will present o
 ur ongoing work on developing a foundation model for tomographic reconstru
 ction that learns generalizable and scalable priors across datasets and im
 aging geometries\, enabling efficient adaptation to downstream reconstruct
 ion tasks.\nTEAMS link \n\nhttps://indico.psi.ch/event/18901/
LOCATION:OHSA/B17
URL:https://indico.psi.ch/event/18901/
END:VEVENT
END:VCALENDAR