Techn. Fakultät Website deprecated and outdated. Click here for the new site. FAU-Logo

Population Modelling

The population modelling group is dedicated to analyse data collected from many individuals in order to be able to model variations within entire populations.

Colloquium Time Table

Responsible PersonTitle


Colloquium Discussion
15.01.2019Shuqing Chen

Journal Club - U-Net Variations & Progressions  1: Opens external link in new windowFully automated organ segmentation in male pelvic CT images 

22.01.2019Franziska KopteMT Final: Liver Tumor Segmentation Using an Interactive Deep Learning Architecture
29.01.2019Theresa Götz3D-2D Neural Networks for dose estimation in nuclear medicine

Maximilian Reymann

Xia Zhong

MT Intro: Denoising in SPECT using Deep Learning

JC: Capsules for Object Segmentation (Opens external link in new windowSegCaps)

05.03.2019Philipp RoserJC: Opens external link in new windowA Probabilistic U-Net for Segmentation of Ambiguous Images
12.03.2019BVM AuthorsBVM 2019 Rehearsal Session
19.03.2019No colloquium due to BVM
26.03.2019No colloquium due to PRS
09.04.2019Maximilian ReymannMT Final: Denoising in SPECT using Deep Learning
07.05.2019Tobias SchmidtBT Intro: Evaluation of Augmented Reality Devices for Employee Training using Machine Learning
11.06.2019Mohammad ZakeriMT Intro & Final: Automatic Glottis segmentation from High-Speed video endoscopy of the larynx
25.06.2019Harb Alnasser AlabdalrahimMT Intro: Automatic Evaluation of the Clock-Drawing Dementia Test using Deep Learning
23.07.2019Tobias SchmidtBT Final: Evaluation of Augmented Reality Devices for Employee Training using Machine Learning
13.08.2019Harb Alnasser AlabdalrahimMT Final: Automatic Evaluation of the Clock-Drawing Dementia Test using Deep Learning
20.08.2019Sina GhasemiMT Final: Head Pose Estimation for Patient Localization Using Deep Neural Network


Mailing list subscription management page for students and guests.

Prior Sessions Opens internal link in current windowLink

Running Projects

X-Ray Imaging Using a Patient Model

This project focuses on estimate a patient model and model based X-ray imaging application. The current research focus is on anatomical patient modelling and model based skin and scatter estimation.

Quantitative DECT with prior atlas knowledge

This project aims at decomposing the materials with anatomy knowledge automatically. An atlas provides the prior knowledge - the anatomic structures. The anatomy knowledge allows then automatic decomposition and classification of the different materials.

Quantification and Respiratory Motion Management for SPECT Imaging

Population-based metrics for diagnosis and patient stratification are important aids for nuclear medicine physicians when dealing with various cardiological, neurological, and oncological indications. A necessary input to such approaches based on SPECT is a standardized, artifact-free dataset. The goals of this project include improvement of standardization through quantitative imaging and reduction of respiratory motion artifacts through data-driven techniques.

Finished Projects

Time-of-Flight 4-D Foot Scan

3-D Reconstruction of the shape of human feet during motion.

Attenuation Correction for Hybrid Imaging Systems

Attenuation maps for correcting PET data needs to be derived from MR information in current PET/MR hybrid systems

Catheter Contact Force

Catheter contact force assessment and evaluation for cardiac ablation procedures.

Motion Compensation

Cardiac and respiratory motion compensation for atrial fibrillation ablation procedures.