Convolutional Neural Networks for multi-organ segmentation of SPECT projections

Type: MA thesis

Status: finished

Date: March 1, 2020 - October 8, 2020

Supervisors: Maximilian Reymann, Dr. Michal Cachovan (Siemens Healthineers), Dr. Philipp Ritt ( Nuklearmedizinische Klinik, Universitätsklinikum Erlangen)

In this work we investigate the usage of deep learning techniques on SPECT data solving a multi-organ segmentation problem. We extract projections from 21 Lu-177 MELP SPECT scans and obtain the corresponding ground truth labels from the accompanied CT scans by forward-projection of 3D CT organ segmentations. We train a U-Net to predict the area of the kidney, spleen, liver, and background seen in the projection data, using a weighted dice loss between prediction and target labels to account for class imbalance.
With our method we achieved a mean dice coefficient of 72 % on the test set, encouraging us to perform further experiments using the U-Net.