Representation learning in PET scans enhanced by semantic and 3D position specific characteristics

Theodoros P. Vagenas, Maria Vakalopoulou, Christos Sachpekidis, Antonia Dimitrakopoulou-Strauss, George K. Matsopoulos

Keywords: AI for Health

TL;DR: Representation learning for regions in PET scans is enhanced through the integration of semantic and 3D position-specific features to produce representations that support clinical diagnosis and characterization of regions of interest.

Abstract:

T. P. Vagenas, M. Vakalopoulou, C. Sachpekidis, A. Dimitrakopoulou-Strauss and G. K. Matsopoulos, "Representation learning in PET scans enhanced by semantic and 3D position specific characteristics," in IEEE Transactions on Medical Imaging, doi: 10.1109/TMI.2025.3566996
https://ieeexplore.ieee.org/abstract/document/10985918
Representation learning methods that discover task and/or data-specific characteristics are very popular for a variety of applications. However, their application to 3D medical images is restricted by the computational cost and their inherent subtle differences in intensities and appearance. In this paper, a novel representation learning scheme for extracting representations capable of distinguishing high-uptake regions from 3D $^{18}$F-Fluorodeoxyglucose positron emission tomography (FDG-PET) images is proposed. In particular, we propose a novel position-enhanced learning scheme effectively incorporating semantic and position-based features through our proposed Position Encoding Block (PEB) to produce highly informative representations. Such representations incorporate both semantic and position-aware features from high-dimensional medical data, leading to general representations with better performance on clinical tasks. To evaluate our method, we conducted experiments on the challenging task of classifying high-uptake regions as either non-tumor or tumor lesions in Metastatic Melanoma (MM). MM is a type of cancer characterized by its rapid spread to various body sites, which leads to low survival rates. Extensive experiments on an in-house and a public dataset of whole-body FDG-PET images indicated an increase of 10.50\% in sensitivity and 4.89\% in F1-score against the baseline representation learning scheme while also outperforming state-of-the-art methods for classifying MM regions of interest. The source code will be available at https://github.com/theoVag/Representation-Learning-Sem-Pos.

Submission Number: 78