Few-Shot Classification and Anatomical Localization of Tissues in SPECT Imaging

2025-02-10

AI Summary

To address the scarcity of tissue annotation data in SPECT medical imaging, this paper proposes a joint classification and anatomical localization framework for cardiac regions—namely, ventricles, myocardium, and liver—under few-shot learning settings. Methodologically, we adapt Prototypical Networks for the first time to SPECT few-shot classification and enhance the PRNet architecture to build an encoder–decoder localization model tailored for 2D SPECT images, incorporating skip connections to strengthen spatial modeling. The input consists of cropped cardiac-region images; the classification branch extracts prototype features via a pre-trained ResNet-18 backbone, while the localization branch performs end-to-end regression of tissue distributions. Experimental results demonstrate a validation classification accuracy of 93.33% and a PRNet reconstruction loss of 1.395—both significantly outperforming baseline methods—thereby validating the feasibility of concurrently achieving high-accuracy classification and geometric localization under few-shot conditions.

Technology Category

• Computer Vision: Medical and Biological Imaging
• Machine Learning: Calibration & Uncertainty Quantification
• Intelligent Robots: Localization, Mapping, and Navigation

Application Category

• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Security and Privacy: Large-scale security measurements

Abstract

Accurate classification and anatomical localization are essential for effective medical diagnostics and research, which may be efficiently performed using deep learning techniques. However, availability of limited labeled data poses a significant challenge. To address this, we adapted Prototypical Networks and the Propagation-Reconstruction Network (PRNet) for few-shot classification and localization, respectively, in Single Photon Emission Computed Tomography (SPECT) images. For the proof of concept we used a 2D-sliced image cropped around heart. The Prototypical Network, with a pre-trained ResNet-18 backbone, classified ventricles, myocardium, and liver tissues with 96.67% training and 93.33% validation accuracy. PRNet, adapted for 2D imaging with an encoder-decoder architecture and skip connections, achieved a training loss of 1.395, accurately reconstructing patches and capturing spatial relationships. These results highlight the potential of Prototypical Networks for tissue classification with limited labeled data and PRNet for anatomical landmark localization, paving the way for improved performance in deep learning frameworks.

Problem

• Few-shot classification SPECT imaging
• Anatomical localization limited labeled data
• Deep learning techniques medical diagnostics

Innovation

• Utilized Prototypical Networks for tissue classification
• Adapted PRNet for anatomical landmark localization
• Achieved high accuracy with limited labeled data