Diffusion Model with Multi-Scale and Attention Mechanisms for SPECT Image Generation
Abstract
Single-photon emission computed tomography (SPECT) is crucial for detecting bone metastases, yet deep learning models are limited by scarce high-quality data. We present a data-augmented Conditional Multi-Scale Diffusion model (C-MSDiff) that integrates a multi-scale convolutional module (MSConv) and a Convolutional Block Attention Module (CBAM) within the conditional diffusion process. C-MSDiff fuses multi-level features and adaptively highlights lesion regions while maintaining consistency between local details and global anatomy. In comprehensive evaluations, it consistently surpasses contemporary generative methods on standard image-quality criteria and better preserves clinically relevant structures. When used to generate training data, C-MSDiff further improves downstream bone-metastasis classification. These results show that C-MSDiff enhances SPECT image quality and strengthens computer-aided diagnosis, offering a practical path to reliable decision support under limited-data conditions.
Read the paper
Related studies
- Learning from First-Generation Qualitative Approaches in the IS Discipline: An Evolutionary View and Some Implications for Authors and Evaluators (PART 1/2)
- Commentary on the clinical and preclinical dosage limits of interstitially administered magnetic fluids for therapeutic hyperthermia based on current practice and efficacy models
- Prosthesis: New Technological Opportunities and Innovative Biomedical Devices
- The history and future of neural modeling for cochlear implants
- Ryanodine receptors: waking up from refractoriness
- Novel Transgenic Rabbit Model Sheds Light on the Puzzling Role of Matrix Metalloproteinase-12 in Atherosclerosis
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.