Novel Model for Classifying Prostate Cancer Lymph Node Metastasis

Strategic Insights -

Researchers from Northeastern University in Shenyang, People's Republic of China, have proposed a novel model for classifying prostate cancer lymph node metastasis using dual-tracer PET/CT images. The model, MECF-Net, integrates 18F-PSMA-1007 and 18F-FDG PET/CT images to enhance feature perception and spatial dimensions. This research has the potential to improve treatment options and prognosis for men with prostate cancer. The model's effectiveness was demonstrated through experimental results on private and public datasets, showcasing superior performance compared to state-of-the-art networks.

Key Takeaways:

  • MECF-Net is a novel model for classifying prostate cancer lymph node metastasis based on dual-tracer PET/CT images.
  • The model integrates 18F-PSMA-1007 and 18F-FDG PET/CT images to enhance feature perception and spatial dimensions.
  • MECF-Net consists of a Multi-Scale Feature Extraction (MSFE) branch, Local-Global Feature Complementary Fusion (LGFCF) branch, and Multi-Feature Adaptive Fusion (MF-AF) module.
  • The model achieved 0.9269 and 0.885 in accuracy (ACC) on private and public datasets, respectively, and 0.9156 and 0.8838 in area under the curve (AUC) on private and public datasets, respectively.
  • The research has been peer-reviewed and published in Biomedical Signal Processing and Control.
  • The study's findings demonstrate the effectiveness of MECF-Net in classifying prostate cancer lymph node metastasis and its potential to improve treatment options and prognosis.

Statistics:

  • ACC on private dataset: 0.9269
  • ACC on public dataset: 0.885
  • AUC on private dataset: 0.9156
  • AUC on public dataset: 0.8838
  • Number of researchers involved: 5 (Huiyan Jiang, Junchen Hao, Yan Cui, Xuena Li, and Guoyu Tong)

Sources:

  • "Mecf-net: a Prostate Cancer Lymph Node Metastasis Classification Method Based On 18f-psma-1007 and 18f-fdg Dual-tracer Pet/ct Image Feature Optimization" (Biomedical Signal Processing and Control, 2025; 105)
  • NewsRx. Researchers from Northeastern University Report Recent Findings in Prostate Cancer (Mecf-net: a Prostate Cancer Lymph Node Metastasis Classification Method Based On 18f-psma-1007 and 18f-fdg Dual-tracer Pet/ct Image Feature Optimization). Health & Medicine Week. July 11, 2025; p 6148.