New publication entitled: “Advanced hybrid deep learning model for enhanced evaluation of osteosarcoma histopathology images” was published in the Frontiers in Medicine.
Recent advances in machine learning, particularly deep learning with CNNs and vision transformers (ViTs), are improving cancer detection through medical image analysis. This study targets osteosarcoma (OS), the most common bone cancer in children and adolescents, using a novel hybrid model combining CNN and ViT to analyze histopathological images. The model classifies images into four categories and achieved outstanding performance on the Cancer Imaging Archive dataset, setting a new benchmark in OS diagnosis with over 99% accuracy.
The study is from Dr. Sepideh Hatamikia, MSc. Arezoo Borji and Dr. Gernot Kronreif.
For more information, please visit:
