Skip to main content
. 2025 Jul 24;15(15):1863. doi: 10.3390/diagnostics15151863
Algorithm 1. Features Concatenation-Based Brain Tumor Classification (FCBTC) Framework

  1. Fine-tune baseline models M1 and M2 with dataset

  2. Extract feature vectors f1 and f2 from M1 and M2 after flattening layer.

  3. Perform concatenation of features: f_con = f1 ‖ f2

  4. The concatenated feature vector is fed into a MLP head z = MLP(f_con)

  5. Final classification by SoftMax layer c = Softmax(z)