Figure 6 | Scientific Reports

Figure 6

From: Deep learning-based classification of blue light cystoscopy imaging during transurethral resection of bladder tumors

Figure 6

Schematic diagram of the CNN fine-tuning considered for identifying bladder cancer. Each pre-trained CNN was fine-tuned by appending two batch normalization layers, a global average pooling layer, dropout layers with the probability of 50%, a dense layer to improve cancer identification, and Softmax activation layer. The last layer delivers different label probabilities for each input image with respect to each classification task.

Back to article page