Table 4.

The classification accuracy results (%) using the DL approach on the KIMIA Path960 dataset (Bi means the fine-tuned blocks).

Block-Wise Fine-Tuning Strategy	Accuracy	Recall	Precision
B6	68.06	68.00	74.51
B6-B5	89.58	90.13	91.17
B6-B4	96.53	96.49	96.37
B6-B3	98.26	97.96	98.12
B6-B2	98.26	98.31	98.48
B6-B1	97.92	98.02	98.23