Artificial Intelligence Meets Whole Slide Images: Deep Learning Model Shapes an Immune-Hot Tumor and Guides Precision Therapy in Bladder Cancer

Figure 7

AI clusters predicted classical molecular subtypes and therapeutic opportunities in the TCGA-BLCA cohort. (a, b) The correlations between AI clusters and seven classical molecular subtype classifications. (c) The correlations between the AI cluster and the enrichment scores of several therapeutic signatures, such as targeted therapy and radiotherapy. (d) ROC curves showed the accuracy of the AI score in predicting classical molecular subtypes. (e, f) The overall mutation rates of neoadjuvant chemotherapy-related genes in the AI cluster subgroups. (g) The difference on the therapeutic sensitivities of six chemotherapy drugs, including cisplatin, docetaxel, gemcitabine, paclitaxel, bleomycin, and camptothecin.

(a)

(b)

(c)

(d)

(e)

(f)

(g)