Prediction of Overall Survival Death < 2 Years in Diffuse Large B-Cell Lymphoma Based on Histological Images and Deep Neural Networks

Diffuse large B-cell lymphoma (DLBCL) is one of the most frequent subtypes of non-Hodgkin lymphoma (NHL). This study is a proof of concept computer vision exercise to support the feasibility of predicting the prognosis of DLBCL using only hematoxylin and eosin (H&E) histological images and deep learning. A conventional series of DLBCL of 114 cases was split into 2 prognostic groups according to the overall survival curve. The curve fitting and slope analysis showed a point of inflection at 2 years, which differentiated patients of “Dead < 2 years” with aggressive (b1 = -0.024), and “Others” with moderate clinical evolution (b1 = -0.003). Twenty different convolutional neural networks (CNNs) were used, and explainable artificial intelligence (XAI) was applied to identify the areas of the images that the network used for classification. The final model based on DarkNet-19 predicted prognosis groups with high performance (test set accuracy = 96.26%). The other performance parameters were precision (94.46%), recall (95.02%), false positive rate (3.07%), specificity (96.93%), and F1 score (94.74%). XAI, including grad-CAM, occlusion sensitivity, and image-LIME, confirmed that the CNN focused on the correct areas. Hybrid partitioning to prevent information leakage with patient-based analysis and image classification between DLBCL and 44 cases of reactive lymphoid tissue were also successfully performed. Correlation with the clinicopathological characteristics found that the Dead < 2 years group was correlated with stage III-IV, International Prognostic Index (IPI) High + High/intermediate, progressive disease, non-GCB cell-of-origin, CD10-, BCL2+, and Epstein-Barr virus (EBER)+. Analysis of the microenvironment, immune checkpoint, cell cycle, and germinal center markers showed that Dead < 2 years had higher IL10, PD-L1, and CD163, and lower E2F1 protein expressions. No differences were found for Ki67, CSF1R, CASP8, TNFAIP8, LMO2, MYC, MDM2, CDK6, and TP53 markers. In conclusion, the overall survival of DLBCL can be predicted using H&E histological images and deep learning using 2 years point (similar to POD24). This trained CNN could be used as a pretrained model for transfer learning in the future.