Published by Nature Communications, 10 June 2022
The pathological identification of lymph node (LN) metastasis is demanding and tedious. Although convolutional neural networks (CNNs) possess considerable potential in improving the process, the ultrahigh-resolution of whole slide images hinders the development of a clinically applicable solution. We design an artificial-intelligence-assisted LN assessment workflow to facilitate the routine counting of metastatic LNs. Unlike previous patch-based approaches, our proposed method trains CNNs by using 5-gigapixel images, obviating the need for lesion-level annotations. Trained on 5907 LN images, our algorithm identifies metastatic LNs in gastric cancer with a slide-level area under the receiver operating characteristic curve (AUC) of 0.9936. Clinical experiments reveal that the workflow significantly improves the sensitivity of micrometastasis identification (81.94% to 95.83%, P < .001) and isolated tumor cells (67.95% to 96.15%, P < .001) in a significantly shorter review time (−31.5%, P < .001). Cross-site evaluation indicates that the algorithm is highly robust (AUC = 0.9829).
Fig: a Dataflow of ESCNN illustrates how patching mechanisms are embedded into image augmentation and the backpropagation algorithm (including forward pass and backward pass). b Procedure of the patch-based affine transformation. c Three threads run concurrently to fully utilize computing resources, including one for loading WSIs (Input/Output-intensive), one for patch-based image augmentation (bus-intensive), and one for patch-based backpropagation (GPU-intensive).
Full Article: https://www.nature.com/articles/s41467-022-30746-1