Development of message passing-based graph convolutional networks for classifying cancer pathology reports

TL;DR

This paper introduces a new graph-based neural network model for efficiently classifying cancer pathology reports and extracting clinical information.

Contribution

The novel FastMPN model improves memory efficiency and flexibility in graph convolutional networks for text classification.

Findings

FastMPN achieved equivalent or better performance than MT-CNN in clinical information extraction tasks.

The model trained a large dataset in under 3 minutes per epoch using a single V100 GPU.

Highly competitive results were obtained for tumor-related information extraction from pathology reports.

Abstract

Applying graph convolutional networks (GCN) to the classification of free-form natural language texts leveraged by graph-of-words features (TextGCN) was studied and confirmed to be an effective means of describing complex natural language texts. However, the text classification models based on the TextGCN possess weaknesses in terms of memory consumption and model dissemination and distribution. In this paper, we present a fast message passing network (FastMPN), implementing a GCN with message passing architecture that provides versatility and flexibility by allowing trainable node embedding and edge weights, helping the GCN model find the better solution. We applied the FastMPN model to the task of clinical information extraction from cancer pathology reports, extracting the following six properties: main site, subsite, laterality, histology, behavior, and grade.

We evaluated the clinical task performance of the FastMPN models in terms of micro- and macro-averaged F1 scores. A comparison was performed with the multi-task convolutional neural network (MT-CNN) model. Results show that the FastMPN model is equivalent to or better than the MT-CNN.

Our implementation revealed that our FastMPN model, which is based on the PyTorch platform, can train a large corpus (667,290 training samples) with 202,373 unique words in less than 3 minutes per epoch using one NVIDIA V100 hardware accelerator. Our experiments demonstrated that using this implementation, the clinical task performance scores of information extraction related to tumors from cancer pathology reports were highly competitive.

Figures