Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Graph Neural Network for Cerebral Blood Flow Prediction With Clinical Datasets

Seungyeon Kim, Wheesung Lee, Sung-Ho Ahn, Do-Eun Lee, Tae-Rin Lee

TL;DR

A graph neural network is proposed to predict blood flow and pressure in previously unseen cerebral vascular network structures that were not included in training data, demonstrating the potential of the GNN for real-time cerebrovascular diagnostics, particularly in handling intricate and pathological vascular networks.

Abstract

Accurate prediction of cerebral blood flow is essential for the diagnosis and treatment of cerebrovascular diseases. Traditional computational methods, however, often incur significant computational costs, limiting their practicality in real-time clinical applications. This paper proposes a graph neural network (GNN) to predict blood flow and pressure in previously unseen cerebral vascular network structures that were not included in training data. The GNN was developed using clinical datasets from patients with stenosis, featuring complex and abnormal vascular geometries. Additionally, the GNN model was trained on data incorporating a wide range of inflow conditions, vessel topologies, and network connectivities to enhance its generalization capability. The approach achieved Pearson's correlation coefficients of 0.727 for pressure and 0.824 for flow rate, with sufficient training data. These findings demonstrate the potential of the GNN for real-time cerebrovascular diagnostics, particularly in handling intricate and pathological vascular networks.

Graph Neural Network for Cerebral Blood Flow Prediction With Clinical Datasets

TL;DR

A graph neural network is proposed to predict blood flow and pressure in previously unseen cerebral vascular network structures that were not included in training data, demonstrating the potential of the GNN for real-time cerebrovascular diagnostics, particularly in handling intricate and pathological vascular networks.

Abstract

Accurate prediction of cerebral blood flow is essential for the diagnosis and treatment of cerebrovascular diseases. Traditional computational methods, however, often incur significant computational costs, limiting their practicality in real-time clinical applications. This paper proposes a graph neural network (GNN) to predict blood flow and pressure in previously unseen cerebral vascular network structures that were not included in training data. The GNN was developed using clinical datasets from patients with stenosis, featuring complex and abnormal vascular geometries. Additionally, the GNN model was trained on data incorporating a wide range of inflow conditions, vessel topologies, and network connectivities to enhance its generalization capability. The approach achieved Pearson's correlation coefficients of 0.727 for pressure and 0.824 for flow rate, with sufficient training data. These findings demonstrate the potential of the GNN for real-time cerebrovascular diagnostics, particularly in handling intricate and pathological vascular networks.

Paper Structure

This paper contains 11 sections, 4 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Method
  3. Vessel Segmentation and Refinement
  4. Link Prediction and Network Construction
  5. Mathematical Model for Blood Flow Prediction
  6. GNN Model
  7. Experiments
  8. Experimental Setup
  9. Discussion
  10. Compliance with Ethical Standards
  11. Acknowledgments

Figures (3)

  • Figure 1: Overall architecture of the proposed gROM framework for predicting flow rate and pressure in cerebral artery networks. The process begins with cerebral artery network data, which is transformed into NetworkX graphs. These graphs are further converted into DGL graphs containing node and edge features. The gROM takes these features along with flow rate ($f$) and pressure ($p$) inputs and applies the gROM for prediction. The intermediate predicted flow rate ($f'$) and pressure ($p'$) are first computed, and after further refinement, the final predicted flow rate ($\hat{f}$) and pressure ($\hat{p}$) are obtained. These final predicted values are compared to the ground truth using the Mean Absolute Error (MAE) as the loss function.
  • Figure 2: Visualization of pressure and flow rate using Dr. NEAR flow, comparing ground truth and model predictions.
  • Figure 3: Comparison of predicted and ground truth values for (a) pressure and (b) flow rate. The red dashed line represents the ideal regression line.