Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

PINNing Cerebral Blood Flow: Analysis of Perfusion MRI in Infants using Physics-Informed Neural Networks

Christoforos Galazis, Ching-En Chiu, Tomoki Arichi, Anil A. Bharath, Marta Varela

TL;DR

A new spatial uncertainty-based physics-informed neural network (PINN), SUPINN, is proposed, to estimate CBF and other parameters from infant ASL data, demonstrating the successful modification of PINNs for accurate multi-parameter perfusion estimation from noisy and limited ASL data in infants.

Abstract

Arterial spin labeling (ASL) magnetic resonance imaging (MRI) enables cerebral perfusion measurement, which is crucial in detecting and managing neurological issues in infants born prematurely or after perinatal complications. However, cerebral blood flow (CBF) estimation in infants using ASL remains challenging due to the complex interplay of network physiology, involving dynamic interactions between cardiac output and cerebral perfusion, as well as issues with parameter uncertainty and data noise. We propose a new spatial uncertainty-based physics-informed neural network (PINN), SUPINN, to estimate CBF and other parameters from infant ASL data. SUPINN employs a multi-branch architecture to concurrently estimate regional and global model parameters across multiple voxels. It computes regional spatial uncertainties to weigh the signal. SUPINN can reliably estimate CBF (relative error $-0.3 \pm 71.7$), bolus arrival time (AT) ($30.5 \pm 257.8$), and blood longitudinal relaxation time ($T_{1b}$) ($-4.4 \pm 28.9$), surpassing parameter estimates performed using least squares or standard PINNs. Furthermore, SUPINN produces physiologically plausible spatially smooth CBF and AT maps. Our study demonstrates the successful modification of PINNs for accurate multi-parameter perfusion estimation from noisy and limited ASL data in infants. Frameworks like SUPINN have the potential to advance our understanding of the complex cardio-brain network physiology, aiding in the detection and management of diseases. Source code is provided at: https://github.com/cgalaz01/supinn.

PINNing Cerebral Blood Flow: Analysis of Perfusion MRI in Infants using Physics-Informed Neural Networks

TL;DR

A new spatial uncertainty-based physics-informed neural network (PINN), SUPINN, is proposed, to estimate CBF and other parameters from infant ASL data, demonstrating the successful modification of PINNs for accurate multi-parameter perfusion estimation from noisy and limited ASL data in infants.

Abstract

Arterial spin labeling (ASL) magnetic resonance imaging (MRI) enables cerebral perfusion measurement, which is crucial in detecting and managing neurological issues in infants born prematurely or after perinatal complications. However, cerebral blood flow (CBF) estimation in infants using ASL remains challenging due to the complex interplay of network physiology, involving dynamic interactions between cardiac output and cerebral perfusion, as well as issues with parameter uncertainty and data noise. We propose a new spatial uncertainty-based physics-informed neural network (PINN), SUPINN, to estimate CBF and other parameters from infant ASL data. SUPINN employs a multi-branch architecture to concurrently estimate regional and global model parameters across multiple voxels. It computes regional spatial uncertainties to weigh the signal. SUPINN can reliably estimate CBF (relative error ), bolus arrival time (AT) (), and blood longitudinal relaxation time () (), surpassing parameter estimates performed using least squares or standard PINNs. Furthermore, SUPINN produces physiologically plausible spatially smooth CBF and AT maps. Our study demonstrates the successful modification of PINNs for accurate multi-parameter perfusion estimation from noisy and limited ASL data in infants. Frameworks like SUPINN have the potential to advance our understanding of the complex cardio-brain network physiology, aiding in the detection and management of diseases. Source code is provided at: https://github.com/cgalaz01/supinn.

Paper Structure

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

Table of Contents

  1. Introduction
  2. Methods
  3. Dataset
  4. Mathematical Model for ASL
  5. Ground Truth Estimation
  6. Loss Function and Training Scheme
  7. PINN Architecture
  8. SUPINN Architecture
  9. Experimental Setup
  10. Results
  11. Discussion

Figures (3)

  • Figure 1: A representative 32-week postmenstrual case showing: A) $T_2$-weighted image highlighting the ASL imaging slice (orange); B) Subsampled perfusion-weighted image time series; and C) The measured perfusion signal of a single voxel over the entire duration, along with the corresponding ground-truth analytical model (see Eq. \ref{['eqn:ode']}).
  • Figure 2: Overview of our proposed SUPINN model, depicted here in a 2-branch variant for illustration purposes, but adaptable to larger configurations. This study employs a 3-branch model based on empirical findings.
  • Figure 3: Panel A shows spatial maps of parameter estimation in deep grey matter for a subject aged 32 weeks. Each row corresponds to the normalized relative error of two parameters: CBF and AT. The columns display the estimation results from four methods: SUPINN, PINN, LSF, and LSF-multi. Panel B depicts the parameter relative error of the models for a single voxel.