BOston Neonatal Brain Injury Data for Hypoxic Ischemic Encephalopathy (BONBID-HIE): I. MRI and Lesion Labeling.

in Scientific data by Rina Bao, Ya'nan Song, Sara V Bates, Rebecca J Weiss, Anna N Foster, Camilo Jaimes, Susan Sotardi, Yue Zhang, Randy L Hirschtick, P Ellen Grant, Yangming Ou

TLDR

  • A new public dataset of neonatal brain diffusion MRI and expert-annotated HIE-related lesions from 133 patients with HIE is now available for developing MRI lesion segmentation methods.
  • The dataset aims to improve understanding and treatment of HIE-related brain injuries by providing a valuable resource for developing more accurate segmentation techniques.
  • Future studies can utilize this dataset to investigate various aspects of HIE-related brain injuries and improve patient care.

Abstract

Hypoxic ischemic encephalopathy (HIE) is a brain injury that occurs in 1 ~ 5/1000 term neonates. Accurate identification and segmentation of HIE-related lesions in neonatal brain magnetic resonance images (MRIs) is the first step toward identifying high-risk patients, understanding neurological symptoms, evaluating treatment effects, and predicting outcomes. We release the first public dataset containing neonatal brain diffusion MRI and expert annotation of lesions from 133 patients diagnosed with HIE. HIE-related lesions in brain MRI are often diffuse (i.e., multi-focal), and small (over half the patients in our data having lesions occupying <1% of the brain volume (including ventricles)). Segmentation for HIE MRI data is remarkably different from, and arguably more challenging than, other segmentation tasks such as brain tumors with focal and relatively large lesions. We hope that this dataset can help fuel the development of MRI lesion segmentation methods for HIE and small diffuse lesions in general.

Overview

  • The study focused on creating a public dataset of neonatal brain diffusion MRI and expert-annotated HIE-related lesions from 133 patients with Hypoxic Ischemic Encephalopathy (HIE).
  • The primary objective was to provide a valuable resource for developing MRI lesion segmentation methods for HIE and small diffuse lesions.
  • The study aimed to bridge the gap in understanding HIE-related brain injuries by providing a dataset for developing more accurate segmentation techniques.

Comparative Analysis & Findings

  • The dataset contains neonatal brain diffusion MRI scans and expert-annotated HIE-related lesions from 133 patients diagnosed with HIE.
  • The lesions are often diffuse (multi-focal) and small, with over half of the patients having lesions occupying less than 1% of the brain volume (including ventricles).
  • The segmentation of HIE-related lesions in brain MRI is challenging due to the diffuse and small nature of the lesions, making it distinct from other segmentation tasks like brain tumors with focal and relatively large lesions.

Implications and Future Directions

  • The dataset has the potential to improve MRI lesion segmentation methods for HIE and small diffuse lesions, which can ultimately lead to better understanding and treatment of HIE-related brain injuries.
  • Future studies can utilize this dataset to develop and evaluate novel segmentation techniques and algorithms, leading to more accurate diagnosis and prognosis of HIE-related brain injuries.
  • The dataset can also be used to investigate the relationship between HIE-related lesions and neurological symptoms, treatment effects, and outcomes, ultimately informing clinical practice and improving patient care.
Read Full Article