https://bonbid-hie2023.grand-challenge.org/
1st BOston Neonatal Brain Injury Dataset for Hypoxic Ischemic Encephalopathy
(BONBID-HIE)
Lesion Segmentation Challenge
Hypoxic ischemic encephalopathy (HIE) is a brain injury that occurs in 1 ~ 5/1000 term-born neonates. HIE affects around 200,000 term-born neonates every year worldwide, costing about $2 billion/year in the US alone, let alone family burdens. Although therapeutic hypothermia can reduce mortality and morbidity, yet around 60% of patients still die or develop neurocognitive deficits by 2 years of age. HIE lesion segmentation is a crucial step in clinical care of HIE. It will lead to a more accurate estimation of prognosis, a better understanding of neurological symptoms, and a timely prediction of response to therapy.
HIE-related brain abnormalities in brain MRI are often diffuse (i.e., multi-focal), and small (over half the patients having lesions occupying <1% of the brain volume). Segmentation for HIE MRI data is remarkably different from and arguably more challenging than other segmentation tasks, such as brain tumors with big and focal lesions. For example, the Dice overlap with U-Net and other state-of-the-art publications on this disease remains at around 0.5 so far, while dice is over 0.8 for brain tumors. Towards accurate early prognosis and medical diagnosis, there is an urgent but unmet need to improve the HIE lesion segmentation performance. The BONBID-HIE lesion segmentation challenge is the first challenge with public MRI data for this important brain abnormality.
Organizers
Rina Bao, PhD Yangming Ou, PhD P. Ellen Grant, MD
Postdoctoral Research Fellow Associate Professor Director of FNNDSC Boston Children's Hospital Boston Children's Hospital Boston Children's Hospital Harvard Medical School Harvard Medical School Harvard Medical School
Contact Info
Rina Bao: rina.bao@childrens.harvard.edu
Yangming Ou: yangming.ou@childrens.harvard.edu