Dataset Information

Convolutional neural network-based segmentation can help in assessing the substantia nigra in neuromelanin MRI.

ABSTRACT:

Purpose

This study aimed to evaluate the accuracy and diagnostic test performance of the U-net-based segmentation method in neuromelanin magnetic resonance imaging (NM-MRI) compared to the established manual segmentation method for Parkinson's disease (PD) diagnosis.

Methods

NM-MRI datasets from two different 3T-scanners were used: a "principal dataset" with 122 participants and an "external validation dataset" with 24 participants, including 62 and 12 PD patients, respectively. Two radiologists performed SNpc manual segmentation. Inter-reader precision was determined using Dice coefficients. The U-net was trained with manual segmentation as ground truth and Dice coefficients used to measure accuracy. Training and validation steps were performed on the principal dataset using a 4-fold cross-validation method. We tested the U-net on the external validation dataset. SNpc hyperintense areas were estimated from U-net and manual segmentation masks, replicating a previously validated thresholding method, and their diagnostic test performances for PD determined.

Results

For SNpc segmentation, U-net accuracy was comparable to inter-reader precision in the principal dataset (Dice coefficient: U-net, 0.83?±?0.04; inter-reader, 0.83?±?0.04), but lower in external validation dataset (Dice coefficient: U-net, 079?±?0.04; inter-reader, 0.85?±?0.03). Diagnostic test performances for PD were comparable between U-net and manual segmentation methods in both principal (area under the receiver operating characteristic curve: U-net, 0.950; manual, 0.948) and external (U-net, 0.944; manual, 0.931) datasets.

Conclusion

U-net segmentation provided relatively high accuracy in the evaluation of the SNpc in NM-MRI and yielded diagnostic performance comparable to that of the established manual method.

SUBMITTER: Le Berre A

PROVIDER: S-EPMC6848644 | biostudies-literature | 2019 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Convolutional neural network-based segmentation can help in assessing the substantia nigra in neuromelanin MRI.

Le Berre Alice A Kamagata Koji K Otsuka Yujiro Y Andica Christina C Hatano Taku T Saccenti Laetitia L Ogawa Takashi T Takeshige-Amano Haruka H Wada Akihiko A Suzuki Michimasa M Hagiwara Akifumi A Irie Ryusuke R Hori Masaaki M Oyama Genko G Shimo Yashushi Y Umemura Atsushi A Hattori Nobutaka N Aoki Shigeki S

Neuroradiology 20190810 12

<h4>Purpose</h4>This study aimed to evaluate the accuracy and diagnostic test performance of the U-net-based segmentation method in neuromelanin magnetic resonance imaging (NM-MRI) compared to the established manual segmentation method for Parkinson's disease (PD) diagnosis.<h4>Methods</h4>NM-MRI datasets from two different 3T-scanners were used: a "principal dataset" with 122 participants and an "external validation dataset" with 24 participants, including 62 and 12 PD patients, respectively. T ...[more]

PMID: 31401723

Similar Datasets

Project description:BackgroundNeuromelanin is a pigment with strong iron-chelating properties preferentially found in dopaminergic neurons of the substantia nigra pars compacta (SNpc). Parkinson's disease is characterized by pronounced, MRI-detectable neuromelanin loss, but the neuroprotective or neurotoxic role of neuromelanin remains debated. Histological studies have demonstrated neuromelanin increases with age, but this has not been confirmed in vivo, and there is uncertainty whether neuromelanin declines, stabilizes, or increases from middle age.MethodsThis study aimed to establish physiological changes of pigmentation of the SNpc using a pooled data set of neuromelanin-sensitive 3T MRI from 134 healthy individuals aged 5-83 years. Neuromelanin-related brightness (regional contrast to ratio) and calibrated hyperintense volumes were analyzed using linear and nonlinear regression models to characterize age effects. Laterality, sex, and subregional effects were also assessed.ResultsFor brightness, age effects were best described as a quadratic trajectory explaining 81.5% of the observed variance in the SNpc showing a strong increase from childhood to adolescence, with plateauing in middle age and a decline in older age. Similar but less pronounced effects were seen in hyperintense volumes. We also show an anterior-posterior gradient in SNpc contrast, larger normalized neuromelanin-rich volume in women > 47 years old, but no laterality effect.ConclusionsUsing optimized neuromelanin MRI in a life span sample, we demonstrate a strong age effect with inverted U-shaped SNpc pigmentation-related contrast from childhood to old age. This age trajectory of physiological SNpc pigmentation needs to be taken into account for diagnostic applications of depigmentation. The study also paves the way for systematic investigations of the mechanisms of neuromelanin in healthy and pathological brain development and aging. © 2018 International Parkinson and Movement Disorder Society.

Dataset Information

Convolutional neural network-based segmentation can help in assessing the substantia nigra in neuromelanin MRI.

Purpose

Methods

Results

Conclusion

Publications

Convolutional neural network-based segmentation can help in assessing the substantia nigra in neuromelanin MRI.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets