RESUMO
INTRODUCTION: The Locus Coeruleus (LC) is linked to the development and pathophysiology of neurodegenerative diseases such as Alzheimer's Disease (AD). Magnetic Resonance Imaging based LC features have shown potential to assess LC integrity in vivo. METHODS: We present a Deep Learning based LC segmentation and feature extraction method: ELSI-Net and apply it to healthy aging and AD dementia datasets. Agreement to expert raters and previously published LC atlases were assessed. We aimed to reproduce previously reported differences in LC integrity in aging and AD dementia and correlate extracted features to cerebrospinal fluid (CSF) biomarkers of AD pathology. RESULTS: ELSI-Net demonstrated high agreement to expert raters and published atlases. Previously reported group differences in LC integrity were detected and correlations to CSF biomarkers were found. DISCUSSION: Although we found excellent performance, further evaluations on more diverse datasets from clinical cohorts are required for a conclusive assessment of ELSI-Nets general applicability.
RESUMO
Expert interpretation of anatomical images of the human brain is the central part of neuroradiology. Several machine learning-based techniques have been proposed to assist in the analysis process. However, the ML models typically need to be trained to perform a specific task, e.g., brain tumour segmentation or classification. Not only do the corresponding training data require laborious manual annotations, but a wide variety of abnormalities can be present in a human brain MRI - even more than one simultaneously, which renders a representation of all possible anomalies very challenging. Hence, a possible solution is an unsupervised anomaly detection (UAD) system that can learn a data distribution from an unlabelled dataset of healthy subjects and then be applied to detect out-of-distribution samples. Such a technique can then be used to detect anomalies - lesions or abnormalities, for example, brain tumours, without explicitly training the model for that specific pathology. Several Variational Autoencoder (VAE) based techniques have been proposed in the past for this task. Even though they perform very well on controlled artificially simulated anomalies, many of them perform poorly while detecting anomalies in clinical data. This research proposes a compact version of the "context-encoding" VAE (ceVAE) model, combined with pre and post-processing steps, creating a UAD pipeline (StRegA), which is more robust on clinical data and shows its applicability in detecting anomalies such as tumours in brain MRIs. The proposed pipeline achieved a Dice score of 0.642 ± 0.101 while detecting tumours in T2w images of the BraTS dataset and 0.859 ± 0.112 while detecting artificially induced anomalies, while the best performing baseline achieved 0.522 ± 0.135 and 0.783 ± 0.111, respectively.
Assuntos
Neoplasias Encefálicas , Processamento de Imagem Assistida por Computador , Encéfalo/diagnóstico por imagem , Encéfalo/patologia , Neoplasias Encefálicas/diagnóstico por imagem , Neoplasias Encefálicas/patologia , Humanos , Processamento de Imagem Assistida por Computador/métodos , Imageamento por Ressonância Magnética/métodos , NeuroimagemRESUMO
PURPOSE: Development and performance measurement of a fully automated pipeline that localizes and segments the locus coeruleus in so-called neuromelanin-sensitive magnetic resonance imaging data for the derivation of quantitative biomarkers of neurodegenerative diseases such as Alzheimer's disease and Parkinson's disease. METHODS: We propose a pipeline composed of several 3D-Unet-based convolutional neural networks for iterative multi-scale localization and multi-rater segmentation and non-deep learning-based components for automated biomarker extraction. We trained on the healthy aging cohort and did not carry out any adaption or fine-tuning prior to the application to Parkinson's disease subjects. RESULTS: The localization and segmentation pipeline demonstrated sufficient performance as measured by Euclidean distance (on average around 1.3mm on healthy aging subjects and 2.2mm in Parkinson's disease subjects) and Dice similarity coefficient (overall around [Formula: see text] on healthy aging subjects and [Formula: see text] for subjects with Parkinson's disease) as well as promising agreement with respect to contrast ratios in terms of intraclass correlation coefficient of [Formula: see text] for healthy aging subjects compared to a manual segmentation procedure. Lower values ([Formula: see text]) for Parkinson's disease subjects indicate the need for further investigation and tests before the application to clinical samples. CONCLUSION: These promising results suggest the usability of the proposed algorithm for data of healthy aging subjects and pave the way for further investigations using this approach on different clinical datasets to validate its practical usability more conclusively.