Using deep Siamese neural networks for detection of brain asymmetries associated with Alzheimer's Disease and Mild Cognitive Impairment

Liu, Chin-Fu; Padhy, Shreyas; Ramachandran, Sandhya; Wang, Victor X.; Efimov, A.Yu.; Bernal, Alonso; Shi, Linyuan; Vaillant, Marc; Ratnanather, J. Tilak; Faria, Andréia V.; Caffo, Brian; Albert, Marilyn; Miller, Michael I.; Initiative, Alzheimer’s Disease Neuroimaging

doi:10.1016/j.mri.2019.07.003

Cited by 68 publications

(29 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This strategy, of including both single blip data, as well as two-blip data, let's the networks learn from distortions in a number of directions. This network architecture mirrors the Siamese [21] and null space [22] network designs.…”

Section: Plos Onementioning

confidence: 97%

Distortion correction of diffusion weighted MRI without reverse phase-encoding scans or field-maps

et al. 2020

View full text Add to dashboard Cite

Diffusion magnetic resonance images may suffer from geometric distortions due to susceptibility induced off resonance fields, which cause geometric mismatch with anatomical images and ultimately affect subsequent quantification of microstructural or connectivity indices. State-of-the art diffusion distortion correction methods typically require data acquired with reverse phase encoding directions, resulting in varying magnitudes and orientations of distortion, which allow estimation of an undistorted volume. Alternatively, additional field maps acquisitions can be used along with sequence information to determine warping fields. However, not all imaging protocols include these additional scans and cannot take advantage of state-of-the art distortion correction. To avoid additional acquisitions, structural MRI (undistorted scans) can be used as registration targets for intensity driven correction. In this study, we aim to (1) enable susceptibility distortion correction with historical and/or limited diffusion datasets that do not include specific sequences for distortion correction and (2) avoid the computationally intensive registration procedure typically required for distortion correction using structural scans. To achieve these aims, we use deep learning (3D U-nets) to synthesize an undistorted b0 image that matches geometry of structural T1w images and intensity contrasts from diffusion images. Importantly, the training dataset is heterogenous, consisting of varying acquisitions of both structural and diffusion. We apply our approach to a withheld test set and show that distortions are successfully corrected after processing. We quantitatively evaluate the proposed distortion correction and intensitybased registration against state-of-the-art distortion correction (FSL topup). The results illustrate that the proposed pipeline results in b0 images that are geometrically similar to non-PLOS ONE

show abstract

Section: Plos Onementioning

confidence: 97%

Distortion correction of diffusion weighted MRI without reverse phase-encoding scans or field-maps

et al. 2020

View full text Add to dashboard Cite

show abstract

“…The model achieves an accuracy of 55.34% for both binary and multiclassification using the ADNI dataset. Chin-Fu Liu et al [21] proposed Siamese neural networks to investigate the discriminative capacity of whole-brain volumetric asymmetry. The team used the MRI Cloud process to create low-dimensional volumetric features for pre-defined atlas brain structures, as well as a unique non-linear kernel method to normalize features and eliminate batch effects across datasets and populations.…”

Section: Related Workmentioning

confidence: 99%

DEMNET: A Deep Learning Model for Early Diagnosis of Alzheimer Diseases and Dementia From MR Images

et al. 2021

View full text Add to dashboard Cite

Alzheimer's Disease (AD) is the most common cause of dementia globally. It steadily worsens from mild to severe, impairing one's ability to complete any work without assistance. It begins to outstrip due to the population ages and diagnosis timeline. For classify cases, existing approaches incorporate medical history, neuropsychological testing, and Magnetic Resonance Imaging (MRI), but efficient procedures remain inconstant due to lack of sensitivity and precision. The Convolutional Neural Network (CNN) utilized to create a framework that can exploit to detect specific Alzheimer's disease characteristics from MRI images. By considering four stages of dementia and conducting a particular diagnosis, the proposed model generates high-resolution disease probability maps from the local brain structure to a multilayer perceptron and provides accurate, intuitive visualizations of individual Alzheimer's disease risk. To avoid the problem of class imbalance, the samples should be evenly distributed among the classes. The obtained MRI image dataset from Kaggle has a major class imbalance problem. A DEMentia NETwork (DEMNET) a CNN model is proposed to detect the dementia stages, which is the primary cause of AD. The DEMNET achieves an accuracy of 95.23%, Area Under Curve (AUC) of 97% and Cohen's Kappa value of 0.93 from the Kaggle dataset, which is superior to existing methods. We also used the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset to predict AD classes in order to assess the efficacy of the proposed model.

show abstract

“…Machine and deep learning are increasingly used in numerous fields. Medical and health applications are among those fields where machine learning and deep learning are used to diagnose, detect, and early predict diseases like Alzheimer's [1], cardiovascular disease [2], cancer [3], and Parkinson's disease [4].…”

Section: Introductionmentioning

confidence: 99%

“…Neural networks, especially deep neural networks, have high classification accuracy; however, these models fail when the number of samples used for training is small. Siamese neural network [1,4] is one type of neural network model that works well under this limitation. Siamese neural network was first presented by [4] for signature verification, and this work was later extended for text similarity [8], face recognition [9,10], video object tracking [11], and other image classification work [1,12].…”

Section: Introductionmentioning

confidence: 99%

“…Siamese neural network [1,4] is one type of neural network model that works well under this limitation. Siamese neural network was first presented by [4] for signature verification, and this work was later extended for text similarity [8], face recognition [9,10], video object tracking [11], and other image classification work [1,12]. is work is motivated by the importance of Parkinson's disease (PD), which is a neurodegenerative disease (NDD) that has motor-related symptoms, like tremor and instability [13].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Data Clustering Improves Siamese Neural Networks Classification of Parkinson’s Disease

2021

View full text Add to dashboard Cite

Parkinson’s disease (PD) is a clinical neurodegenerative disease having symptoms like tremor, rigidity, and postural disability. According to Harvard, about 60,000 of American citizens are diagnosed with PD yearly, with more than 10 million people infected worldwide. An estimate of 4% of the people have PD before they reach the age 50; however, the incident increases with age. Diagnosis of PD relies on the expertise of the physician and depends on several established clinical criteria. This makes the diagnosis subjective and inefficient. Hence, continuous efforts are being made to enhance the diagnosis of PD using deep learning approaches that rely on experienced neurologists. Siamese neural networks mainly work on two different input vectors and are used in comparison of output vectors. Moreover, clustering a dataset before applying classification enhances the distribution of similar samples among groups. In addition, applying the Siamese network can overcome the limitation of samples per class in the dataset by guiding the network to learn differences between samples rather than focusing on learning specific classes. In this paper, a Siamese neural network is applied to diagnose PD. Siamese networks predict the sample class by estimating how similar a sample is to other samples. The idea behind this work is clustering the dataset before training the network, as different pairs that belong to the same cluster are candidates to be mistaken by the network and assumed to be matched pairs. To overcome this problem, the dataset is first clustered, and then the architecture feeds the network to pairs of the same cluster. The proposed framework is concerned with comparing the performance when using clustered against unclustered data. The proposed framework outperforms the conventional framework without clustering. The accuracy achieved for classifying unclustered PD patients reached 76.75%, while it reached 84.02% for clustered data, outperforming the same technique on unclustered data. The significance of this study is in the enhanced performance achieved due to the clustering of data, which shows a promising framework to enhance the diagnostic capability of computer-aided disease diagnostic tools.

show abstract

Using deep Siamese neural networks for detection of brain asymmetries associated with Alzheimer's Disease and Mild Cognitive Impairment

Cited by 68 publications

References 47 publications

Distortion correction of diffusion weighted MRI without reverse phase-encoding scans or field-maps

Distortion correction of diffusion weighted MRI without reverse phase-encoding scans or field-maps

DEMNET: A Deep Learning Model for Early Diagnosis of Alzheimer Diseases and Dementia From MR Images

Data Clustering Improves Siamese Neural Networks Classification of Parkinson’s Disease

Contact Info

Product

Resources

About