Automating Carotid Intima-Media Thickness Video Interpretation with Convolutional Neural Networks

Shin, Jae Y.; Tajbakhsh, Nima; Hurst, R. Todd; Kendall, Christopher; Liang, Jianming

doi:10.1109/cvpr.2016.277

Cited by 47 publications

(29 citation statements)

References 29 publications

(43 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Applications of CNNs in medical image analysis are not limited to only computer-aided detection systems, however. CNNs have recently been used for carotid intima-media thickness measurement in ultrasound images [22], pancreas segmentation in CT images [23], brain tumor segmentation in magnetic resonance imaging (MRI) scans [24], multimodality isointense infant brain image segmentation [25], neuronal membrane segmentation in electron microscopy images [26], and knee cartilage segmentation in MRI scans [27].…”

Section: Related Workmentioning

confidence: 99%

Convolutional Neural Networks for Medical Image Analysis: Full Training or Fine Tuning?

Tajbakhsh

Shin

Gurudu

et al. 2016

IEEE Trans. Med. Imaging

Self Cite

2,385

1,325

View full text Add to dashboard Cite

Training a deep convolutional neural network (CNN) from scratch is difficult because it requires a large amount of labeled training data and a great deal of expertise to ensure proper convergence. A promising alternative is to fine-tune a CNN that has been pre-trained using, for instance, a large set of labeled natural images. However, the substantial differences between natural and medical images may advise against such knowledge transfer. In this paper, we seek to answer the following central question in the context of medical image analysis: Can the use of pre-trained deep CNNs with sufficient fine-tuning eliminate the need for training a deep CNN from scratch? To address this question, we considered four distinct medical imaging applications in three specialties (radiology, cardiology, and gastroenterology) involving classification, detection, and segmentation from three different imaging modalities, and investigated how the performance of deep CNNs trained from scratch compared with the pre-trained CNNs fine-tuned in a layer-wise manner. Our experiments consistently demonstrated that 1) the use of a pre-trained CNN with adequate fine-tuning outperformed or, in the worst case, performed as well as a CNN trained from scratch; 2) fine-tuned CNNs were more robust to the size of training sets than CNNs trained from scratch; 3) neither shallow tuning nor deep tuning was the optimal choice for a particular application; and 4) our layer-wise fine-tuning scheme could offer a practical way to reach the best performance for the application at hand based on the amount of available data.

show abstract

Section: Related Workmentioning

confidence: 99%

Convolutional Neural Networks for Medical Image Analysis: Full Training or Fine Tuning?

Tajbakhsh

Shin

Gurudu

et al. 2016

IEEE Trans. Med. Imaging

Self Cite

2,385

1,325

View full text Add to dashboard Cite

show abstract

“…The solution developed in this work is inspired by recent works reported in the area of deep learning, where CNNs are outperforming classical methods in many medical tasks [20]. The first attempt in using a CNN for the measurement of carotid IMT has been made only recently [21]. The exploitation of temporal redundancy on US sequences was shown to be a solution for improving overall detection results of the fetal heart [22], where a CNN coupled with a recurrent neural network (RNN) is used.…”

Section: Related Workmentioning

confidence: 99%

Real-time diameter of the fetal aorta from ultrasound

Savioli

Grisan

Visentin

et al. 2019

Neural Comput & Applic

View full text Add to dashboard Cite

The automatic analysis of ultrasound sequences can substantially improve the efficiency of clinical diagnosis. This article presents an attempt to automate the challenging task of measuring the vascular diameter of the fetal abdominal aorta from ultrasound images. We propose a neural network architecture consisting of three blocks: a convolutional neural network (CNN) for the extraction of imaging features, a convolution gated recurrent unit (C-GRU) for exploiting the temporal redundancy of the signal, and a regularized loss function, called CyclicLoss, to impose our prior knowledge about the periodicity of the observed signal. The solution is investigated with a cohort of 25 ultrasound sequences acquired during the third-trimester pregnancy check, and with 1000 synthetic sequences. In the extraction of features, it is shown that a shallow CNN outperforms two other deep CNNs with both the real and synthetic cohorts, suggesting that echocardiographic features are optimally captured by a reduced number of CNN layers. The proposed architecture, working with the shallow CNN, reaches an accuracy substantially superior to previously reported methods, providing an average reduction of the mean squared error from 0.31 (state-of-the-art) to 0.09 mm 2 , and a relative error reduction from 8.1 to 5.3%. The mean execution speed of the proposed approach of 289 frames per second makes it suitable for real-time clinical use. Nicoló Savioli and Enrico Grisan equally contributed to the work.

show abstract

“…There are techniques based on statistical modelling [11,14], Haugh transforms [15,22] and Nakagami distribution [33]. Recently, most of the techniques are based on deep learning methods [18,19,29,34]. Menchen-Lara et al has used ELM-AE with SLFN and estimated IMT, but the performance of the system is still below the active contour-based method proposed by the same group [22].…”

Section: Discussionmentioning

confidence: 99%

Arterial Parameters and Elasticity Estimation in Common Carotid Artery Using Deep Learning Approach

Patel¹,

Jain²

2019

IJIGSP

View full text Add to dashboard Cite

The risk of cardiovascular diseases is growing worldwide, and its early detection is necessary to reduce the level of risk. Structural parameters of the carotid artery as intima-media thickness and functional parameters such as arterial elasticity are directly associated with cardiovascular diseases. Segmentation of the carotid artery is required to measure the structural parameters and its temporal value that is used to estimate the arterial elasticity. This paper has two primary objectives: (i) Segmentation of the sequence of carotid artery ultrasound to measure temporal value of intimamedia thickness and lumen-diameter, and (ii) Young's modulus of elasticity estimation. The proposed segmentation method uses the contextual feature of the image pattern and is based on multi-layer extreme learning machine auto-encoder network. This segmentation method has two parts: (a) region of interest localization and (b) lumen-intima interface and mediaadventitia interface detection at the far wall. ROI localization algorithm divides the ultrasound frame into columns and also divides each column into overlapping blocks, ensuring that every column has a region of interest block. A multi-layer extreme learning machine with auto-encoder is trained with labelled data and in testing; system classifies the blocks into 'region of interest' and 'non-region of interest'. Pixels belonging to the region of interest are classified in the first part and a similar network-based method is proposed for lumenintima and media-adventitia interface detection at the near wall of the carotid artery. Structural parameter of the artery, intima-media thickness and lumen diameter are measured in a sequence of images of the cardiac cycle. The temporal values of structural parameters are used to estimate the young's modulus of elasticity. .

show abstract

Automating Carotid Intima-Media Thickness Video Interpretation with Convolutional Neural Networks

Cited by 47 publications

References 29 publications

Convolutional Neural Networks for Medical Image Analysis: Full Training or Fine Tuning?

Convolutional Neural Networks for Medical Image Analysis: Full Training or Fine Tuning?

Real-time diameter of the fetal aorta from ultrasound

Arterial Parameters and Elasticity Estimation in Common Carotid Artery Using Deep Learning Approach

Contact Info

Product

Resources

About