Transfer learning in deep neural network based under-sampled MR image reconstruction

Arshad, Madiha; Qureshi, Mahmood; Inam, Omair; Omer, Hammad

doi:10.1016/j.mri.2020.09.018

Cited by 15 publications

(13 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On account of that, rather than the hand-crafted feature extraction approaches performed in prior studies, the application of deep learning-based, particularly transfer learning-based methods in eye state detection, has emerged as a promising capability in terms of both speed and utility [10,26]. Transfer learning is an approach to dealing with modest changes between datasets by applying the knowledge learnt by a neural network from one task to another independent learning assignment [30]. In countless fields, such as medical image analysis, transfer learning is favored when there is inadequate data in the datasets during the learning process [30,31].…”

Section: Cnnmentioning

confidence: 99%

“…Transfer learning is an approach to dealing with modest changes between datasets by applying the knowledge learnt by a neural network from one task to another independent learning assignment [30]. In countless fields, such as medical image analysis, transfer learning is favored when there is inadequate data in the datasets during the learning process [30,31]. ImageNet [32], a large image database dedicated for use in visual object recognition software research, hosted a competition in 2012, and since then, the outstanding success of CNN methods has substantially extended its application in the field of computer vision.…”

Section: Cnnmentioning

confidence: 99%

See 1 more Smart Citation

An Eye State Recognition System Using Transfer Learning: AlexNet-Based Deep Convolutional Neural Network

Kayadibi

Güraksın

Ergün

et al. 2022

Int J Comput Intell Syst

View full text Add to dashboard Cite

For eye state recognition (closed or open), a mechanism based on deep convolutional neural network (DCNN) using the Zhejiang University (ZJU) and Closed Eyes in the Wild (CEW) dataset, has been proposed in this paper. In instances where blinking is consequential, eye state recognition plays a critical part for the development of human–machine interaction (HMI) solutions. To accomplish this objective, pre-trained CNN architectures on ImageNet were first trained on the both dataset, which included both open and closed-eye states, and then they were tested, and their performance was quantified. The AlexNet design has proven to be more successful owing to these assessments. The ZJU and CEW datasets were leveraged to train the DCNN architecture, which was constructed employing AlexNet modifications for performance enhancement. On the both datasets, the suggested DCNN architecture was tested for performance. The achieved DCNN design was found to have 97.32% accuracy, 95.37% sensitivity, 97.97% specificity, 93.99% precision, 94.67% F1 score, and 99.37% AUC values in the ZJU dataset, while it was found to have 97.93% accuracy, 98.74% sensitivity, 97.15% specificity, 97.11% precision, 97.92% F1 score, and 99.69% AUC values in the CEW dataset. Accordingly, when compared to CNN architectures, it scored the maximum performance. At the same time, the DCNN architecture proposed on the ZJU and CEW datasets has been confirmed to be an acceptable and productive solution for eye state recognition depending on the outcomes compared to the studies in the literature. This method may contribute to the development of HMI systems by adding to the literature on eye state recognition.

show abstract

Section: Cnnmentioning

confidence: 99%

Section: Cnnmentioning

confidence: 99%

An Eye State Recognition System Using Transfer Learning: AlexNet-Based Deep Convolutional Neural Network

Kayadibi

Güraksın

Ergün

et al. 2022

Int J Comput Intell Syst

View full text Add to dashboard Cite

show abstract

“…Operationally, challenges exist with dealing with a heterogenous fleet of scanners that may be at different stages in their life cycle. Image quality may vary depending on brand, capabilities and whether the scanner is 1.5 T versus 3 T. Arshad et al, demonstrated that these problems can be addressed with Transfer Learning techniques that showed improved generalizability for: images acquired from scanners with different magnetic field strengths, MR images of different anatomies, and MR images under-sampled by different acceleration factors [7] , [8] . This can be particularly valuable in oncologic patients where comparison to prior imaging and consistent protocols are important in restaging and post-treatment scans.…”

Section: Imaging Acquisition Optimizationmentioning

confidence: 99%

Artificial intelligence in oncologic imaging

Chen

Terzić²,

Becker

et al. 2022

European Journal of Radiology Open

View full text Add to dashboard Cite

“…More recently deep learning-based methods such as convolutional neural networks (CNNs) and generative adversarial networks (GAN) have shown promising results to accelerate the MR imaging data acquisition process [15]. These methods apply deep learning-based reconstruction schemes to create highquality images from undersampled MR data [9,[16][17][18][19][20][21].…”

Section: Introductionmentioning

confidence: 99%

Deep learning–based acceleration of Compressed Sense MR imaging of the ankle

et al. 2022

View full text Add to dashboard Cite

Objectives To evaluate a compressed sensing artificial intelligence framework (CSAI) to accelerate MRI acquisition of the ankle. Methods Thirty patients were scanned at 3T. Axial T2-w, coronal T1-w, and coronal/sagittal intermediate-w scans with fat saturation were acquired using compressed sensing only (12:44 min, CS), CSAI with an acceleration factor of 4.6–5.3 (6:45 min, CSAI2x), and CSAI with an acceleration factor of 6.9–7.7 (4:46 min, CSAI3x). Moreover, a high-resolution axial T2-w scan was obtained using CSAI with a similar scan duration compared to CS. Depiction and presence of abnormalities were graded. Signal-to-noise and contrast-to-noise were calculated. Wilcoxon signed-rank test and Cohen’s kappa were used to compare CSAI with CS sequences. Results The correlation was perfect between CS and CSAI2x (κ = 1.0) and excellent for CS and CSAI3x (κ = 0.86–1.0). No significant differences were found for the depiction of structures between CS and CSAI2x and the same abnormalities were detected in both protocols. For CSAI3x the depiction was graded lower (p ≤ 0.001), though most abnormalities were also detected. For CSAI2x contrast-to-noise fluid/muscle was higher compared to CS (p ≤ 0.05), while no differences were found for other tissues. Signal-to-noise and contrast-to-noise were higher for CSAI3x compared to CS (p ≤ 0.05). The high - resolution axial T2-w sequence specifically improved the depiction of tendons and the tibial nerve (p ≤ 0.005). Conclusions Acquisition times can be reduced by 47% using CSAI compared to CS without decreasing diagnostic image quality. Reducing acquisition times by 63% is feasible but should be reserved for specific patients. The depiction of specific structures is improved using a high-resolution axial T2-w CSAI scan. Key Points • Prospective study showed that CSAI enables reduction in acquisition times by 47% without decreasing diagnostic image quality. • Reducing acquisition times by 63% still produces images with an acceptable diagnostic accuracy but should be reserved for specific patients. • CSAI may be implemented to scan at a higher resolution compared to standard CS images without increasing acquisition times.

show abstract

Transfer learning in deep neural network based under-sampled MR image reconstruction

Cited by 15 publications

References 16 publications

An Eye State Recognition System Using Transfer Learning: AlexNet-Based Deep Convolutional Neural Network

An Eye State Recognition System Using Transfer Learning: AlexNet-Based Deep Convolutional Neural Network

Artificial intelligence in oncologic imaging

Deep learning–based acceleration of Compressed Sense MR imaging of the ankle

Contact Info

Product

Resources

About