A Deep Convolutional Neural Network for Annotation of Magnetic Resonance Imaging Sequence Type

Ranjbar, Sara; Singleton, Kyle; Jackson, Pamela; Rickertsen, Cassandra; Whitmire, Scott A.; Clark-Swanson, Kamala; Mitchell, J. Ross; Swanson, Kristin R.; Hu, Leland

doi:10.1007/s10278-019-00282-4

Cited by 27 publications

(21 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 1 provides a comparison between DNE, DRL, and prior Supervised Deep Learning using transfer learning approaches to sequence identification [12,13,8]. Although we have the smallest training set and CNN, DNE achieves 100% testing set accuracy.…”

Section: Resultsmentioning

confidence: 99%

“…MRI sequence identification is often helpful when attempting to acquire images of a particular sequence when natural language processing fails due to variable naming schemes. Prior supervised deep learning approaches have shown success in this task [12,13], reaching up to 99 − 100% testing set accuracy. However, they both relied on very large CNNs (Ima-geNet, GoogleNet, and VGGnet) that had been pre-trained on enormous image databases (ImageNet, 1.2 million images.)…”

Section: Prior Applications Of Deep Neuroevolution To Radiologymentioning

confidence: 99%

See 1 more Smart Citation

Deep Neuroevolution Squeezes More out of Small Neural Networks and Small Training Sets: Sample Application to MRI Brain Sequence Classification

Stember¹,

Shalu²

2021

Preprint

View full text Add to dashboard Cite

Purpose Deep Neuroevolution (DNE) holds the promise of providing radiology artificial intelligence (AI) that performs well with small neural networks and small training sets. We seek to realize this potential via a proof-of-principle application to MRI brain sequence classification.Materials and Methods We analyzed a training set of 20 patients, each with four sequences/weightings: T1, T1 post-contrast, T2, and T2-FLAIR. We trained the parameters of a relatively small convolutional neural network (CNN) as follows: First, we randomly mutated the CNN weights. We then measured the CNN training set accuracy, using the latter as the fitness evaluation metric. The fittest "child" CNNs were identified. We incorporated their mutations into the "parent" CNN. This selectively mutated parent became the next generation's parent CNN. We repeated this process for approximately 50,000 generations.Results DNE achieved monotonic convergence to 100% training set accuracy. DNE also converged monotonically to 100 % testing set accuracy.Conclusion DNE can achieve perfect accuracy with small training sets and small CNNs. Particularly when combined with Deep Reinforcement Learning, DNE may provide a path forward in the quest to make radiology AI more human-like in its ability to learn. DNE may very well turn out to be a key component of the much-anticipated "meta-learning" regime of radiology AI; algorithms that can adapt to new tasks and new image types, similar to human radiologists.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Prior Applications Of Deep Neuroevolution To Radiologymentioning

confidence: 99%

Deep Neuroevolution Squeezes More out of Small Neural Networks and Small Training Sets: Sample Application to MRI Brain Sequence Classification

Stember¹,

Shalu²

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Sara Ranjhbhar et al [12] demonstrated automatic annotation of MRI sequence types by using a deep CNN which can successfully detect patterns. Current works struggle with enhanced modalities and generate good results on non-enhanced T1W MRI scans.…”

Section: Amri Data and Processingmentioning

confidence: 99%

“…Also, it is advantageous in reducing computational costs.Various researchers have ne-tuned only a few layers of traditional pre-trained CNN architectures. Muhammad Sara Ranjbar et al[12] used a variation of VGG Net architecture instead of using full architecture for annotation of magnetic resonance imaging sequence type. Mariana Pereira et al[36] utilize 2 layers of Resnet34 architecture instead of using the full pre-trained model and proposed a prolonged 2D model for multiclass Alzheimer's Disease classi cation by achieving an accuracy of 68.6% and performs nicely compared to the state-of-the-art AD classi cation method.…”

mentioning

confidence: 99%

A Survey on Convolutional Neural Networks for MRI Analysis

Hardaha¹,

Reddy²,

Parne³

2021

Preprint

View full text Add to dashboard Cite

As of late, Convolutional Neural Networks have been very successful in segmentation and classification tasks. Magnetic resonance imaging (MRI) is a favored medical imaging method that comes up with interesting information for the diagnosis of different diseases.MR method is getting to be exceptionally well-known due to its non-invasive rule and for this reason, automated processing of this sort of image is getting noticed. MRI is effectively and widely used for tumor detection. Brain tumor detection is a popular medical application of MRI. Automating segmentation using CNN assists radiologists to lessen the high manual workload of tumor evaluation. CNN classification accuracy depends on network parameters and training data. CNN has the benefit of learning image features automatically directly out of multi-modal MRI images. In this survey paper, we have presented a summary of CNN's recent advancement in its technique applied on MRI images. The aim of this survey is to discuss various architectures and factors affecting the performance of CNN for learning features from different available MRI datasets. Based on the survey, section III (CNN for MRI analysis) comprises three subsections: A) MRI data and processing, B) CNN dimensionality, C) CNN architectures.

show abstract

“…Township health work site timely obtains and prints diagnostic reports to patients. For difficult cases, township doctors can upload the examination data to the central hospital for imaging consultation and generally complete the diagnosis or consultation within 1 hour [9].…”

Section: Researchmentioning

confidence: 99%

The Application of the Big Data Medical Imaging System in Improving the Medical and Health Examination

Zhou

Ouyang

2021

Journal of Healthcare Engineering

View full text Add to dashboard Cite

To explore the application effect of the big data medical imaging tertiary diagnostic system in improving the medical and health examination, cases in township health centers were collected by the medical imaging tertiary diagnosis system. Clinical cases examined by the tertiary diagnostic system of big data medical imaging will be set as the observation group. Clinical cases not involved in the tertiary diagnostic system of big data medical imaging were set as the control group. The qualified rate, film positive rate, and film diagnosis accuracy between the two groups are compared, and X-ray perspective, X-ray examination, and CT multiple medical imaging examinations are used in two groups. The experimental results showed that the pass rate was 86.57%, positive rate was 72.32%, and diagnosis rate was 80.17%. Pass rate, positive rate, and diagnostic accuracy were higher than the control group ( P < 0.05 ). X-line film is the most cost effective. CT examination has a high diagnostic sensitivity and can achieve a clear diagnosis of the benign and malignant diseases. The three-level diagnosis system of medical imaging has significantly improved and improved the technical level in the medical and health examination, which has good practical value.

show abstract

A Deep Convolutional Neural Network for Annotation of Magnetic Resonance Imaging Sequence Type

Cited by 27 publications

References 25 publications

Deep Neuroevolution Squeezes More out of Small Neural Networks and Small Training Sets: Sample Application to MRI Brain Sequence Classification

Deep Neuroevolution Squeezes More out of Small Neural Networks and Small Training Sets: Sample Application to MRI Brain Sequence Classification

A Survey on Convolutional Neural Networks for MRI Analysis

The Application of the Big Data Medical Imaging System in Improving the Medical and Health Examination

Contact Info

Product

Resources

About