Dataset Information

Using Deep Learning for the Classification of Images Generated by Multifocal Visual Evoked Potential.

ABSTRACT: Multifocal visual evoked potential (mfVEP) is used for assessing visual functions in patients with pituitary adenomas. Images generated by mfVEP facilitate evaluation of visual pathway integrity. However, lack of healthy controls, and high time consumption for analyzing data restrict the use of mfVEP in clinical settings; moreover, low signal-noise-ratio (SNR) in some images further increases the difficulty of analysis. I hypothesized that automated workflow with deep learning could facilitate analysis and correct classification of these images. A total of 9,120 images were used in this study. The automated workflow included clustering ideal and noisy images, denoising images using an autoencoder algorithm, and classifying normal and abnormal images using a convolutional neural network. The area under the receiver operating curve (AUC) of the initial algorithm (built on all the images) was 0.801 with an accuracy of 79.9%. The model built on denoised images had an AUC of 0.795 (95% CI: 0.773-0.817) and an accuracy of 78.6% (95% CI: 76.8-80.0%). The model built on ideal images had an AUC of 0.985 (95% CI: 0.976-0.994) and an accuracy of 94.6% (95% CI: 93.6-95.6%). The ensemble model achieved an AUC of 0.908 and an accuracy of 90.8% (sensitivity: 94.3%; specificity: 87.7%). The automated workflow for analyzing mfVEP plots achieved high AUC and accuracy, which suggests its possible clinical use.

SUBMITTER: Qiao N

PROVIDER: S-EPMC6085441 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Using Deep Learning for the Classification of Images Generated by Multifocal Visual Evoked Potential.

Qiao Nidan N

Frontiers in neurology 20180803

Multifocal visual evoked potential (mfVEP) is used for assessing visual functions in patients with pituitary adenomas. Images generated by mfVEP facilitate evaluation of visual pathway integrity. However, lack of healthy controls, and high time consumption for analyzing data restrict the use of mfVEP in clinical settings; moreover, low signal-noise-ratio (SNR) in some images further increases the difficulty of analysis. I hypothesized that automated workflow with deep learning could facilitate a ...[more]

PMID: 30123181

Similar Datasets

Project description:BackgroundMultifocal visual evoked potential (MF-VEP) assesses a wider visual field than full-field VEP (FF-VEP) and potentially offers a more precise analysis of optic nerve injury and repair following optic neuritis. MF-VEP may offer advantages over FF-VEP as an endpoint in clinical trials of remyelinating therapies.ObjectiveMF-VEP testing was used to study changes in visual pathways in 48% of RENEW [phase II, opicinumab (anti-LINGO-1; BIIB033) vs. placebo after first acute unilateral optic neuritis] participants.MethodsThis exploratory MF-VEP RENEW substudy compared mean outcomes at weeks 24 and 32 among participants in the intent-to-treat (ITT; n = 39; 72% female; mean age: 32.3 years) and per-protocol (PP; n = 31; 71% female; mean age: 32.2 years) populations in affected and fellow eye latency from fellow eye baseline latency and affected and fellow eye amplitude from their own baselines. Treatment differences were evaluated using analysis of covariance (week 24) and a mixed-effect model of repeated measures (week 32). Last observation carried forward was used to impute missing data at week 24.ResultsA trend for improvement in affected eye MF-VEP latency with opicinumab versus placebo was seen in the ITT and PP populations at weeks 24 and 32. Both treatment groups in the ITT population experienced partial recovery of amplitude in the affected eye at week 32. Notably, the mean change in fellow eye amplitude at weeks 24 and 32 was - 17.57 and - 31.41 nanovolts (nV) in placebo but only - 0.59 and 1.93 nV in the opicinumab group [differences at weeks 24 and 32: 16.98 nV (p = 0.050) and 33.33 nV (p < 0.01), respectively].ConclusionResults from this substudy showed advantages of MF-VEP over FF-VEP in multicenter studies of central nervous system reparative therapies and provide novel evidence that fellow eye visual pathway amplitude loss occurs after optic neuritis but can potentially be prevented by opicinumab treatment.RegistrationClinicalTrials.gov identifier NCT01721161.

Project description:A proper echocardiographic study requires several video clips recorded from different acquisition angles for observation of the complex cardiac anatomy. However, these video clips are not necessarily labeled in a database. Identification of the acquired view becomes the first step of analyzing an echocardiogram. Currently, there is no consensus whether the mislabeled samples can be used to create a feasible clinical prediction model of ejection fraction (EF). The aim of this study was to test two types of input methods for the classification of images, and to test the accuracy of the prediction model for EF in a learning database containing mislabeled images that were not checked by observers. We enrolled 340 patients with five standard views (long axis, short axis, 3-chamber view, 4-chamber view and 2-chamber view) and 10 images in a cycle, used for training a convolutional neural network to classify views (total 17,000 labeled images). All DICOM images were rigidly registered and rescaled into a reference image to fit the size of echocardiographic images. We employed 5-fold cross validation to examine model performance. We tested models trained by two types of data, averaged images and 10 selected images. Our best model (from 10 selected images) classified video views with 98.1% overall test accuracy in the independent cohort. In our view classification model, 1.9% of the images were mislabeled. To determine if this 98.1% accuracy was acceptable for creating the clinical prediction model using echocardiographic data, we tested the prediction model for EF using learning data with a 1.9% error rate. The accuracy of the prediction model for EF was warranted, even with training data containing 1.9% mislabeled images. The CNN algorithm can classify images into five standard views in a clinical setting. Our results suggest that this approach may provide a clinically feasible accuracy level of view classification for the analysis of echocardiographic data.

Dataset Information

Using Deep Learning for the Classification of Images Generated by Multifocal Visual Evoked Potential.

Publications

Using Deep Learning for the Classification of Images Generated by Multifocal Visual Evoked Potential.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure