In the presence of strong reflecting surfaces, the detector in SS-OCT may saturate, leading to loss of information within affected A-scans and potentially disturbing axial artifacts in affected B-scans or volumes. In this work, we trained an image-based neural network to detect and remove such artifacts and restore the underlying structure by means of image inpainting. For this purpose, sets of paired images were generated from raw OCT spectra, with one image intact and the other suffering from simulated detector saturation. We demonstrate the effectiveness of the proposed method qualitatively and quantitatively.
Optical coherence tomography can provide visualizations of the eye both in diagnostic and surgical settings. However, noise limits the achievable image quality, especially in scenarios in which multi-frame averaging is not available. In this work, we present high-quality OCT image denoising using deep learning, only requiring unpaired volumetric capture scans for training. It is shown that, by exploiting neighboring B-scans, an artificial neural network for denoising OCT images can be trained based on a state-of-the-art approach which usually requires repeated scans from the exact same location. The effect of denoising is demonstrated for B-scans and volumetric renderings during and after mock cataract surgery on ex-vivo porcine eyes.
Diabetic retinopathy is the leading cause of vision loss. Optical coherence tomography-angiography is emerging as the potentially most promising technique for diagnosing DR. We circumvent the necessity for strong labels in these data with our multiple instance learning (MIL)-based network, MIL-ResNet14. MIL-ResNet14 is evaluated against two other proven capable classifiers, Res-Net14 and VGG16. All networks were assessed quantitatively by numerical classification values. MIL-ResNet14 showcased superior numerical classification abilities and turned to identify lesions more reliably. We conclude that MIL has a regularizing effect on inexactly labeled data and is a more reliable classifier than previously proposed methods.
Previously introduced deep learning classifiers were able to support diabetic biomarker detection in OCTA en face images, but require pixel-by-pixel expert labeling, which is a labor-intensive and expensive process. We present a multiple-instance learning-based network, MIL-ResNet,14 that detects clinically relevant diabetic retinopathy biomarkers in a wide-angle (65°) OCTA dataset with high accuracy without annotation. We evaluated our proposed architecture against two well-established machine learning classifiers, ResNet14 and VGG16. The dataset we used for this study was acquired with a MHz A-scan rate swept source OCT device. We used a total of 352 en face images representing the retinal vasculature over an 18 mm x 18 mm field of view. MIL-ResNet14 outperformed the other two networks with an F-score of 0.95, a precision of 0.909 and an area under the curve of 0.973. In addition, we were able to demonstrate that MIL-ResNet14 paid special attention to relevant biomarkers such as ischemic areas and retinal vascular abnormalities by saliency overlay of gradient-weighted class activation maps on top of the en face images. Thus, OCTA could be used as a powerful diagnostic decision support tool for clinical ophthalmic screening in combination with our MIL approach.
As in other imaging modalities, noise decreases image quality in optical coherence tomography (OCT), which is especially problematic in real-time intra-surgical application, where multi-frame averaging is not available. In this work, we present an adapted self-supervised training approach to train a blind-spot denoising network for OCT data. With the proposed method, the stability of the method is improved, avoiding the occurrence of artifacts by increasing realism of training data. We show that using this approach, the quality of two-dimensional B-scans can be improved qualitatively and quantitatively even without paired training data. This improvement is also translated into live volumetric renderings composed of denoised two-dimensional scans, even when using only very small network complexities due to harsh time constraints.
In this work, we propose to utilize an end-to-end deep learning approach for the reconstruction of structural OCT images based on the rich information contained in raw OCT data alone instead of performing signal processing with manual tuning of the associated system parameters.
The proposed deep learning approach already yields promising results on a small training data set of widefield OCT images. Qualitative results suggest that the neural network is able to implicitly learn the full signal processing pipeline and its inherent system parameters but is strongly impacted by the data variability seen during training.
We present a multiple instance learning-based network, MIL-ResNet14, detecting biomarkers for diabetic retinopathy in a widefield optical coherence tomography angiography dataset with high accuracy, without the necessity of annotations other than the information of whether a scan stems from a diabetic patient or not. Previously introduced deep learning-based classifiers were able to support the detection of diabetic biomarkers in OCTA images, however, require expert labeling on a pixel-level, a labor-intensive and expensive process. We evaluated our proposed architecture against two proven-capable classifiers, ResNet14 and VGG16. The dataset we applied for this study was acquired with a MHz A-Scan rate widefield Swept Source-OCT device. We utilized a total of 352 en face images, displaying retinal vasculature over a field of view of 18 mm x 18 mm. MIL-ResNet14 outperformed both other networks with an F-score of 0.95, a precision of 0.909 and an Area Under the Curve of 0.973. In addition, we could show via saliency overlays of gradient-weighted class activation mappings onto the en face images, that MIL-ResNet14 pays special attention to clinically relevant biomarkers like ischemic areas and retinal vessel anomalies. This could therefore function as a vigorous diagnostic decision support tool for clinical ophthalmologic screenings.
Noise decreases image quality in optical coherence tomography (OCT) and can obscure important features in real-time visualizations. In this work, we show that a neural network can be applied to denoise volumetric OCT data for intra-surgical visualization in real-time. We adapt a self-supervised training approach, not requiring any paired data for training. Several optimizations and trade-offs in deployment are required, with which we achieved processing times of only few milliseconds. While still being limited by the real-time requirements, denoising in this scenario can enhance surface visibility, and therefore allow guidance for more precise intra-surgical maneuvers.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.