Deep learning based noise reduction method for automatic 3D segmentation of the anterior of lamina cribrosa in optical coherence tomography volumetric scans

Abstract

A deep-learning (DL) based noise reduction algorithm, in combination with a vessel shadow compensation method and a three-dimensional (3D) segmentation technique, has been developed to achieve, to the authors best knowledge, the first automatic segmentation of the anterior surface of the lamina cribrosa (LC) in volumetric ophthalmic optical coherence tomography (OCT) scans. The present DL-based OCT noise reduction algorithm was trained without the need of noise-free ground truth images by utilizing the latest development in deep learning of de-noising from single noisy images, and was demonstrated to be able to cover more locations in the retina and disease cases of different types to achieve high robustness. Compared with the original single OCT images, a 6.6 dB improvement in peak signal-to-noise ratio and a 0.65 improvement in the structural similarity index were achieved. The vessel shadow compensation method analyzes the energy profile in each A-line and automatically compensates the pixel intensity of locations underneath the detected blood vessel. Combining the noise reduction algorithm and the shadow compensation and contrast enhancement technique, medical experts were able to identify the anterior surface of the LC in 98.3% of the OCT images. The 3D segmentation algorithm employs a two-round procedure based on gradients information and information from neighboring images. An accuracy of 90.6% was achieved in a validation study involving 180 individual B-scans from 36 subjects, compared to 64.4% in raw images. This imaging and analysis strategy enables the first automatic complete view of the anterior LC surface, to the authors best knowledge, which may have the potentials in new LC parameters development for glaucoma diagnosis and management.

Quantitative and Qualitative Evaluation of Supercontinuum Laser‐Induced Cutaneous Thermal Injuries and Their Repair With OCT Images

“…We will find the damage threshold (ED 50 ) [14] of a supercontinuum laser for mouse skin. The injuries in the OAC map were manually delineated; therefore, an automatic segmentation and intelligent algorithm needs to be developed [43]. Furthermore, the current depth information and 3D visualization of injured spots should be acquired through segmenting the injured areas, and the 3D injured areas should be segmented in combination with a deep learning algorithm [44], enabling a stereoimaging quantitative evaluation of the 3D morphological information to effectively present the injured structure.…”

Diagnosis of Choroidal Disease With Deep Learning-Based Image Enhancement and Volumetric Quantification of Optical Coherence Tomography

“…The algorithm is trained and validated with all OCT scan patterns, including macular scans, disc scans, and wide scans, and can enhance the image quality to a level comparable to that of the 128× registration average. 19 Third, a shadow reduction method is performed to minimize shadows cast by retinal vessels in the choroid ( Fig. 1 d) because these have a similar appearance to the choroidal vessels and if not removed will lead to artifacts during vessel segmentation.…”

“…1 d) because these have a similar appearance to the choroidal vessels and if not removed will lead to artifacts during vessel segmentation. 19 The algorithm minimizes the shadows by normalizing each A-line in the OCT volume with a filtered energy profile that compensates the reduced energy caused by shadowing; it is universally applicable to all scan patterns and does not change the contrast of the original image. Further, an attenuation compensation ( Fig 1 e) is executed to improve the image contrast in the deep choroid.…”

Deep Learning in Biomedical Optics

“…There exist other sources of noise to further degrade the image quality when the signal level is low. Denoising and despeckling are important applications of DNNs, which are often trained with the averaged reduced‐noise image as the “ground truth” in a U‐Net and ResNet [182,183]. GAN has also been applied and provided improved visual perception than the DNNs trained with only the least‐squares loss function [180] (Fig.…”