Automated machine learning-based classification of proliferative and non-proliferative diabetic retinopathy using optical coherence tomography angiography vascular density maps

OCT angiography 2023 update: focus on diabetic retinopathy

Optical coherence tomography angiography (OCTA) has become part of the clinical practice and its growing applications are in continuous development. Coherently with the growing concern about the human and economic cost of diabetes, diabetic retinopathy (DR) was the most popular topic for OCTA studies in the past year. The analysis of the literature reveals that applications of OCTA in DR are in continuous growth. In particular, ultrawide field (UWF) OCTA and artificial intelligence (AI) based on OCTA images are affirming as the new frontiers of scientific research in the field. Diagnostic accuracy of AI methods based on OCTA is equal or superior to the one based on OCT methods and also bears potential to detect systemic associations. UWF OCTA is noninvasive method that is reaching similar accuracy of FA in detection of neovascularization and intraretinal microvascular abnormalities (IRMAs) and has allowed better characterization of microvascular peripherical changes in DR. Lastly, deep capillary plexus (DCP) characteristics seem to play a pivotal role in the development of diabetic macular edema (DME) and refinement of biomarkers for different phenotypes of DME and diabetic macular ischemia (DMI) is currently on its way.

DISCOVER: 2-D multiview summarization of Optical Coherence Tomography Angiography for automatic diabetic retinopathy diagnosis

Diabetic Retinopathy (DR), an ocular complication of diabetes, is a leading cause of blindness worldwide. Traditionally, DR is monitored using Color Fundus Photography (CFP), a widespread 2-D imaging modality. However, DR classifications based on CFP have poor predictive power, resulting in suboptimal DR management. Optical Coherence Tomography Angiography (OCTA) is a recent 3-D imaging modality offering enhanced structural and functional information (blood flow) with a wider field of view. This paper investigates automatic DR severity assessment using 3-D OCTA. A straightforward solution to this task is a 3-D neural network classifier. However, 3-D architectures have numerous parameters and typically require many training samples. A lighter solution consists in using 2-D neural network classifiers processing 2-D en-face (or frontal) projections and/or 2-D cross-sectional slices. Such an approach mimics the way ophthalmologists analyze OCTA acquisitions: (1) en-face flow maps are often used to detect avascular zones and neovascularization, and (2) cross-sectional slices are commonly analyzed to detect macular edemas, for instance. However, arbitrary data reduction or selection might result in information loss. Two complementary strategies are thus proposed to optimally summarize OCTA volumes with 2-D images: (1) a parametric en-face projection optimized through deep learning and (2) a cross-sectional slice selection process controlled through gradient-based attribution. The full summarization and DR classification pipeline is trained from end to end. The automatic 2-D summary can be displayed in a viewer or printed in a report to support the decision. We show that the proposed 2-D summarization and classification pipeline outperforms direct 3-D classification with the advantage of improved interpretability.

Assessment of area and structural irregularity of retinal layers in diabetic retinopathy using machine learning and image processing techniques

Diabetes retinopathy prevention necessitates early detection, monitoring, and treatment. Non-invasive optical coherence tomography (OCT) shows structural changes in the retinal layer. OCT image evaluation necessitates retinal layer segmentation. The ability of our automated retinal layer segmentation to distinguish between normal, non-proliferative (NPDR), and proliferative diabetic retinopathy (PDR) was investigated in this study using quantifiable biomarkers such as retina layer smoothness index (SI) and area (S) in horizontal and vertical OCT images for each zone (fovea, superior, inferior, nasal, and temporal). This research includes 84 eyes from 57 individuals. The study shows a significant difference in the Area (S) of inner nuclear layer (INL) and outer nuclear layer (ONL) in the horizontal foveal zone across the three groups (p < 0.001). In the horizontal scan, there is a significant difference in the smoothness index (SI) of the inner plexiform layer (IPL) and the upper border of the outer plexiform layer (OPL) among three groups (p < 0.05). There is also a significant difference in the area (S) of the OPL in the foveal zone among the three groups (p = 0.003). The area (S) of the INL in the foveal region of horizontal slabs performed best for distinguishing diabetic patients (NPDR and PDR) from normal individuals, with an accuracy of 87.6%. The smoothness index (SI) of IPL in the nasal zone of horizontal foveal slabs was the most accurate at 97.2% in distinguishing PDR from NPDR. The smoothness index of the top border of the OPL in the nasal zone of horizontal slabs was 84.1% accurate in distinguishing NPDR from PDR. Smoothness index of IPL in the temporal zone of horizontal slabs was 89.8% accurate in identifying NPDR from PDR patients. In conclusion, optical coherence tomography can assess the smoothness index and irregularity of the inner and outer plexiform layers, particularly in the nasal and temporal regions of horizontal foveal slabs, to distinguish non-proliferative from proliferative diabetic retinopathy. The evolution of diabetic retinopathy throughout severity levels and its effects on retinal layer irregularity need more study.

IMoVR-Net: A robust interpretable network for multi-ocular lesion recognition from TAO facial images

Thyroid-associated ophthalmopathy (TAO) is an organ-specific autoimmune disease that seriously affects patient's life and health. However, the early diagnosis of TAO is highly dependent on the physician's subjective experience. Moreover, the currently proposed deep learning networks for eye diseases do not provide robust interpretability concerning feature learning paradigm, model structure, and the number of neurons. But the mentioned components are very important for model interpretability and are key factors that severely affect the transparency of the model. Therefore, a robust interpretable multi-orientation visual recognition network (IMoVR-Net) for TAO multi-ocular lesion recognition is proposed in this paper. Firstly, a multi-orientation visual cascaded encoder composed of the DenseGabor module and the dilated Gabor convolution group is proposed to achieve the fine extraction of multi-directional TAO lesion features by using a novel feature learning paradigm called alternating filtering. Besides, combining information theory and topology tool, an optimization rule based on topological energy entropy is proposed to provide a solid interpretable theory for determining the model structure. Finally, a clustering correlation analysis method is developed to accomplish the determination of the number of convolutional hidden neurons, providing robust interpretability for the selection of the number of neurons. Compared to other advanced models, the IMoVR-Net achieved state-of-the-art performance on different TAO ocular datasets with an average accuracy, sensitivity, precision, and F1 score of 0.878, 87.27 %, 0.875, and 87.39 %, respectively. The IMoVR-Net has good clinical application prospects due to its strong recognition ability and robust interpretability in feature extraction paradigm, model structure, and number of convolutional neurons.

Wavelet scattering transform application in classification of retinal abnormalities using OCT images

To assist ophthalmologists in diagnosing retinal abnormalities, Computer Aided Diagnosis has played a significant role. In this paper, a particular Convolutional Neural Network based on Wavelet Scattering Transform (WST) is used to detect one to four retinal abnormalities from Optical Coherence Tomography (OCT) images. Predefined wavelet filters in this network decrease the computation complexity and processing time compared to deep learning methods. We use two layers of the WST network to obtain a direct and efficient model. WST generates a sparse representation of the images which is translation-invariant and stable concerning local deformations. Next, a Principal Component Analysis classifies the extracted features. We evaluate the model using four publicly available datasets to have a comprehensive comparison with the literature. The accuracies of classifying the OCT images of the OCTID dataset into two and five classes were [Formula: see text] and [Formula: see text], respectively. We achieved an accuracy of [Formula: see text] in detecting Diabetic Macular Edema from Normal ones using the TOPCON device-based dataset. Heidelberg and Duke datasets contain DME, Age-related Macular Degeneration, and Normal classes, in which we achieved accuracy of [Formula: see text] and [Formula: see text], respectively. A comparison of our results with the state-of-the-art models shows that our model outperforms these models for some assessments or achieves nearly the best results reported so far while having a much smaller computational complexity.