Enhancing Readability and Detection of Age-Related Macular Degeneration Using Optical Coherence Tomography Imaging: An AI Approach.

Artificial intelligence has been used effectively in medical diagnosis. The objective of this project is to examine the application of a collective AI model using weighted fusion of predicted probabilities from different AI architectures to diagnose various retinal conditions based on optical coherence tomography (OCT). A publicly available Noor dataset, comprising 16,822, images from 554 retinal OCT scans of 441 patients, was used to predict a diverse spectrum of age-related macular degeneration (AMD) stages: normal, drusen, or choroidal neovascularization. These predictions were compared with predictions from ResNet, EfficientNet, and Attention models, respectively, using precision, recall, F1 score, and confusion matric and receiver operating characteristics curves. Our collective model demonstrated superior accuracy in classifying AMD compared to individual ResNet, EfficientNet, and Attention models, showcasing the effectiveness of using trainable weights in the ensemble fusion process, where these weights dynamically adapt during training rather than being fixed values. Specifically, our ensemble model achieved an accuracy of 91.88%, precision of 92.54%, recall of 92.01%, and F1 score of 92.03%, outperforming individual models. Our model also highlights the refinement process undertaken through a thorough examination of initially misclassified cases, leading to significant improvements in the model's accuracy rate to 97%. This study also underscores the potential of AI as a valuable tool in ophthalmology. The proposed ensemble model, combining different mechanisms highlights the benefits of model fusion for complex medical image analysis.

Concurrent Learning Approach for Estimation of Pelvic Tilt from Anterior-Posterior Radiograph.

Accurate and reliable estimation of the pelvic tilt is one of the essential pre-planning factors for total hip arthroplasty to prevent common post-operative complications such as implant impingement and dislocation. Inspired by the latest advances in deep learning-based systems, our focus in this paper has been to present an innovative and accurate method for estimating the functional pelvic tilt (PT) from a standing anterior-posterior (AP) radiography image. We introduce an encoder-decoder-style network based on a concurrent learning approach called VGG-UNET (VGG embedded in U-NET), where a deep fully convolutional network known as VGG is embedded at the encoder part of an image segmentation network, i.e., U-NET. In the bottleneck of the VGG-UNET, in addition to the decoder path, we use another path utilizing light-weight convolutional and fully connected layers to combine all extracted feature maps from the final convolution layer of VGG and thus regress PT. In the test phase, we exclude the decoder path and consider only a single target task i.e., PT estimation. The absolute errors obtained using VGG-UNET, VGG, and Mask R-CNN are 3.04 ± 2.49, 3.92 ± 2.92, and 4.97 ± 3.87, respectively. It is observed that the VGG-UNET leads to a more accurate prediction with a lower standard deviation (STD). Our experimental results demonstrate that the proposed multi-task network leads to a significantly improved performance compared to the best-reported results based on cascaded networks.

Multi-scale convolutional neural network for automated AMD classification using retinal OCT images.

BACKGROUND AND OBJECTIVE: Age-related macular degeneration (AMD) is the most common cause of blindness in developed countries, especially in people over 60 years of age. The workload of specialists and the healthcare system in this field has increased in recent years mainly due to three reasons: 1) increased use of retinal optical coherence tomography (OCT) imaging technique, 2) prevalence of population aging worldwide, and 3) chronic nature of AMD. Recent advancements in the field of deep learning have provided a unique opportunity for the development of fully automated diagnosis frameworks. Considering the presence of AMD-related retinal pathologies in varying sizes in OCT images, our objective was to propose a multi-scale convolutional neural network (CNN) that can capture inter-scale variations and improve performance using a feature fusion strategy across convolutional blocks. METHODS: Our proposed method introduces a multi-scale CNN based on the feature pyramid network (FPN) structure. This method is used for the reliable diagnosis of normal and two common clinical characteristics of dry and wet AMD, namely drusen and choroidal neovascularization (CNV). The proposed method is evaluated on the national dataset gathered at Hospital (NEH) for this study, consisting of 12649 retinal OCT images from 441 patients, and the UCSD public dataset, consisting of 108312 OCT images from 4686 patients. RESULTS: Experimental results show the superior performance of our proposed multi-scale structure over several well-known OCT classification frameworks. This feature combination strategy has proved to be effective on all tested backbone models, with improvements ranging from 0.4% to 3.3%. In addition, gradual learning has proved to be effective in improving performance in two consecutive stages. In the first stage, the performance was boosted from 87.2%±2.5% to 92.0%±1.6% using pre-trained ImageNet weights. In the second stage, another performance boost from 92.0%±1.6% to 93.4%±1.4% was observed as a result of fine-tuning the previous model on the UCSD dataset. Lastly, generating heatmaps provided additional proof for the effectiveness of our multi-scale structure, enabling the detection of retinal pathologies appearing in different sizes. CONCLUSION: The promising quantitative results of the proposed architecture, along with qualitative evaluations through generating heatmaps, prove the suitability of the proposed method to be used as a screening tool in healthcare centers assisting ophthalmologists in making better diagnostic decisions.

Classification of Breast Cancer Lesions in Ultrasound Images by Using Attention Layer and Loss Ensemble in Deep Convolutional Neural Networks.

The reliable classification of benign and malignant lesions in breast ultrasound images can provide an effective and relatively low-cost method for the early diagnosis of breast cancer. The accuracy of the diagnosis is, however, highly dependent on the quality of the ultrasound systems and the experience of the users (radiologists). The use of deep convolutional neural network approaches has provided solutions for the efficient analysis of breast ultrasound images. In this study, we propose a new framework for the classification of breast cancer lesions with an attention module in a modified VGG16 architecture. The adopted attention mechanism enhances the feature discrimination between the background and targeted lesions in ultrasound. We also propose a new ensembled loss function, which is a combination of binary cross-entropy and the logarithm of the hyperbolic cosine loss, to improve the model discrepancy between classified lesions and their labels. This combined loss function optimizes the network more quickly. The proposed model outperformed other modified VGG16 architectures, with an accuracy of 93%, and also, the results are competitive with those of other state-of-the-art frameworks for the classification of breast cancer lesions. Our experimental results show that the choice of loss function is highly important and plays a key role in breast lesion classification tasks. Additionally, by adding an attention block, we could improve the performance of the model.

Using a deep convolutional network to predict the longitudinal dispersion coefficient.

Given the interest in accurately predicting the Longitudinal Dispersion Coefficient (Dx) within the fields of hydraulic and water quality modeling, a wide range of methods have been used to estimate this parameter. In order to improve the accuracy of Dx predictions, this paper proposes the use of a Deep Convolutional Network (DCN), a sub-field of machine learning. The proposed deep neural network architecture consists of two parts; first, a one-dimensional convolutional neural network (CNN) to build informative feature maps, and second, a stack of deep, fully connected layers to estimate pollution dispersion (as Dx) in streams. To accurately predict Dx the developed model draws upon a large and diverse array of datasets in the form of three dimensionless parameters: Width/Depth (W/H), Velocity/Shear Velocity (U/u*), and Longitudinal Dispersion Coefficient/(Depth * Shear Velocity) (Dx /Hu*). The model's accuracy is compared to that of several empirical models using a number of statistical measures. In addition, the DCN model results are compared with artificial neural network (ANN) and support vector machine (SVM) models implemented in this research and also similar studies applying various machine learning models (ML) towards Dx prediction. The statistical evaluation indicates that the DCN model outperforms the tested empirical, ANN, SVM and ML models with a significant difference. Additionally, five-fold cross-validation is performed to analyze the sensitivity and dependency of the DCN model's results on dataset selection, which shows that the dataset selection process does not significantly affect the model's accuracy. Since both ML and empirical models are, in general, poor predictors of the upper and lower ranges of Dx values, the DCN model's predictions of Dx in six different extreme-value ranges are assessed. The DCN model shows excellent accuracy in estimating Dx over the full possible range of data. In comparison with the empirical and ML models mentioned above, the DCN model more accurately predicts Dx values from river geometry and hydraulic datasets, with low errors across all ranges of Dx. The most significant advantage of DCN is that it tries to learn high-level features from data in an incremental manner.

Proposing a convolutional neural network for stress assessment by means of derived heart rate from functional near infrared spectroscopy.

BACKGROUND: Stress is known as one of the major factors threatening human health. A large number of studies have been performed in order to either assess or relieve stress by analyzing the brain and heart-related signals. METHOD: In this study, a method based on the Convolutional Neural Network (CNN) approach is proposed to assess stress induced by the Montreal Imaging Stress Task. The proposed model is trained on the heart rate signal derived from functional Near-Infrared Spectroscopy (fNIRS), which is referred to as HRF. In this regard, fNIRS signals of 20 healthy volunteers were recorded using a configuration of 23 channels located on the prefrontal cortex. The proposed deep learning system consists of two main parts where in the first part, the one-dimensional convolutional neural network is employed to build informative activation maps, and then in the second part, a stack of deep fully connected layers is used to predict the stress existence probability. Thereafter, the employed CNN method is compared with the Dense Neural Network, Support Vector Machine, and Random Forest regarding various classification metrics. RESULTS: Results clearly showed the superiority of CNN over all other methods. Additionally, the trained HRF model significantly outperforms the model trained on the filtered fNIRS signals, where the HRF model could achieve 98.69 ± 0.45% accuracy, which is 10.09% greater than the accuracy obtained by the fNIRS model. CONCLUSIONS: Employment of the proposed deep learning system trained on the HRF measurements leads to higher stress classification accuracy than the accuracy reported in the existing studies where the same experimental procedure has been done. Besides, the proposed method suggests better stability with lower variation in prediction. Furthermore, its low computational cost opens up the possibility to be applied in real-time monitoring of stress assessment.

Fully automatic estimation of pelvic sagittal inclination from anterior-posterior radiography image using deep learning framework.

BACKGROUND AND OBJECTIVE: Malposition of the acetabular component causes dislocation and prosthetic impingement after Total Hip Arthroplasty (THA), which significantly affects the postoperative quality of life and implant longevity. The position of the acetabular component is determined by the Pelvic Sagittal Inclination (PSI), which not only varies among different people but also changes in different positions. It is important to recognize individual dynamic changes of the PSI for patient-specific planning of the THA. Previously PSI was estimated by registering the CT and radiography images. In this study, we introduce a new method for accurate estimation of functional PSI without requiring CT image in order to lower radiation exposure of the patient which opens up the possibility of increasing its application in a larger number of hospitals where CT is not acquired as a routine protocol. METHODS: The proposed method consists of two main steps: First, the Mask R-CNN framework was employed to segment the pelvic shape from the background in the radiography images. Then, following the segmentation network, another convolutional network regressed the PSI angle. We employed a transfer learning paradigm where the network weights were initialized by non-medical images followed by fine-tuning using radiography images. Furthermore, in the training process, augmented data was generated to improve the performance of both networks. We analyzed the role of segmentation network in our system and investigated the Mask R-CNN performance in comparison with the U-Net, which is commonly used for the medical image segmentation. RESULTS: In this study, the Mask R-CNN utilizing multi-task learning, transfer learning, and data augmentation techniques achieve 0.960 ± 0.008 DICE coefficient, which significantly outperforms the U-Net. The cascaded system is capable of estimating the PSI with 4.04° ± 3.39° error for the radiography images. CONCLUSIONS: The proposed framework suggests a fully automatic and robust estimation of the PSI using only an anterior-posterior radiography image.