Optical coherence tomography (OCT) is a noninvasive, high-resolution imaging technique widely used in clinical practice to depict the structure of the retina. Over the past few decades, ophthalmologists have used OCT to diagnose, monitor, and treat retinal diseases. However, manual analysis of the complicated retinal layers using two colors, black and white, is time consuming. Although ophthalmologists have more experience, their results may be prone to erroneous diagnoses. Therefore, in this study, we propose an automatic method for diagnosing five retinal diseases based on the use of hybrid and ensemble deep learning (DL) methods. DL extracts a thousand constitutional features from images as features for training classifiers. The machine learning method classifies the extracted features and fuses the outputs of the two classifiers to improve classification performance. The distribution probabilities of two classifiers of the same class are aggregated; then, class prediction is made using the class with the highest probability. The limited dataset is resolved by the fine-tuning of classification knowledge and generating augmented images using transfer learning and data augmentation. Multiple DL models and machine learning classifiers are used to access a suitable model and classifier for the OCT images. The proposed method is trained and evaluated using OCT images collected from a hospital and exhibits a classification accuracy of 97.68% (InceptionResNetV2, ensemble: Extreme gradient boosting (XG-Boost) and k-nearest neighbor (k-NN). The experimental results show that our proposed method can improve the OCT classification performance; moreover, in the case of a limited dataset, the proposed method is critical to develop accurate classifications.

The ability of any steganography system to correctly retrieve the secret message is the primary criterion for measuring its efficiency. Recently, researchers have tried to generate a new natural image driven from only the secret message bits rather than using a cover to embed the secret message within it; this is called the stego image. This paper proposes a new secured coverless steganography system using a generative mathematical model based on semi Quick Response (QR) code and maze game image generation. This system consists of two components. The first component contains two processes, encryption process, and hiding process. The encryption process encrypts secret message bits in the form of a semi-QR code image whereas the hiding process conceals the pregenerated semi-QR code in the generated maze game image. On the other hand, the second component contains two processes, extraction and decryption, which are responsible for extracting the semi-QR code from the maze game image and then retrieving the original secret message from the extracted semi-QR code image, respectively. The results were obtained using the bit error rate (BER) metric. These results confirmed that the system achieved high hiding capacity, good performance, and a high level of robustness against attackers compared with other coverless steganography methods.

Manual inspection of fruit diseases is a time-consuming and costly because it is based on naked-eye observation. The authors present computer vision techniques for detecting and classifying fruit leaf diseases. Examples of computer vision techniques are preprocessing original images for visualization of infected regions, feature extraction from raw or segmented images, feature fusion, feature selection, and classification. The following are the major challenges identified by researchers in the literature: (i) low-contrast infected regions extract irrelevant and redundant information, which misleads classification accuracy; (ii) irrelevant and redundant information may increase computational time and reduce the designed model’s accuracy. This paper proposed a framework for fruit leaf disease classification based on deep hierarchical learning and best feature selection. In the proposed framework, contrast is first improved using a hybrid approach, and then data augmentation is used to solve the problem of an imbalanced dataset. The next step is to use a pre-trained deep model named Darknet53 and fine-tune it. Next, deep transfer learning-based training is carried out, and features are extracted using an activation function on the average pooling layer. Finally, an improved butterfly optimization algorithm is proposed, which selects the best features for classification using machine learning classifiers. The experiment was carried out on augmented and original fruit datasets, yielding a maximum accuracy of 99.6% for apple diseases, 99.6% for grapes, 99.9% for peach diseases, and 100% for cherry diseases. The overall average achieved accuracy is 99.7%, higher than previous techniques.

Melanoma is a skin disease with high mortality rate while early diagnoses of the disease can increase the survival chances of patients. It is challenging to automatically diagnose melanoma from dermoscopic skin samples. Computer-Aided Diagnostic (CAD) tool saves time and effort in diagnosing melanoma compared to existing medical approaches. In this background, there is a need exists to design an automated classification model for melanoma that can utilize deep and rich feature datasets of an image for disease classification. The current study develops an Intelligent Arithmetic Optimization with Ensemble Deep Transfer Learning Based Melanoma Classification (IAOEDTT-MC) model. The proposed IAOEDTT-MC model focuses on identification and classification of melanoma from dermoscopic images. To accomplish this, IAOEDTT-MC model applies image preprocessing at the initial stage in which Gabor Filtering (GF) technique is utilized. In addition, U-Net segmentation approach is employed to segment the lesion regions in dermoscopic images. Besides, an ensemble of DL models including ResNet50 and ElasticNet models is applied in this study. Moreover, AO algorithm with Gated Recurrent Unit (GRU) method is utilized for identification and classification of melanoma. The proposed IAOEDTT-MC method was experimentally validated with the help of benchmark datasets and the proposed model attained maximum accuracy of 92.09% on ISIC 2017 dataset.