Search

IntroductionInfectious keratitis is a vision threatening disease. Bacterial and fungal keratitis are often confused in the early stages, so right diagnosis and optimized treatment for causative organisms is crucial. Antibacterial and antifungal medications are completely different, and the prognosis for fungal keratitis is even much worse. Since the identification of microorganisms takes a long time, empirical treatment must be started according to the appearance of the lesion before an accurate diagnosis. Thus, we developed an automated deep learning (DL) based diagnostic system of bacterial and fungal keratitis based on the anterior segment photographs using two proposed modules, Lesion Guiding Module (LGM) and Mask Adjusting Module (MAM).MethodsWe used 684 anterior segment photographs from 107 patients confirmed as bacterial or fungal keratitis by corneal scraping culture. Both broad- and slit-beam images were included in the analysis. We set baseline classifier as ResNet-50. The LGM was designed to learn the location information of lesions annotated by ophthalmologists and the slit-beam MAM was applied to extract the correct feature points from two different images (broad- and slit-beam) during the training phase. Our algorithm was then externally validated using 98 images from Google image search and ophthalmology textbooks.ResultsA total of 594 images from 88 patients were used for training, and 90 images from 19 patients were used for test. Compared to the diagnostic accuracy of baseline network ResNet-50, the proposed method with LGM and MAM showed significantly higher accuracy (81.1 vs. 87.8%). We further observed that the model achieved significant improvement on diagnostic performance using open-source dataset (64.2 vs. 71.4%). LGM and MAM module showed positive effect on an ablation study.DiscussionThis study demonstrated that the potential of a novel DL based diagnostic algorithm for bacterial and fungal keratitis using two types of anterior segment photographs. The proposed network containing LGM and slit-beam MAM is robust in improving the diagnostic accuracy and overcoming the limitations of small training data and multi type of images.

Although the hyperspectral image (HSI) classification has adopted deep neural networks (DNNs) and shown remarkable performances, there is a lack of studies of the adversarial vulnerability for the HSI classifications. In this paper, we propose a novel HSI classification framework robust to adversarial attacks. To this end, we focus on the unique spectral characteristic of HSIs (i.e., distinctive spectral patterns of materials). With the spectral characteristic, we present the random spectral sampling and spectral shape feature encoding for the robust HSI classification. For the random spectral sampling, spectral bands are randomly sampled from the entire spectrum for each pixel of the input HSI. Also, the overall spectral shape information, which is robust to adversarial attacks, is fed into the shape feature extractor to acquire the spectral shape feature. Then, the proposed framework can provide the adversarial robustness of HSI classifiers via randomization effects and spectral shape feature encoding. To the best of our knowledge, the proposed framework is the first work dealing with the adversarial robustness in the HSI classification. In experiments, we verify that our framework improves the adversarial robustness considerably under diverse adversarial attack scenarios, and outperforms the existing adversarial defense methods.

The stripe fixed pattern noise (FPN) of infrared images significantly corrupts image quality, so that most infrared imaging systems suffer from the degradation of visibility and detectability during operation. Therefore, the FPN de-striping method, which eliminates stripe patterns without substantial loss of image information, remains a core technology in the field of infrared image processing. In this article, we propose the dual-branch structure based FPN de-striping deep convolutional neural network (DBS-DCN) to effectively extract structural features of FPN and preserve the image details in a single infrared image. In addition, we have established the parametric FPN model through the diagnostic experiments of infrared images based on the physical principle of an infrared detector and its signal response. We have optimized each parameter of the FPN model using measured data, which acquired on a wide range of detector temperatures. Further, we generate the training data using our FPN model to ensure stable learning performance against various stripe patterns. We performed comparative experiments with state-of-the-art methods using artificially corrupted infrared images and real corrupted infrared data, and our proposed method achieved outstanding de-striping results in both qualitative and quantitative evaluation compared to existing methods.

The MPEG-21 Multimedia Framework allows visually impaired users to have an improved access to visual content by enabling content adaptation techniques such as color compensation. However, one important issue is the method to create and interpret the standardized CVD descriptions when making the use of generic color vision tests. In Part I of our study to tackle the issue, we present a novel computerized hue test (CHT) to examine and quantify CVD, which allows reproducing and manipulating test colors for the purposes of computer simulation and analysis of CVD. Both objective evaluation via color difference measurement and subjective evaluation via clinical experiment showed that the CHT works well as a color vision test: it is highly correlated with the Farnsworth-Munsell 100 Hue (FM100H) test and allows for a more elaborate and correct color reproduction than the FM100H test.

A color compensation scheme has been developed to enhance the perception of people with color vision deficiency (CVD) and for people suffering from anomalous trichromacy. It is operated within the MPEG-21 Multimedia Framework, which provides a standardized description of CVD. The basic idea behind the proposed color compensation consists of simulating the path of human color perception. As such, compensated color is realized by relying on the spectral cone sensitivities of the human eye and the spectral emission functions of the display device. For quantified color compensation, the spectral sensitivity of anomalous cones has been modeled according to the deficiency degree of the standardized CVD description. The latter is based on the error score of a computerized hue test (CHT), developed in Part I of our study. Given the anomalous cone spectra, the reduction of error score on the CHT after color compensation was measured in each deficiency degree. The quantitative relationship of color compensation with the error score is linearly regressed, based on the deficiency degree with the least error score after color compensation as well as the error score before color compensation.