Deep learning for terahertz image denoising in nondestructive historical document analysis.

Historical documents contain essential information about the past, including places, people, or events. Many of these valuable cultural artifacts cannot be further examined due to aging or external influences, as they are too fragile to be opened or turned over, so their rich contents remain hidden. Terahertz (THz) imaging is a nondestructive 3D imaging technique that can be used to reveal the hidden contents without damaging the documents. As noise or imaging artifacts are predominantly present in reconstructed images processed by standard THz reconstruction algorithms, this work intends to improve THz image quality with deep learning. To overcome the data scarcity problem in training a supervised deep learning model, an unsupervised deep learning network (CycleGAN) is first applied to generate paired noisy THz images from clean images (clean images are generated by a handwriting generator). With such synthetic noisy-to-clean paired images, a supervised deep learning model using Pix2pixGAN is trained, which is effective to enhance real noisy THz images. After Pix2pixGAN denoising, 99% characters written on one-side of the Xuan paper can be clearly recognized, while 61% characters written on one-side of the standard paper are sufficiently recognized. The average perceptual indices of Pix2pixGAN processed images are 16.83, which is very close to the average perceptual index 16.19 of clean handwriting images. Our work has important value for THz-imaging-based nondestructive historical document analysis.

Maximum a posteriori signal recovery for optical coherence tomography angiography image generation and denoising.

Optical coherence tomography angiography (OCTA) is a novel and clinically promising imaging modality to image retinal and sub-retinal vasculature. Based on repeated optical coherence tomography (OCT) scans, intensity changes are observed over time and used to compute OCTA image data. OCTA data are prone to noise and artifacts caused by variations in flow speed and patient movement. We propose a novel iterative maximum a posteriori signal recovery algorithm in order to generate OCTA volumes with reduced noise and increased image quality. This algorithm is based on previous work on probabilistic OCTA signal models and maximum likelihood estimates. Reconstruction results using total variation minimization and wavelet shrinkage for regularization were compared against an OCTA ground truth volume, merged from six co-registered single OCTA volumes. The results show a significant improvement in peak signal-to-noise ratio and structural similarity. The presented algorithm brings together OCTA image generation and Bayesian statistics and can be developed into new OCTA image generation and denoising algorithms.

Automatic dementia screening and scoring by applying deep learning on clock-drawing tests.

Dementia is one of the most common neurological syndromes in the world. Usually, diagnoses are made based on paper-and-pencil tests and scored depending on personal judgments of experts. This technique can introduce errors and has high inter-rater variability. To overcome these issues, we present an automatic assessment of the widely used paper-based clock-drawing test by means of deep neural networks. Our study includes a comparison of three modern architectures: VGG16, ResNet-152, and DenseNet-121. The dataset consisted of 1315 individuals. To deal with the limited amount of data, which also included several dementia types, we used optimization strategies for training the neural network. The outcome of our work is a standardized and digital estimation of the dementia screening result and severity level for an individual. We achieved accuracies of 96.65% for screening and up to 98.54% for scoring, overcoming the reported state-of-the-art as well as human accuracies. Due to the digital format, the paper-based test can be simply scanned by using a mobile device and then be evaluated also in areas where there is a staff shortage or where no clinical experts are available.

Correction propagation for user-assisted optical coherence tomography segmentation: general framework and application to Bruch's membrane segmentation.

Optical coherence tomography (OCT) is a commonly used ophthalmic imaging modality. While OCT has traditionally been viewed cross-sectionally (i.e., as a sequence of B-scans), higher A-scan rates have increased interest in en face OCT visualization and analysis. The recent clinical introduction of OCT angiography (OCTA) has further spurred this interest, with chorioretinal OCTA being predominantly displayed via en face projections. Although en face visualization and quantitation are natural for many retinal features (e.g., drusen and vasculature), it requires segmentation. Because manual segmentation of volumetric OCT data is prohibitively laborious in many settings, there has been significant research and commercial interest in developing automatic segmentation algorithms. While these algorithms have achieved impressive results, the variability of image qualities and the variety of ocular pathologies cause even the most robust automatic segmentation algorithms to err. In this study, we develop a user-assisted segmentation approach, complementary to fully-automatic methods, wherein correction propagation is used to reduce the burden of manually correcting automatic segmentations. The approach is evaluated for Bruch's membrane segmentation in eyes with advanced age-related macular degeneration.

Virtual cleaning and unwrapping of non-invasively digitized soiled bamboo scrolls.

In ancient China, symbols and drawings captured on bamboo and wooden slips were used as main communication media. Those documents are very precious for cultural heritage and research, but due to aging processes, the discovered pieces are sometimes in a poor condition and contaminated by soil. Manual cleaning of excavated slips is a demanding and time-consuming task in which writings can be accidentally deleted. To counter this, we propose a novel approach based on conventional 3-D X-ray computed tomography to digitize such historical documents without before manual cleaning. By applying a virtual cleaning and unwrapping algorithm, the entire scroll surface is remapped into 2-D such that the hidden content becomes readable. We show that the technique also works for heavily soiled scrolls, enabling an investigation of the content by the naked eye without the need for manual labor. This digitization also allows for recovery of potentially erased writings and reconstruction of the original spatial information.

Browsing through sealed historical manuscripts by using 3-D computed tomography with low-brilliance X-ray sources.

Severely damaged historical documents are extremely fragile. In many cases, their secrets remain concealed beneath their cover. Recently, non-invasive digitization approaches based on 3-D scanning have demonstrated the ability to recover single pages or letters without the need to open the manuscripts. This can even be achieved using conventional micro-CTs without the need for synchrotron hardware. However, not all manuscripts may be suited for such techniques due to their material and X-ray properties. In order to recommend which manuscripts and which inks are best suited for such a process, we investigate six inks that were commonly used in ancient times: malachite, three types of iron gall, Tyrian purple, and buckthorn. Image contrast is explored over the complete pipeline, from the X-ray CT scan and page extraction to the virtual flattening of the page image. We demonstrate, that all inks containing metallic particles are visible in the output, a decrease of the X-ray energy enhances the readability, and that the visibility highly depends on the X-ray attenuation of the ink's metallic ingredients and their concentration. Based on these observations, we give recommendations on how to select the appropriate imaging parameters.