[Digital Image Processing and Deep Neural Networks in Ophthalmology - Current Trends]. / Digitale Bildverarbeitung und Tiefe Neuronale Netze in der Augenheilkunde ­ aktuelle Trends.

The use of deep neural networks ("deep learning") creates new possibilities in digital image processing. This approach has been widely applied and successfully used for the evaluation of image data in ophthalmology. In this article, the methodological approach of deep learning is examined and compared to the classical approach for digital image processing. The differences between the approaches are discussed and the increasingly important role of training data for model generation is explained. Furthermore, the approach of transfer learning for deep learning is presented with a representative data set from the field of corneal confocal microscopy. In this context, the advantages of the method and the specific problems when dealing with medical microscope data will be discussed.

Numerical time-domain simulation of diffusive ultrasound in concrete.

Certain aspects of diffusive ultrasound fields in concrete are still unknown and thus, systematic parameter studies using numerical time-domain simulations of the ultrasonic propagation process could lead to further insights into theoretical and experimental questions. In the present paper, the elastodynamic finite integration technique (EFIT) is used to simulate a diffusive reverberation measurement at a concrete specimen taking aggregates, pores, and viscoelastic damping explicitly into account. The numerical results for dissipation and diffusivity are compared with theoretical models. Moreover, the influence of air-filled pores in the cement matrix is demonstrated.

Detection of buried reference structures by use of atomic force acoustic microscopy.

The miniaturization of micro- and nanoelectronic components requires new methods for the inspection of buried inner structures at the nanoscale. We used the atomic force acoustic microscopy technique (AFAM) to image subsurface defects. This technique combines high lateral resolution with the capability to determine local elastic properties of materials near the surface. As the structures buried near the surface change the effective tip-sample contact stiffness it is possible to detect them. For the verification of the detection capabilities of AFAM we fabricated well-defined buried void structures with different geometries and dimensions. Large, thin, plate like structures of silicon nitride with a local filling were our first test samples. Then, sets of nine small, square, thin plates with thicknesses increasing stepwise from 30 to 270 nm were etched in a thinned silicon wafer. The last two samples contained wedge structures of widths varying between 1.6 and 10 µm. Our results showed that it was possible to detect buried void structures at depths between 180 and 900 nm. We also observed that the depths at which the buried defects can be detected by the use of the AFAM method depend on the defect dimensions and geometry, and on the mismatch in the elastic properties of the sample and the defects. The experimental results obtained for the groups of small, thin plates were verified by quantitative analysis via finite element method (FEM) simulations.

Image reconstruction of the subbasal nerve plexus with in vivo confocal microscopy.

PURPOSE: To overcome the anterior corneal mosaic (ACM) phenomenon in in vivo confocal laser scanning microscopy (CLSM) and to reconstruct undistorted images of the subbasal nerve plexus (SNP), facilitating morphometric analysis in the presence of ACM ridges. METHODS: CLSM was performed in five healthy volunteers. An original image processing algorithm based on phase correlation was used to analyze and reduce motion distortions in volume scan image sequences. Three-dimensional tracing of the SNP was performed to reconstruct images containing only the SNP layer, with nerve fibers clearly visible even in ACM areas. RESULTS: Real-time mapping of the SNP revealed the presence of ridges with K-structures underneath them in all cases. The occurrence of K-structures correlated directly with development of ACM observed by slit lamp and resulted in massive deformation at the level of Bowman's membrane, seriously interfering with examination of SNP structures. The average elevation of ACM ridges was 20.6 µm (range, 8.7-34.0 µm). The novel method presented permitted reconstruction of the SNP layer in regions of ACM. CONCLUSIONS: The described method allows the precise analysis and elimination of motion artifacts in CLSM volume scans, in conjunction with the capability to reconstruct SNP structures even in the presence of severe ACM. The robustness and automation of the described algorithms require ongoing development, but this will provide a sound basis for extended studies of corneal nerve regeneration or degeneration and for use in clinical practice.