Photovoltaic is one of the most important renewable energy sources for dealing with world-wide steadily increasing energy consumption. This raises the demand for fast and scalable automatic quality management during production and operation. However, the detection and segmentation of cracks on electroluminescence (EL) images of mono-or polycrystalline solar modules is a challenging task. In this work, we propose a weakly supervised learning strategy that only uses image-level annotations to obtain a method that is capable of segmenting cracks on EL images of solar cells. We use a modified ResNet-50 to derive a segmentation from network activation maps. We use defect classification as a surrogate task to train the network. To this end, we apply normalized L p normalization to aggregate the activation maps into single scores for classification. In addition, we provide a study how different parameterizations of the normalized L p layer affect the segmentation performance. This approach shows promising results for the given task. However, we think that the method has the potential to solve other weakly supervised segmentation problems as well.

No abstract

Introduction In addition to a wide range of disease-related burdens, patients with multiple sclerosis (MS) are challenged by long-term medication and complex disease management plans, often leading to decreased adherence. Earlier studies have indicated that patient support through intensified communication can positively impact adherence. During recent years, the use of mobile health applications has gained importance in patient communication and disease management. Objective Our objective was to describe a novel software application providing innovative features that engage patient–physician contact and to evaluate users’ acceptance. Methods A novel software application ensuring data security was developed. Various innovative modules have been implemented, enabling bidirectional communication between treating physicians and patients, supporting therapy monitoring and management, and allowing the collection of large sets of anonymous patient data. The currently conducted FASTER study will analyze the acceptance of patients and physicians using this novel application. Results PatientConcept is a free app available for download since 2016. Meanwhile, it has been successfully implemented. First preliminary results indicate high acceptance among users. The clinical benefit will have to be tested in larger patient populations. Conclusion This ID-associated application may provide a secure, feasible, and cost-optimized possibility to intensify and simplify the communication between patients and their treating physicians, thus promising an exceptional benefit to both.

Abstract. In image registration, a proper transformation should be topology preserving. Especially for landmark-based image registration, if the displacement of one landmark is larger enough than those of neighbourhood landmarks, topology violation will be occurred. This paper aim to analyse the topology preservation of some Radial Basis Functions (RBFs) which are used to model deformations in image registration. Matérn functions are quite common in the statistic literature (see, e.g. [9,12]). In this paper, we use them to solve the landmark-based image registration problem. We present the topology preservation properties of RBFs in one landmark and four landmarks model respectively. Numerical results of three kinds of Matérn transformations are compared with results of Gaussian, Wendland's, and Wu's functions.

Although Bacillus Calmette Guérin (BCG) remains a mainstay of adjuvant treatment in high-risk, non-muscle-invasive bladder cancer, BCG failure occurs in up to 40% of patients, with radical cystectomy (RC) as the inevitable therapeutic consequence. Current data suggest that PD-L1 immunosuppressive signaling is responsible for BCG failure, supporting the therapeutic rationale of combining checkpoint inhibitors with BCG. To address the immune cascade in 19 RC specimens obtained after BCG failure, we applied a small immunohistochemical (IHC) panel consisting of selected markers (PD-L1, GATA-3, a disintegrin and metalloproteinase (ADAM) proteases, IL-10/IL-10R). A modified quick score was used for IHC semi-quantification of these markers in tumor cells (TC) and immune cells (IC) within two different regions: muscle-invasive bladder cancer (MIBC) and primary/concurrent carcinoma in situ (CIS). Contrary to expectation, PD-L1 was consistently low, irrespective of tumor region and cell type. Intriguingly, expression of ADAM17, which has been reported to release membrane-bound PD-L1, was high in both tumor regions and cell types. Moreover, expression of GATA3, IL-10, and IL-10R was also increased, indicative of a generally immunosuppressive tumor microenvironment in BCG failure. ADAM10 expression was associated with advanced tumor disease at RC. Our findings raise the possibility that ADAM proteases may cleave PD-L1 from the surface of bladder TC and possibly also from IC. Therefore, IHC assessment of PD-L1 expression seems to be insufficient and should be supplemented by ADAM10/17 in patients with BCG failure.