MISM: A Medical Image Segmentation Metric for Evaluation of Weak Labeled Data.

Performance measures are an important tool for assessing and comparing different medical image segmentation algorithms. Unfortunately, the current measures have their weaknesses when it comes to assessing certain edge cases. These limitations arise when images with a very small region of interest or without a region of interest at all are assessed. As a solution to these limitations, we propose a new medical image segmentation metric: MISm. This metric is a composition of the Dice similarity coefficient and the weighted specificity. MISm was investigated for definition gaps, an appropriate scoring gradient, and different weighting coefficients used to propose a constant value. Furthermore, an evaluation was performed by comparing the popular metrics in the medical image segmentation and MISm using images of magnet resonance tomography from several fictitious prediction scenarios. Our analysis shows that MISm can be applied in a general way and thus also covers the mentioned edge cases, which are not covered by other metrics, in a reasonable way. In order to allow easy access to MISm and therefore widespread application in the community, as well as reproducibility of experimental results, we included MISm in the publicly available evaluation framework MISeval.

Retinoschisin and novel Na/K-ATPase interaction partners Kv2.1 and Kv8.2 define a growing protein complex at the inner segments of mammalian photoreceptors.

The RS1 gene on Xp 22.13 encodes retinoschisin which is known to directly interact with the retinal Na/K-ATPase at the photoreceptor inner segments. Pathologic mutations in RS1 cause X-linked juvenile retinoschisis (XLRS), a hereditary retinal dystrophy in young males. To further delineate the retinoschisin-Na/K-ATPase complex, co-immunoprecipitation was performed with porcine and murine retinal lysates targeting the ATP1A3 subunit. This identified the voltage-gated potassium (Kv) channel subunits Kv2.1 and Kv8.2 as direct interaction partners of the retinal Na/K-ATPase. Colocalization of the individual components of the complex was demonstrated at the membrane of photoreceptor inner segments. We further show that retinoschisin-deficiency, a frequent consequence of molecular pathology in XLRS, causes mislocalization of the macromolecular complex during postnatal retinal development with a simultaneous reduction of Kv2.1 and Kv8.2 protein expression, while the level of retinal Na/K-ATPase expression remains unaffected. Patch-clamp analysis revealed no effect of retinoschisin-deficiency on Kv channel mediated potassium ion currents in vitro. Together, our data suggest that Kv2.1 and Kv8.2 together with retinoschisin and the retinal Na/K-ATPase are integral parts of a macromolecular complex at the photoreceptor inner segments. Defective compartmentalization of this complex due to retinoschisin-deficiency may be a crucial step in initial XLRS pathogenesis.

Retinoschisin and Cardiac Glycoside Crosstalk at the Retinal Na/K-ATPase.

Purpose: Mutations in the RS1 gene, which encodes retinoschisin, cause X-linked juvenile retinoschisis, a retinal dystrophy in males. Retinoschisin specifically interacts with the retinal sodium-potassium adenosine triphosphatase (Na/K-ATPase), a transmembrane ion pump. Na/K-ATPases also bind cardiac glycosides, which control the activity of the pump and have been linked to disturbances in retinal homeostasis. In this study, we investigated the crosstalk between retinoschisin and cardiac glycosides at the retinal Na/K-ATPase and the consequences of this interplay on retinal integrity. Methods: The effect of cardiac glycosides (ouabain and digoxin) on the binding of retinoschisin to the retinal Na/K-ATPase was investigated via western blot and immunocytochemistry. Also, the influence of retinoschisin on the binding of cardiac glycosides was analyzed via enzymatic assays, which quantified cardiac glycoside-sensitive Na/K-ATPase pump activity. Moreover, retinoschisin-dependent binding of tritium-labeled ouabain to the Na/K-ATPase was determined. Finally, a reciprocal effect of retinoschisin and cardiac glycosides on Na/K-ATPase localization and photoreceptor degeneration was addressed using immunohistochemistry in retinoschisin-deficient murine retinal explants. Results: Cardiac glycosides displaced retinoschisin from the retinal Na/K-ATPase; however, retinoschisin did not affect cardiac glycoside binding. Notably, cardiac glycosides reduced the capacity of retinoschisin to regulate Na/K-ATPase localization and to protect against photoreceptor degeneration. Conclusions: Our findings reveal opposing effects of retinoschisin and cardiac glycosides on retinal Na/K-ATPase binding and on retinal integrity, suggesting that a fine-tuned interplay between both components is required to maintain retinal homeostasis. This observation provides new insight into the mechanisms underlying the pathological effects of cardiac glycoside treatment on retinal integrity.

Identification of the retinoschisin-binding site on the retinal Na/K-ATPase.

X-linked juvenile retinoschisis (XLRS) is a hereditary retinal dystrophy, caused by mutations in the RS1 gene which encodes the secreted protein retinoschisin. In recent years, several molecules have been proposed to interact with retinoschisin, including the retinal Na/K-ATPase, L-voltage gated Ca2+ channels, and specific sugars. We recently showed that the retinal Na/K-ATPase consisting of subunits ATP1A3 and ATP1B2 is essential for anchoring retinoschisin to plasma membranes and identified the glycosylated ATP1B2 subunit as the direct interaction partner for retinoschisin. We now aimed to precisely map the retinoschisin binding domain(s) in ATP1B2. In general, retinoschisin binding was not affected after selective elimination of individual glycosylation sites via site-directed mutagenesis as well as after full enzymatic deglycosylation of ATP1B2. Applying the interface prediction tool PresCont, two putative protein-protein interaction patches ("patch I" and "patch II") consisting each of four hydrophobic amino acid stretches on the ATP1B2 surface were identified. These were consecutively altered by site-directed mutagenesis. Functional assays with the ATP1B2 patch mutants identified patch II and, specifically, the associated amino acid at position 240 (harboring a threonine in ATP1B2) as crucial for retinoschisin binding to ATP1B2. These and previous results led us to suggest an induced-fit binding mechanism for the interaction between retinoschisin and the Na/K-ATPase, which is dependent on threonine 240 in ATP1B2 allowing the accommodation of hyperflexible retinoschisin spikes by the associated protein-protein interaction patch on ATP1B2.

Pathomechanism of mutated and secreted retinoschisin in X-linked juvenile retinoschisis.

Mutations in the RS1 gene encoding retinoschisin cause X-linked juvenile retinoschisis (XLRS), a hereditary retinal dystrophy in males. While most of the XLRS associated mutations strongly interfere with cellular secretion, this is not true for mutants RS1-F108C, -R141G, -R141H, -R182C, -H207Q and -R209H. Native retinoschisin builds double-octamers and binds to retinal membranes, interacting with the retinal Na/K-ATPase. Functionally, it regulates MAP kinase signaling and Na/K-ATPase localization, and hampers photoreceptor degeneration. In this study, we investigated the capacity of the retinoschisin mutants still secreted extracellularly to fulfil these tasks. We addressed secretion and oligomerization of the heterologously expressed mutants as well as their binding to recombinant retinal Na/K-ATPases and murine retinoschisin-deficient (Rs1h-/Y) retinal and non-retinal explants. This has refined the categorization of secreted retinoschisin mutants: (i) no octamerization, unspecific membrane binding (RS1-F108C and -R182C), (ii) double-octamerization but no membrane binding (RS1-R141H), and (iii) double-octamerization and unspecific membrane binding (RS1-R141G, -H207Q, and -R209H). Notably, selected mutants of all categories (RS1-F108C, -R141H, and -R209H) failed to regulate retinal MAP kinase signaling and Na/K-ATPase localization in Rs1h-/Y retinal explants, and could not attenuate photoreceptor degeneration. Bioinformatic modeling of the secreted mutants depicted prominent alterations in the spatial and temporal conformation of a substructure called "spike 3" and its vicinity, implying a crucial role of this substructure for binding capacity and specificity. Taken together, our data point to a pathomechanism for secreted retinoschisin mutants, specifically to disturbances of the retinoschisin interface accompanied by unphysiological membrane interactions and impaired regulatory functions.

Alpha-syntrophin null mice are protected from non-alcoholic steatohepatitis in the methionine-choline-deficient diet model but not the atherogenic diet model.

Adipose tissue dysfunction contributes to the pathogenesis of non-alcoholic steatohepatitis (NASH). The adapter protein alpha-syntrophin (SNTA) is expressed in adipocytes. Knock-down of SNTA increases preadipocyte proliferation and formation of small lipid droplets, which are both characteristics of healthy adipose tissue. To elucidate a potential protective role of SNTA in NASH, SNTA null mice were fed a methionine-choline-deficient (MCD) diet or an atherogenic diet which are widely used as preclinical NASH models. MCD diet mediated loss of fat mass was largely improved in SNTA-/- mice compared to the respective wild type animals. Hepatic lipids were mostly unchanged while the oxidative stress marker malondialdehyde was only induced in the wild type mice. The expression of inflammatory markers and macrophage immigration into the liver were reduced in SNTA-/- animals. This protective function of SNTA loss was absent in atherogenic diet induced NASH. Here, hepatic expression of inflammatory and fibrotic genes was similar in both genotypes though mutant mice gained less body fat during feeding. Hepatic cholesterol and ceramide were strongly induced in both strains upon feeding the atherogenic diet, while hepatic sphingomyelin, phosphatidylserine and phosphatidylethanolamine levels were suppressed. SNTA deficient mice are protected from fat loss and NASH in the experimental MCD model. NASH induced by an atherogenic diet is not influenced by loss of SNTA. The present study suggests the use of different experimental NASH models to study the pathophysiological role of proteins like SNTA in NASH.

Model-supported phototrophic growth studies with Scenedesmus obtusiusculus in a flat-plate photobioreactor.

Light-dependent growth of microalgae can vary remarkably depending on the cultivation system and microalgal strain. Cell size and the pigmentation of each strain, as well as reactor geometry have a great impact on absorption and scattering behavior within a photobioreactor. In this study, the light-dependent, cell-specific growth kinetics of a novel green algae isolate, Scenedesmus obtusiusculus, was studied in a LED-illuminated flat-plate photobioreactor on a lab-scale (1.8 L, 0.09 m2 ). First, pH-controlled batch processes were performed with S. obtusiusculus at different constant incident photon flux densities. The best performance was achieved by illuminating S. obtusiusculus with 1400 µmol photons m-2 s-1 at the surface of the flat-plate photobioreactor, resulting in the highest biomass concentration (4.95 ± 0.16 gCDW L-1 within 3.5 d) and the highest specific growth rate (0.22 h-1 ). The experimental data were used to identify the kinetic parameters of different growth models considering light inhibition for S. obtusiusculus. Light attenuation within the flat-plate photobioreactor was considered by varying light transfer models. Based on the identified kinetic growth model of S. obtusiusculus, an optimum growth rate of 0.22 h-1 was estimated at a mean integral photon flux density of 1072 µmol photons m-2 s-1 with the Beer-Lambert law and 1590 µmol photons m-2 s-1 with Schuster's light transfer model in the flat-plate photobioreactor. LED illumination was, thus, increased to keep the identified optimum mean integral photon flux density constant in the batch process assuming Schuster's light transfer model. Compared to the same constant incident photon flux density (1590 µmol photons m-2 s-1 ), biomass concentration was up to 24% higher using the lighting profile until a dry cell mass concentration of 14.4 ± 1.4 gCDW L-1 was reached. Afterward, the biomass concentration remained constant, whereas cell growth continued in the batch process with constant incident photon flux density. Finally, biomass concentration was 15.5 ± 1.5 gCDW L-1 and, thus, 7% higher compared to the corresponding batch process with lighting profile. Biotechnol. Bioeng. 2017;114: 308-320. © 2016 Wiley Periodicals, Inc.

Crystalline order and topological charges on capillary bridges.

We numerically investigate crystalline order on negative Gaussian curvature capillary bridges. In agreement with the experimental results in [W. Irvine et al., Nature, Pleats in crystals on curved surfaces, 2010, 468, 947] we observe for decreasing integrated Gaussian curvature, a sequence of transitions, from no defects to isolated dislocations, pleats, scars and isolated sevenfold disclinations. We especially focus on the dependency of topological charge on the integrated Gaussian curvature, for which we observe, again in agreement with the experimental results, no net disclination for an integrated curvature down to -10, and an approximately linear behavior from there on until the disclinations match the integrated curvature of -12. In contrast to previous studies in which ground states for each geometry are searched for, we here show that the experimental results, which are likely to be in a metastable state, can be best resembled by mimicking the experimental settings and continuously changing the geometry. The obtained configurations are only low energy local minima. The results are computed using a phase field crystal approach on catenoid-like surfaces and are highly sensitive to the initialization.

Hybrid column generation and large neighborhood search for the dial-a-ride problem.

Demographic change towards an ever aging population entails an increasing demand for specialized transportation systems to complement the traditional public means of transportation. Typically, users place transportation requests, specifying a pickup and a drop off location and a fleet of minibuses or taxis is used to serve these requests. The underlying optimization problem can be modeled as a dial-a-ride problem. In the dial-a-ride problem considered in this paper, total routing costs are minimized while respecting time window, maximum user ride time, maximum route duration, and vehicle capacity restrictions. We propose a hybrid column generation and large neighborhood search algorithm and compare different hybridization strategies on a set of benchmark instances from the literature.

Solving the dynamic ambulance relocation and dispatching problem using approximate dynamic programming.

Emergency service providers are supposed to locate ambulances such that in case of emergency patients can be reached in a time-efficient manner. Two fundamental decisions and choices need to be made real-time. First of all immediately after a request emerges an appropriate vehicle needs to be dispatched and send to the requests' site. After having served a request the vehicle needs to be relocated to its next waiting location. We are going to propose a model and solve the underlying optimization problem using approximate dynamic programming (ADP), an emerging and powerful tool for solving stochastic and dynamic problems typically arising in the field of operations research. Empirical tests based on real data from the city of Vienna indicate that by deviating from the classical dispatching rules the average response time can be decreased from 4.60 to 4.01 minutes, which corresponds to an improvement of 12.89%. Furthermore we are going to show that it is essential to consider time-dependent information such as travel times and changes with respect to the request volume explicitly. Ignoring the current time and its consequences thereafter during the stage of modeling and optimization leads to suboptimal decisions.

Ambulance location and relocation problems with time-dependent travel times.

EMERGENCY SERVICE PROVIDERS ARE FACING THE FOLLOWING PROBLEM: how and where to locate vehicles in order to cover potential future demand effectively. Ambulances are supposed to be located at designated locations such that in case of an emergency the patients can be reached in a time-efficient manner. A patient is said to be covered by a vehicle if (s)he can be reached by an ambulance within a predefined time limit. Due to variations in speed and the resulting travel times it is not sufficient to solve the static ambulance location problem once using fixed average travel times, as the coverage areas themselves change throughout the day. Hence we developed a multi-period version, taking into account time-varying coverage areas, where we allow vehicles to be repositioned in order to maintain a certain coverage standard throughout the planning horizon. We have formulated a mixed integer program for the problem at hand, which tries to optimize coverage at various points in time simultaneously. The problem is solved metaheuristically using variable neighborhood search. We show that it is essential to consider time-dependent variations in travel times and coverage respectively. When ignoring them the resulting objective will be overestimated by more than 24%. By taking into account these variations explicitly the solution on average can be improved by more than 10%.

Carcinoma lobular invasivo no homem / Invasive lobular carcinoma in men

Os autores relatam um caso de carcinoma lobular invasivo em homem, ocorrência extremamente rara nesse sexo, enquanto corresponde a aproximadamente 10% dos carcinomas mamários femininos. Revisamos a literatura e encontramos 22 casos, apenas 1 publicado no Brasil, em 1997.

The authors report a case of invasive lobular carcinoma in man, which is extremely rare, while accord approximately 10% of breast cancers in women. We reviewed literature and found 22 cases, which of just one published in Brazil in 1997.