Classifiers for Ischemic Stroke Lesion Segmentation: A Comparison Study

Motivation

Ischemic stroke, triggered by an obstruction in the cerebral blood supply, leads to infarction of the affected brain tissue. An accurate and reproducible automatic segmentation is of high interest, since the lesion volume is an important end-point for clinical trials. However, various factors, such as the high variance in lesion shape, location and appearance, render it a difficult task.

Methods

In this article, nine classification methods (e.g. Generalized Linear Models, Random Decision Forests and Convolutional Neural Networks) are evaluated and compared with each other using 37 multiparametric MRI datasets of ischemic stroke patients in the sub-acute phase in terms of their accuracy and reliability for ischemic stroke lesion segmentation. Within this context, a multi-spectral classification approach is compared against mono-spectral classification performance using only FLAIR MRI datasets and two sets of expert segmentations are used for inter-observer agreement evaluation.

Results and Conclusion

The results of this study reveal that high-level machine learning methods lead to significantly better segmentation results compared to the rather simple classification methods, pointing towards a difficult non-linear problem. The overall best segmentation results were achieved by a Random Decision Forest and a Convolutional Neural Networks classification approach, even outperforming all previously published results. However, none of the methods tested in this work are capable of achieving results in the range of the human observer agreement and the automatic ischemic stroke lesion segmentation remains a complicated problem that needs to be explored in more detail to improve the segmentation results.     

Development of an intein-mediated split–Cas9 system for gene therapy

Using CRISPR/Cas9, it is possible to target virtually any gene in any organism. A major limitation to its application in gene therapy is the size of Cas9 (>4 kb), impeding its efficient delivery via recombinant adeno-associated virus (rAAV). Therefore, we developed a split–Cas9 system, bypassing the packaging limit using split-inteins. Each Cas9 half was fused to the corresponding split-intein moiety and, only upon co-expression, the intein-mediated trans-splicing occurs and the full Cas9 protein is reconstituted. We demonstrated that the nuclease activity of our split-intein system is comparable to wild-type Cas9, shown by a genome-integrated surrogate reporter and by targeting three different endogenous genes. An analogously designed split-Cas9D10A nickase version showed similar activity as Cas9D10A. Moreover, we showed that the double nick strategy increased the homologous directed recombination (HDR). In addition, we explored the possibility of delivering the repair template accommodated on the same dual-plasmid system, by transient transfection, showing an efficient HDR. Most importantly, we revealed for the first time that intein-mediated split–Cas9 can be packaged, delivered and its nuclease activity reconstituted efficiently, in cells via rAAV.     

SNAP-25 is a promising novel cerebrospinal fluid biomarker for synapse degeneration in Alzheimer’s disease

Background

Synaptic degeneration is an early pathogenic event in Alzheimer’s disease, associated with cognitive impairment and disease progression. Cerebrospinal fluid biomarkers reflecting synaptic integrity would be highly valuable tools to monitor synaptic degeneration directly in patients. We previously showed that synaptic proteins such as synaptotagmin and synaptosomal-associated protein 25 (SNAP-25) could be detected in pooled samples of cerebrospinal fluid, however these assays were not sensitive enough for individual samples.

Results

We report a new strategy to study synaptic pathology by using affinity purification and mass spectrometry to measure the levels of the presynaptic protein SNAP-25 in cerebrospinal fluid. By applying this novel affinity mass spectrometry strategy on three separate cohorts of patients, the value of SNAP-25 as a cerebrospinal fluid biomarker for synaptic integrity in Alzheimer’s disease was assessed for the first time. We found significantly higher levels of cerebrospinal fluid SNAP-25 fragments in Alzheimer’s disease, even in the very early stages, in three separate cohorts. Cerebrospinal fluid SNAP-25 differentiated Alzheimer’s disease from controls with area under the curve of 0.901 (P?<?0.0001).

Conclusions

We developed a sensitive method to analyze SNAP-25 levels in individual CSF samples that to our knowledge was not possible previously. Our results support the notion that synaptic biomarkers may be important tools for early diagnosis, assessment of disease progression, and to monitor drug effects in treatment trials.

     

Cenobamate: Neuroprotective Potential of a New Antiepileptic Drug

Neurochemical Research - Central nervous system (CNS) injuries annually afflict approximately 2.7 million people in United States only, inflicting costs of nearly 100 billion US dollars. The...     

Single cell transcriptomics of neighboring hyphae of <Emphasis Type="Italic">Aspergillus niger</Emphasis>

Single cell profiling was performed to assess differences in RNA accumulation in neighboring hyphae of the fungus Aspergillus niger. A protocol was developed to isolate and amplify RNA from single hyphae or parts thereof. Microarray analysis resulted in a present call for 4 to 7% of the A. niger genes, of which 12% showed heterogeneous RNA levels. These genes belonged to a wide range of gene categories.     

Cholesterol's interactions with serine phospholipids — A comparison of N-palmitoyl ceramide phosphoserine with dipalmitoyl phosphatidylserine

In this study we have prepared ceramide phosphoserine (CerPS) and examined its sterol-interacting properties. CerPS is a hydrogen-bonding sphingolipid, but its head group differs from that found in sphingomyelin (SM). Based on diphenylhexatriene steady-state anisotropy measurements, we observed that fully hydrated N-palmitoyl CerPS had a gel-to-liquid crystalline phase transition temperature of about 51 °C in 50 mM sodium phosphate buffer (pH 7.4). This was close to the T_m measured for 1,2-dipalmitoyl-sn-glycero-3-phosphoserine (DPPS) bilayers (T_m 50.5 °C). Based on cholestatrienol (CTL) quenching experiments in liquid disordered ternary bilayers (containing 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphcholine; POPC), cholesterol/CTL formed sterol-enriched ordered domains with CerPS. These had similar thermostability as the sterol domains formed with N-palmitoyl SM. Cholesterol failed to form sterol-enriched ordered domains with DPPS under comparable conditions. Based on the equilibrium partitioning of CTL, we observed that the affinity of sterol for bilayers containing POPC/CerPS/cholesterol (6:3:1 by mol) was much higher than the affinity measured for control fluid POPC/cholesterol (9:1 by mol) bilayers, but slightly less than seen for comparable PSM-containing bilayers. We conclude that the phosphoserine head group was less efficient than the phosphocholine head group in stabilizing sterol/sphingolipid interaction. However, hydrogen bonding apparently can overcome some of the negative effects of the phosphoserine head group, since CerPS interacted more favorably with cholesterol compared to DPPS.