[Purpose] The purpose of this study was to examine the effects on stroke patients of trunk stabilization exercise on different support surfaces. [Subjects and Methods] Sixteen stroke patients with onset of stroke six months earlier or longer were randomly and equally assigned to group I (exercise performed on a stable support surface) and group II (exercise performed on an unstable support surface). The two groups conducted the trunk stabilization exercises on the respective support surfaces, in addition to existing rehabilitation exercises five times per week for 12 weeks. Changes in the cross-sectional area (CSA) of the muscles were examined using computed tomography (CT), and changes in the balance ability were assessed using a measuring system and the trunk impairment scale (TIS). [Results] In group I, there was a significant increase in the CSA of the mulifidus muscle on the side contralateral to the brain lesion and in the paravertebral and multifidus muscles on the side ipsilateral to the brain lesion. In group II, there was a significant increase in the CSA of the paravertebral and multifidus muscles on the side contralateral to the brain lesion and on the side ipsilateral to the brain lesion. In terms of changes in balance ability, the sway path (SP) and TIS significantly improved in group I, and the SP, sway area (SA), and TIS significantly improved in group II . [Conclusion] Exercise on the unstable support surface enhanced the size of the cross-sectional area of the trunk muscles and balance ability significantly more than exercise on the stable support surface.

We consider a novel question answering (QA) task where the machine needs to read from large streaming data (long documents or videos) without knowing when the questions will be given, which is difficult to solve with existing QA methods due to their lack of scalability. To tackle this problem, we propose a novel end-to-end deep network model for reading comprehension, which we refer to as Episodic Memory Reader (EMR) that sequentially reads the input contexts into an external memory, while replacing memories that are less important for answering unseen questions. Specifically, we train an RL agent to replace a memory entry when the memory is full, in order to maximize its QA accuracy at a future timepoint, while encoding the external memory using either the GRU or the Transformer architecture to learn representations that considers relative importance between the memory entries. We validate our model on a synthetic dataset (bAbI) as well as real-world large-scale textual QA (TriviaQA) and video QA (TVQA) datasets, on which it achieves significant improvements over rulebased memory scheduling policies or an RLbased baseline that independently learns the query-specific importance of each memory.

Phospholipase C-γl (PLC-γl) expression is associated with cellular transformation. Notably, PLC-γ is up-regulated in colorectal cancer tissue and breast carcinoma. Because exotoxins released by Clostridium botulinum have been shown to induce apoptosis and promote growth arrest in various cancer cell lines, we examined here the potential of Clostridium difficile toxin A to selectively induce apoptosis in cells transformed by PLC-γl overexpression. We found that PLC-γl-transformed cells, but not vectortransformed (control) cells, were highly sensitive to C. difficile toxin A-induced apoptosis and mitotic inhibition. Moreover, expression of the proapoptotic Bcl2 family member, Bim, and activation of caspase-3 were significantly up-regulated by toxin A in PLC-γl-transformed cells. Toxin A-induced cell rounding and paxillin dephosphorylation were also significantly higher in PLC-γl-transformed cells than in control cells. These findings suggest that C. difficile toxin A may have potential as an anticancer agent against colorectal cancers and breast carcinomas in which PLC-γl is highly up-regulated.

No abstract

No abstract