Background and Objective: It was the aim of this study to determine the prevalence of anti-aquaporin 4 antibody (anti-AQP4 Ab) and long spinal cord lesions in neuromyelitis optica (NMO) and multiple sclerosis (MS) patients in Taiwan. Asia has a relatively high rate of NMO compared with MS patients. Anti-AQP4 Ab is an important marker for NMO worldwide, but serological data and clinical profiles of NMO patients in Taiwan have not been reported. Methods: This retrospective study compared the clinical symptoms, demographics, spinal cord lesion length and AQP4 Ab status of 34 patients with NMO with 34 patients diagnosed with conventional MS. Results: Our NMO patients were predominantly middle-aged women (median age 45 years), exhibited many relapses (1.0/year) and displayed a higher Expanded Disability Status Scale score (4.75) than conventional MS patients. NMO patients exhibited long spinal cord lesions as detected by MRI. Forty-one percent of the NMO patients had detectable anti-AQP4 Ab. The Expanded Disability Status Scale score was significantly higher in AQP4 Ab– NMO patients. Conclusion: The prevalence of AQP4 Ab in a Taiwanese NMO group was 41%. Long spinal cord lesions and detection of AQP4 Ab helped to differentiate NMO patients from MS patients. Long spinal cord lesions with the anti-AQP4 Ab test may allow for an earlier diagnosis of NMO and improve therapeutic decisions.

A multidrug-resistant (MDR) cell line isolated from HOB1 lymphoma cells was characterized. The MDR phenotype in this cell line was typified by resistance to vincristine with varying degrees of cross-resistance to Adriamycin, colcemid and actinomycin D. Decreased intracellular [³H]vincristine with concurrent increase in the expression of a 170-kDa membrane glycoprotein (P-glycoprotein) suggested a plausible underlying mechanism for the development of resistance. Amplification of the mdr1gene as well as a homogeneous staining region on the long arm of the 7th chromosome was observed. Moreover, metabolic studies with [¹⁴C]glucosamine or [¹⁴C]mannose indicated differential expressions of membrane glycoproteins between the drug-sensitive parental and drug-resistant descendant cells. It is concluded that the development of drug resistance in HOB1 lymphoma cells was strongly correlated with the overexpression of P-glycoprotein.

No abstract

For timing-sensitive edge applications, the demand for efficient lightweight machine learning solutions has increased recently. Tree ensembles are among the state-of-the-art in many machine learning applications. While single decision trees are comparably small, an ensemble of trees can have a significant memory footprint leading to cache locality issues, which are crucial to performance in terms of execution time. In this work, we analyze memory-locality issues of the two most common realizations of decision trees, i.e. native and if-else trees. We highlight, that both realizations demand a more careful memory layout to improve caching behavior and maximize performance. We adopt a probabilistic model of decision tree inference to find the best memory layout for each tree at the application layer. Further, we present an efficient heuristic to take architecture-dependent information into account thereby optimizing the given ensemble for a target computer architecture. Our code-generation framework, which is freely available on an open-source repository, produces optimized code sessions while preserving the structure and accuracy of the trees. With several real-world data sets, we evaluate the elapsed time of various tree realizations on server hardware as well as embedded systems for Intel and ARM processors. Our optimized memory layout achieves a reduction in execution time up to 75 % execution for server-class systems, and up to 70 % for embedded systems, respectively.

No abstract