POOL and DOOM a survey of esprit 415 subproject A, Philips research laboratories

Author(s):  
Wim Bronnenberg
Keyword(s):  
2014 ◽  
Vol 45 (2) ◽  
pp. 27-31
Author(s):  
Dirk van Delft
Keyword(s):  

Vacuum ◽  
1967 ◽  
Vol 17 (5) ◽  
pp. 273
Author(s):  
Sir John Cockcroft
Keyword(s):  

Blood ◽  
2017 ◽  
Vol 130 (Suppl_1) ◽  
pp. 996-996
Author(s):  
David Dale ◽  
Audrey Anna Bolyard ◽  
Tracy M. Marrero ◽  
David A. Weinstein ◽  
Cornelia Zeidler ◽  
...  

Abstract Background: Glycogen-storage disease 1b (GSD 1b) is characterized by neutropenia, hepatomegaly, hypoglycemia, enterocolitis and recurrent severe infections. For more than 25 years, many GSD 1b patients worldwide have been treated with granulocyte colony-stimulating factor (G-CSF). Methods: The Severe Chronic Neutropenia International Registry (SCNIR) enrolls and prospectively follows GSD 1b patients with both severe (ANC< 0.5 x 109/L) and mild or moderate neutropenia (ANC< 1.5 x 109/L). For this report we analyzed records for 83 patients enrolled through the SCNIR offices in Seattle, Washington and Hannover, Germany. Findings - Clinical outcomes: Overall most patients and parents report very favorable outcomes with increases in neutrophils, reduced infections, reduced symptoms of enterocolitis and improved quality of life. Neutropenia: Because splenomegaly may increase markedly with G-CSF treatment, clinical experience dictates that the G-CSF dose should be carefully monitored to maintain the ANC at 1.0 to 2.0 x 109/L. Summary data for treatment of 83 patients are shown in Table 1. Myeloid malignancies: We assessed the risk of GSD 1b patients developing myeloid malignancies from review of the medical literature and patients prospectively followed through the SCNIR. We identified 1 case of AML prior to availability of G-CSF and 6 cases of AML in G-CSF treated patients, 4 from the cohort followed prospectively by the SCNIR. Table 2 shows detailed data for the 7 patients with GSD1b who have developed AML. Conclusions: In the 83 SCNIR patients followed for a mean (median) of 12.2 (11.0) years there are 4 cases (4/83, 4.8%) of AML, or 1 case/253 years of G-CSF treatment. This is lower than the estimated risk of evolution to AML in the major genetic subgroups of severe congenital neutropenia treated with G-CSF, e.g., harboring ELANE, HAX1, SDS mutations . The numbers of patients in other genetic subgroups, such as G6PC3, JAGN1, TCIRG1, etc. are too small to make any reliable comparison about the risk of developing leukemia. (Skokowa J et al. Nat Rev Dis Primers. 2017 Jun 8. doi: 10.1038/nrdp.2017.32. [Epub ahead of print]). Disclosures Dale: Amgen: Consultancy, Research Funding; Sanofi Aventis: Consultancy, Editor, Current Opinions in Hematology, Honoraria; Cellerant: Membership on an entity's Board of Directors or advisory committees; Genzyme (now owned by Sanofi-Aventis): Consultancy, Patents & Royalties, Research Funding; Genkyotex: Other: DSMB (work completed 6/2015); Hospira: Consultancy; Prolong: Consultancy; Boheringer-Ingelheim: Consultancy; Coherus: Consultancy; Omeros: Other: DSMB; Shire: Other: Independent Review Board; X4Pharma: Research Funding; Philips: Research Funding; Wolter Kluwer: Other: Editor, Current Opinions in Hematology; WedMD/Medscape: Membership on an entity's Board of Directors or advisory committees; National Institutes of Health: Research Funding; University of Washington: Employment, Research Funding; American College of Physicians: Other: Editor and author; GlaxoSmithKline: Equity Ownership; Johnson&Johnson: Equity Ownership. Weinstein: Vitaflo: Research Funding; Dimension Therapeutics: Membership on an entity's Board of Directors or advisory committees, Research Funding.


Author(s):  
H. NEY ◽  
V. STEINBISS ◽  
R. HAEB-UMBACH ◽  
B.-H. TRAN ◽  
U. ESSEN

This paper gives an overview of a research system for phoneme based, large vocabulary continuous speech recognition. The system to be described has been applied to the SPICOS task, the DARPA RM task and a 12000 word dictation task. Experimental results for these three tasks will be presented. Like many other systems, the recognition architecture is based on an integrated statistical approach. In this paper, we describe the characteristic features of the system as opposed to other systems: (1) The Viterbi criterion is consistently applied both in training and testing. (2) Continuous mixture densities are used without any tying or smoothing; this approach can be viewed as a sort of ‘statistical template matching’. (3) Time-synchronous beam search is used consistently throughout all tasks; extensions using a tree organization of the vocabulary and phoneme lookahead are presented so that a 12000 word task can be handled.


Sign in / Sign up

Export Citation Format

Share Document