Validation Study on Automated Sleep Stage Scoring Using a Deep Learning Algorithm.

Background and Objectives: Polysomnography is manually scored by sleep experts. However, manual scoring is a time-consuming and labor-intensive task. The goal of this study was to verify the accuracy of automated sleep-stage scoring based on a deep learning algorithm compared to manual sleep-stage scoring. Materials and Methods: A total of 602 polysomnography datasets from subjects (Male:Female = 397:205) aged 19 to 65 years (mean age, 43.8, standard deviation = 12.2) were included in the study. The performance of the proposed model was evaluated based on kappa value and bootstrapped point-estimate of median percent agreement with a 95% bootstrap confidence interval and R = 1000. The proposed model was trained using 482 datasets and validated using 48 datasets. For testing, 72 datasets were selected randomly. Results: The proposed model exhibited good concordance rates with manual scoring for stages W (94%), N1 (83.9%), N2 (89%), N3 (92%), and R (93%). The average kappa value was 0.84. For the bootstrap method, high overall agreement between the automated deep learning algorithm and manual scoring was observed in stages W (98%), N1 (94%), N2 (92%), N3 (99%), and R (98%) and total (96%). Conclusions: Automated sleep-stage scoring using the proposed model may be a reliable method for sleep-stage classification.

Polymorphisms of the prion protein gene (PRNP) in a Korean population.

Human prion protein gene (PRNP) has been considered to be involved in the susceptibility of humans to prion diseases. Polymorphisms of methionine (Met)/valine (Val) at codon 129 and of glutamic acid (Glu)/lysine (Lys) at codon 219 are thought to play an important role in susceptibility to sporadic, iatrogenic and variant Creutzfeldt-Jakob disease (CJD). Although the genotype distribution of polymorphisms in PRNP open reading frame (ORF) has been reported in many European populations, among Asian groups, it has been reported only in the Japanese population. We examined the PRNP polymorphisms in 529 healthy Koreans. We observed that genotype frequencies at codon 129 was 94.33% Met/Met, 5.48% Met/Val, and 0.19% Val/Val with an allele frequency of 0.971:0.029 Met:Val, and that genotype frequencies at codon 219 was 92.06% Glu/Glu, 7.94% Glu/Lys, and 0% Lys/Lys with an allele frequency of 0.96:0.04 Glu:Lys. The frequencies of the Glu/Glu genotype ( chi(2)=10.075, P=0.0015) and of the Glu allele ( chi(2)=9.486, P=0.0021) at codon 219 were significantly higher in the Korean population than the Japanese population. In addition, the genotype frequency of heterozygotes (12.7%) at codons 129 or/and 219 was significantly lower in Koreans than in people from Great Britain ( chi(2)=89.52, P<0.0001). The deletion rate of one octarepeat (R2 deletion) was 0.38%, with 99.62% undeleted homozygotes and 0% deleted homozygote. To our knowledge, the R2 octarepeat deletion has never been found in people from countries other than Korea. The data of PRNP polymorphism at codon 219 suggest that Koreans may be more sensitive to sporadic CJD than the Japanese population.

Regional outbreak of mumps due to genotype H in Korea in 1999.

Serological and virological studies were carried out of a mumps outbreak which occurred in one region, Yoeju County, Southeast of Seoul in Korea from September to December, 1999. Sera from 736 children at 8-13 years of age of patients with mumps and healthy children were tested for mumps-specific antibodies by enzyme immunoassay. The overall IgM positive rate was 7.6% (56/736), compared with 69.8% (514/736) for IgG. Of the 49 children with both IgG and IgM, 32 were also confirmed by both clinical and serological diagnosis. IgM antibodies were detected even in the samples collected up to 3 months after the onset of symptoms. Although 436 children had been vaccinated before the outbreak, 27 (6.2%) were found to be IgM positive, particularly 6 (4.4%) of 136 were positive serologically despite a second-dose vaccinees. Sequence analysis of the small hydrophobic (SH) gene of 4 mumps viruses isolated from 42 saliva specimens revealed that these were related to the genotype H, but distinguishable from European strains. This is the first study on the outbreak due to mumps virus genotype H and provides information to assess the understanding of recent outbreaks of mumps in Korea.

Complete nucleotide sequence of a mumps virus genotype I strain isolated in Korea.

The complete nucleotide sequence of mumps virus isolated in Korea, Dg1062/Korea/98 (Dg1062), was determined. As other mumps viruses, its genome was to be 15,384 nucleotides (nts) in length and encoded seven proteins. The both 5' and 3' ends were confirmed to be 55 and 24 nts by RACE method, respectively. The full-length nucleotide sequence of Dg1062 isolate differed from other strains by 2.9-6.8% in the nucleotide sequence level, resulting in 206 nucleotide and 54 amino acid substitutions which were observed in only Dg1062 isolate relative to the consensus sequences of other strains. Despite the variations of amino acids over the full genome including HN gene, it might be considered that this isolate have no significant variations in the antigenic sites. This result is the first report of full-length genome of genotype I strain and provides an overview on the diversity of genetic characteristics of circulating mumps virus.

Molecular characterization of two genotypes of mumps virus circulated in Korea during 1998-2001.

Sequence analyses of the entire small hydrophobic (SH) and hemagglutinin-neuraminidase (HN) genes of mumps viruses circulated in Korea from 1998 to 2001 showed that these isolates were grouped into two genotypes, H and I. While genotype I was predominant throughout the country during this period, genotype H was found in the restricted region, 1999. The nucleotide and deduced amino acid sequences of Korean isolates showed the type-specific changes including the signature motif at positions 28-30 in the SH gene and the neutralizing epitopes in the HN gene. Particularly, Asian strains including Korean isolates and European strains differed from 2.3 to 3.8% at the nucleotide sequence level in the SH gene although they belonged to the same genotype H. Furthermore, none of Korean isolates were genetically related to the vaccine strains used in Korea. The results provide important information to understand the epidemiology of mumps infection and to facilitate the development of more efficient vaccine program in Korea.