Objective To investigate the molecular characteristics and epidemiological signification of patients with low-level HBsAg. Methods PCR and gene sequencing were used to detect HBV DNA and Tyr-Met-Asp-Asp(YMDD) mutant in 136 serum samples with low-level HBsAg and 44 sernm samples with high-level HBsAg. Genotyping was performed in 47 cases with HBV DNA 10~5 copies/L by concentration method and 37 cases with high-level HBsAg. S gene sequences and serotypes were analyzed in 14 cases with HBV DNA 105 copies/L and 29 cases with high-level HBsAg. S gene sequences were compared with the consensus sequence of Chinese strain by BioEdit software. Results The HBV DNA-positive rate, YMDD mutation rate and HBV DNA load (logarithm) in low-level and high-level HBsAg group were 34.6% (47/136), 0% (0/136), 6.5±1.4 and 84.1% (37/44), 9.1% (4/44), 8.9±1.8, respectively. There was statistically significant differences between two groups (for concentration method,χ~2 = 30.8, P < 0.05; for direct method, χ~2 = 53.5, P < 0.05; for YMDD mutation ratio, P = 0.003, For HBV DNA (log), t = 6.5, P < 0.05). The genotypes in low-level HBsAg group included type B (16/47), type C (5/47) and non-classified ones(26/47). There were significant differences between two groups (χ~2=21.8, P <0.01). The serotypss included adw (7/14), ayw (4/14), adr (2/14) and ayr (1/14). There were significant differences in genotypes (χ~2 = 13.5, P < 0.05) but not in serotypes between two groups (χ~2 = 4.7, P >0.05). S gene sequencing results showed no S gnne variation was detected, but there were 6 single nucleotide polymorphisms in 16 cases, which would not result in the alternation of amino acid. Conclusions Low-replication phenomenon of HBV DNA was present in patients with low-level HBsAg. The major genotyps and serotype was type B and adw/ayw, respectively. Polymorphic variants have been found in the S gene. The existence of low-level HBsAg might be related with its own molecular characteristics resulting in low expression of HBsAg or immune tolerance induced by low-level HBsAg after HBV infection.

Electrocardiographic (ECG) signal are a kind of basic physiological signals of human body, and are very important in clinical diagnosis. But the ECG signals from body surface are often interfered by noises such as 50 Hz noise, baseline displacemant, electromyography (EMG) noise and edv. These noises bring obstacle to the diagnosis of cardiovascular diseases. To eliminate the ECG signals noises mentioned above,this paper adopts LMS adaptive algorithm and wavelet transform theory to design three kinds of digital adaptive filters-adaptive noise cancellation filter, wavelet transform filter and adaptive signal dividing filter to filter the corresponding noises. The results show that the three kinds of noises existing in the ECG signal have been efficiently eliminated.
Algorithms ; Electrocardiography ; Fourier Analysis ; Humans ; Signal Processing, Computer-Assisted ; instrumentation ; Stochastic Processes ; Time Factors