We studied by questionnaire 530 subjects with chronic myeloid leukaemia (CML) in Hubei Province during the recent SARS-CoV-2 epidemic. Five developed confirmed (N = 4) or probable COVID-19 (N = 1). Prevalence of COVID-19 in our subjects, 0.9% (95% Confidence Interval, 0.1, 1.8%) was ninefold higher than 0.1% (0, 0.12%) reported in normals but lower than 10% (6, 17%) reported in hospitalised persons with other haematological cancers or normal health-care providers, 7% (4, 12%). Co-variates associated with an increased risk of developing COVID-19 amongst persons with CML were exposure to someone infected with SARS-CoV-2 (P = 0.037), no complete haematologic response (P = 0.003) and comorbidity(ies) (P = 0.024). There was also an increased risk of developing COVID-19 in subjects in advanced phase CML (P = 0.004) even when they achieved a complete cytogenetic response or major molecular response at the time of exposure to SARS-CoV-2. 1 of 21 subjects receiving 3rd generation tyrosine kinase-inhibitor (TKI) developed COVID-19 versus 3 of 346 subjects receiving imatinib versus 0 of 162 subjects receiving 2nd generation TKIs (P = 0.096). Other co-variates such as age and TKI-therapy duration were not significantly associated with an increased risk of developing COVID-19. Persons with these risk factors may benefit from increased surveillance of SARS-CoV-2 infection and possible protective isolation.

No abstract

This paper presents an effective image retrieval method by combining high-level features from convolutional neural network (CNN) model and low-level features from dot-diffused block truncation coding (DDBTC). The low-level features, e.g., texture and color, are constructed by vector quantization -indexed histogram from DDBTC bitmap, maximum, and minimum quantizers. Conversely, high-level features from CNN can effectively capture human perception. With the fusion of the DDBTC and CNN features, the extended deep learning two-layer codebook features is generated using the proposed two-layer codebook, dimension reduction, and similarity reweighting to improve the overall retrieval rate. Two metrics, average precision rate and average recall rate (ARR), are employed to examine various data sets. As documented in the experimental results, the proposed schemes can achieve superior performance compared with the state-of-the-art methods with either low-or high-level features in terms of the retrieval rate. Thus, it can be a strong candidate for various image retrieval related applications.

No abstract