Time series forecasting is crucial in many fields, yet current deep learning models struggle with noise, data sparsity, and capturing complex multi-scale patterns. This paper presents MFF-FTNet, a novel framework addressing these challenges by combining contrastive learning with multi-scale feature extraction across both frequency and time domains. MFF-FTNet introduces an adaptive noise augmentation strategy that adjusts scaling and shifting factors based on the statistical properties of the original time series data, enhancing model resilience to noise. The architecture is built around two complementary modules: a Frequency-Aware Contrastive Module (FACM) that refines spectral representations through frequency selection and contrastive learning, and a Complementary Time Domain Contrastive Module (CTCM) that captures both short- and long-term dependencies using multi-scale convolutions and feature fusion. A unified feature representation strategy enables robust contrastive learning across domains, creating an enriched framework for accurate forecasting. Extensive experiments on five real-world datasets demonstrate that MFF-FTNet significantly outperforms state-of-the-art models, achieving a 7.7% MSE improvement on multivariate tasks. These findings underscore MFF-FTNet's effectiveness in modeling complex temporal patterns and managing noise and sparsity, providing a comprehensive solution for both long- and short-term forecasting.
Numerous studies have suggested that the hemoglobin-to-red blood cell distribution width ratio (HRR) is associated with the onset, progression, and prognosis of various diseases. However, to the best of our knowledge, no research has conducted statistical analyses to determine the association between HRR and stroke. This cross-sectional study was conducted among adults with complete data on hemoglobin-red cell distribution width ratio (HRR) and stroke from the 2005–2018 National Health and Nutrition Examination Survey (NHANES). HRR was calculated by dividing hemoglobin (HGB) in grams per deciliter (g/dL) by the red blood cell distribution width (RDW). Weighted multivariable logistic regression and generalized additive models were employed to investigate the independent and nonlinear relationships between HRR and stroke. Threshold effects were assessed using two-piece linear regression models. Additionally, subgroup analyses and interaction tests were performed. A total of 36,215 participants were included, with 2.92% classified as stroke patients. The prevalence of stroke decreased across increasing tertiles of HRR (Q1: 5.07%; Q2: 2.63%; Q3: 1.69%; P < 0.0001). A negative association between HRR and stroke was observed in both unadjusted and adjusted models. In Model III, each one-unit increase in HRR was associated with a 58% reduction in the likelihood of stroke (Model III: OR = 0.42, 95% CI: 0.29–0.63). Subgroup analyses and interaction tests revealed that the association between HRR and stroke was BMI-dependent (P < 0.05), with the negative association only observed in participants with BMI < = 25 and BMI 25–30, but not in those with BMI > 30. Using a two-piece linear regression model, a threshold effect was identified at a BMI of 30 (K = 1.16). To the left of this breakpoint, there was a negative association between HRR and stroke (OR = 0.55, 95% CI: 0.27–0.97), whereas no such association was detected to the right of the breakpoint (OR = 6.49, 95% CI: 0.75–56.11). HRR is negatively associated with the likelihood of stroke, with a lower risk of stroke in individuals with higher HRR levels.
The major molecular events in the genesis of most breast cancers are unknown. However, human papillomaviruses (HPV) have been reported to be found in a significant portion of breast cancers of women with concomitant cervical intraepithelial neoplasia III. To investigate a potential HPV-breast cancer link, we carried out a small survey to identify HPV in unselected, general breast cancer tissues.Deoxyribonucleic acid (DNA) was isolated from 17 breast cancer tissues (and one cervical swab) taken from our local, randomly selected patient population. Two different previously characterized broad-spectrum primer sets (targeting the E6/E7 or L1 regions) were used to amplify HPV DNA, and another primer set was used to amplify the ColE1/pBR322 origin of replication by polymerase chain reaction amplification. The polymerase chain reaction product DNA was analyzed by dot blot hybridization with HPV-16, -18, -31, or pRB322 DNA probes. Total cellular DNA was also analyzed by one- and two-dimensional Southern blot analysis. Finally, the E6/E7 polymerase chain reaction products were cloned, sequenced, and compared to previously cloned HPV types.Polymerase chain reaction/dot blot analysis by both the HPV E6-E7 and L1 primer sets identified the same 6 out of 17 (35%) breast cancers as being HPV positive. ColE1/pBR322 origin targeted polymerase chain reaction/dot blot analysis failed to identify plasmid contamination. One- and two-dimensional Southern blot analysis showed that the breast cancers specimens contained significant levels of HPV DNA and that the viral DNA was largely episomal. The sequences of the HPV clones demonstrated that HPV-16, -18, and possibly type 11 were present within the breast cancer specimens. Furthermore, the HPV sequences cloned from the cervical swab and breast cancer of the same patient were found to be identical.These data suggest that HPV may be associated with a significant subset of breast cancers, and further suggest that additional studies are warranted.
The radioprotective effects of dimethyl sulfoxide (DMSO) have been known for many years, and the suppression of hydroxyl (OH) radicals induced by ionizing radiation has been thought to be the main cause of this effect. However, the DMSO concentration used was very high, and might be toxic, in earlier studies. In the present study, we administered a lower, non-toxic concentration (0.5%, i.e., 64 mM) of DMSO before irradiation and examined its radioprotective effects. Colony formation assay and micronucleus assay showed significant radioprotective effects in CHO, but not in xrs5, which is defective in the repair function of DNA double-strand breaks. The levels of phosphorylated H2AX and the formation of 53BP1 foci 15 minutes after irradiation, which might reflect initial DNA double-strand breaks, in DMSO-treated CHO cells were similar to those in non-treated cells, suggesting that the radioprotective effects were not attributable to the suppression of general indirect action in the lower concentration of DMSO. On the other hand, 2 hours after irradiation, the average number of 53BP1 foci, which might reflect residual DNA double-strand breaks, was significantly decreased in DMSO-treated CHO cells compared to non-treated cells. The results indicated that low concentration of DMSO exerts radioprotective effects through the facilitation of DNA double-strand break repair rather than through the suppression of indirect action.
Two species of marine microalgae, Isochrysis galbana 8701 and Skeletonema costatum, were exposed to low concentration of anthracene, and their cell density, contents of chlorophyll a, carotinoid and protein, and superoxide dismutase (SOD) activity were examined. It was showed that low concentration of anthracene (1.5-6.0 micrograms.L) could obviously stimulate the growth of these microalgae, and their protein, chlorophyll a, and carotinoid contents increased with increasing cell density. SOD of treated groups remained high activity, compared with the controlled group during the whole experiment.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)