Enhancing cross-subject EEG emotion recognition through multi-source manifold metric transfer learning.

Transfer learning (TL) has demonstrated its efficacy in addressing the cross-subject domain adaptation challenges in affective brain-computer interfaces (aBCI). However, previous TL methods usually use a stationary distance, such as Euclidean distance, to quantify the distribution dissimilarity between two domains, overlooking the inherent links among similar samples, potentially leading to suboptimal feature mapping. In this study, we introduced a novel algorithm called multi-source manifold metric transfer learning (MSMMTL) to enhance the efficacy of conventional TL. Specifically, we first selected the source domain based on Mahalanobis distance to enhance the quality of the source domains and then used manifold feature mapping approach to map the source and target domains on the Grassmann manifold to mitigate data drift between domains. In this newly established shared space, we optimized the Mahalanobis metric by maximizing the inter-class distances while minimizing the intra-class distances in the target domain. Recognizing that significant distribution discrepancies might persist across different domains even on the manifold, to ensure similar distributions between the source and target domains, we further imposed constraints on both domains under the Mahalanobis metric. This approach aims to reduce distributional disparities and enhance the electroencephalogram (EEG) emotion recognition performance. In cross-subject experiments, the MSMMTL model exhibits average classification accuracies of 88.83 % and 65.04 % for SEED and DEAP, respectively, underscoring the superiority of our proposed MSMMTL over other state-of-the-art methods. MSMMTL can effectively solve the problem of individual differences in EEG-based affective computing.

Alfosbuvir plus Daclatasvir for Treatment of Chronic Hepatitis C Virus Infection in China.

INTRODUCTION: A pan-genotypic and effective treatment regimen for patients with chronic hepatitis C virus (HCV) infection remains an unmet medical need in China. Alfosbuvir is a novel potent HCV NS5B polymerase inhibitor in development for the treatment of chronic HCV infection. We conducted a phase 3 study to evaluate the efficacy and safety of alfosbuvir in combination with daclatasvir in Chinese patients with HCV infection. METHODS: All patients received 600 mg alfosbuvir tablets plus 60 mg daclatasvir tablets once daily for 12 weeks. The primary endpoint was sustained virological response 12 weeks after the end of treatment (SVR12). A follow-up visit was done at week 4 and 12, and those who achieved SVR12 were followed up at post-treatment week 24. RESULTS: Of the 326 patients who received at least one dose of the study drug, 320 (98.2% [95% confidence interval (CI): 96.5%-99.5%]) achieved sustained virological response at post-treatment week 12 (SVR12), which was superior to the historical SVR12 rate of 88% (p < 0.0001). The SVR12 rates were similar regardless of most baseline characteristics. The most common adverse event (AE) (≥ 10%) was hypercholesterolemia. Serious adverse events (SAEs) were reported in 25 (7.7%) patients, none of which was judged to be related to the study drug. The majority of AEs were mild to moderate in severity. CONCLUSIONS: Alfosbuvir plus daclatasvir for 12 weeks was highly effective and safe in Chinese patients infected with HCV genotype 1, 2, 3, or 6, suggesting that this regimen could be a promising option for HCV treatment in China irrespective of genotype. TRIAL REGISTRATION: ClinicalTrial.gov identifier, NCT04070235.

Cross-subject EEG emotion recognition using multi-source domain manifold feature selection.

Recent researches on emotion recognition suggests that domain adaptation, a form of transfer learning, has the capability to solve the cross-subject problem in Affective brain-computer interface (aBCI) field. However, traditional domain adaptation methods perform single to single domain transfer or simply merge different source domains into a larger domain to realize the transfer of knowledge, resulting in negative transfer. In this study, a multi-source transfer learning framework was proposed to promote the performance of multi-source electroencephalogram (EEG) emotion recognition. The method first used the data distribution similarity ranking (DDSA) method to select the appropriate source domain for each target domain off-line, and reduced data drift between domains through manifold feature mapping on Grassmann manifold. Meanwhile, the minimum redundancy maximum correlation algorithm (mRMR) was employed to select more representative manifold features and minimized the conditional distribution and marginal distribution of the manifold features, and then learned the domain-invariant classifier by summarizing structural risk minimization (SRM). Finally, the weighted fusion criterion was applied to further improve recognition performance. We compared our method with several state-of-the-art domain adaptation techniques using the SEED and DEAP dataset. Results showed that, compared with the conventional MEDA algorithm, the recognition accuracy of our proposed algorithm on SEED and DEAP dataset were improved by 6.74% and 5.34%, respectively. Besides, compared with TCA, JDA, and other state-of-the-art algorithms, the performance of our proposed method was also improved with the best average accuracy of 86.59% on SEED and 64.40% on DEAP. Our results demonstrated that the proposed multi-source transfer learning framework is more effective and feasible than other state-of-the-art methods in recognizing different emotions by solving the cross-subject problem.

Safety, efficacy, and pharmacokinetics of SH-1028 in EGFR T790M-positive advanced non-small cell lung cancer patients: A dose-escalation phase 1 study.

BACKGROUND: SH-1028 is a new third-generation EGFR tyrosine kinase inhibitors (TKI) to benefit patients with EGFR T790M-mutated NSCLC. Here, the authors report its clinical safety, preliminary efficacy, and pharmacokinetic (PK) profile for the first time. METHODS: Patients with EGFR T790M mutation, locally advanced non-small cell lung cancer (NSCLC), or metastatic NSCLC who had progressed after previous EGFR TKI therapy were eligible. Patients received SH-1028 at five oral dose levels (60 mg, 100 mg, 200 mg, 300 mg, and 400 mg) once daily until disease progression, unacceptable toxicity, or patient withdrawal. The primary end points were the safety, dose-limiting toxicity (DLT), maximum-tolerated dose (MTD), and PK profile. Secondary end points included objective response rate (ORR), disease control rate (DCR), progression-free survival (PFS), etc. RESULTS: Data cut off on December 31, 2020, a total of 20 patients were enrolled during the trial, two of three patients in 300 mg cohort experienced a DLT, and no DLT was observed in 240 mg cohort, 240 mg was determined to be the MTD of SH-1028. A total of 95.0% (19 of 20) of patients reported treatment-related adverse events (TRAEs), and the incidence of serious adverse events was 20.0% (4 of 20). The ORR and DCR of the 200 mg cohort were 75% (95% confidence interval [CI], 19.41-99.37) and 75.0% (95% CI, 19.41-99.37), respectively. The overall ORR was 40% (95% CI, 19.12-63.95), and DCR was 70.0% (95% CI, 45.72-88.11). According to the PK profile, the dosage regimen for future studies was determined as 200 mg once daily. CONCLUSIONS: SH-1028 showed a manageable safety and promising antitumor activity in patients with EGFR T790M mutation at the dose of 200 mg once daily. PLAIN LANGUAGE SUMMARY: Lung cancer has a high morbidity and mortality, with an estimated 1.8 million deaths in 2020. Non-small cell lung cancer accounts for approximately 85% of lung cancer. First- or second-generation EGFR TKIs' weak selectivity often led to the occurrence of treatment-related adverse events, such as interstitial lung disease, rash, diarrhea, etc., along with acquired drug resistance within approximately 1 year. A dose of 200 mg of SH-1028 once daily showed a preliminary antitumor activity with manageable safety in patients with EGFR T790M mutation.

The lncRNA PVT1/miR-423-5p axis is positively correlated with the severity of bronchial asthma.

Bronchial asthma (BA) is a serious problem affecting the quality of life of patients. Long noncoding RNAs (lncRNAs) are involved in BA. This study set out to investigate expressions of PVT1/miR-423-5p in the serum of BA patients and its clinical value on BA diagnosis and evaluation. This study included the same number (N = 100) of patients with BA at remission (BA-R), BA at exacerbation (BA-E), and healthy controls. PVT1 level was increased in BA-R and BA-E patients, and PVT1 level was higher in BA-E patients than BA-R patients. miR-141-3p targeted PVT1. miR-423-5p was downregulated in the serum of BA patients and was negatively correlated with PVT1. Area under ROC curve of PVT1/miR-423-5p axis on BA-R patients was 0.837 with sensitivity 0.74, specificity 0.84, while that of BA-E was 0.974 with sensitivity 0.87 and specificity 0.95. PVT1/miR-423-5p axis was negatively correlated with FEV1/FVC, FEV1% pred, and IL-10, and positively correlated with IgE, TNF-α, and IL-6. PVT1 and PVT1/miR-423-5p axis were associated with increased severity while miR-423-5p axis was negatively associated with BA severity. In conclusion, increased levels of PVT1/miR-423-5p had higher diagnostic efficiency on BA patients, especially patients with acute exacerbation.