Biomedical imaging plays a critical role in medical diagnostics and research, yet image noise remains a significant challenge that hinders accurate analysis. To address this issue, we propose BDNet, a real-time biomedical image denoising network optimized for enhancing gradient and high-frequency information while effectively suppressing noise. The network adopts a lightweight U-Net-inspired encoder–decoder architecture, incorporating a Convolutional Block Attention Module at the bottleneck to refine spatial and channel-wise feature extraction. A novel gradient-based loss function—combining Sobel operator-derived gradient loss with L1, L2, and LSSIM losses—ensures faithful preservation of fine structural details. Extensive experiments on the Fluorescence Microscopy Denoising (FMD) dataset demonstrate that BDNet achieves state-of-the-art performance across multiple metrics, including PSNR, RMSE, SSIM, and LPIPS, outperforming both convolutional and Transformer-based models in accuracy and efficiency. With its superior denoising capability and real-time inference speed, BDNet provides an effective and practical solution for improving biomedical image quality, particularly in fluorescence microscopy applications.

As one of the leading cancers threatening women’s lives and health, breast cancer is challenging to treat and often irreversible in advanced cases, highlighting the critical importance of early detection and intervention. In recent years, fluorescent probe technology, a revolutionary in vivo imaging tool, has gained attention in medical research for its ability to improve tumor visualization significantly. This review focuses on recent advances in intelligent, responsive fluorescent probes, particularly in the field of breast cancer, which are divided into five categories, near-infrared responsive, fluorescein-labeled, pH-responsive, redox-dependent, and enzyme-triggered fluorescent probes, each of which has a different value for application based on its unique biological response mechanism. In addition, this review also covers the strategy of combining fluorescent probes with various anti-tumor drugs, aiming to reveal the possibility of synergistic effects between the two in breast cancer treatment and provide a solid theoretical platform for the clinical translation of fluorescent probe technology, which is expected to promote the expansion of cancer treatment technology.

Purposes: Osimertinib, one of the third-generation EGFR-tyrosine kinase inhibitors (TKIs) designed to target EGFR T790M mutation, significantly improves the prognosis of lung cancer. However, drug resistance still happens and MET amplification is responsible for one of the main causes. Fluorescence in situ hybridization (FISH) is the gold standard for MET amplification detection, but fundamentally limited by observer subjectivity. Herein, we assessed the value of next-generation sequencing (NGS) method in MET amplification detection in non-small cell lung cancer (NSCLC), as well as revealed the mutation profiling of NSCLC patients with osimertinib resistance to provide some valuable clues to the mechanisms of resistance.Methods: A total of 317 cancer tissue samples from 317 NSCLC patients at time of progression following osimertinib were submitted to NGS and only 96 tissues were tested by FISH simultaneously. With FISH results as gold standard, enumeration algorithm was applied to establish the optimal model for identifying MET amplification using gene copy number (GCN) data.Results: The optimal model for identifying MET amplification was constructed based on the GCN of MET, BRAF, CDK6 and CYP3A4, which achieved a 74.0% overall agreement with FISH and performed well in identifying MET amplification except polysomy with a sensitivity of 85.7% and a specificity of 93.9%. The inconsistency between NGS and FISH occurred mainly in polysomy subtype, while MET GCN ≥ 5 could be reliably recognized by NGS. Moreover, the most frequently mutated genes in NSCLC patients with osimertinib resistance were EGFR (59.94%), followed by TP53 (43.85%), NRG1 (9.46%), PIK3CA (6.31%), and ATM (5.36%). The known resistance mechanisms, including MET amplification, EGFR (C797S, L718Q/R), TP53, CDK4, CDK6, CDKN2A, BRAF, KRAS, NRAS and PIK3CA mutations were also disclosed in our cohort.Conclusions: NGS assay can achieve a high concordance with FISH in MET amplification detection and has advantages in portraying various genetic alterations, which is of worthy in clinical promotion.

Digital polymerase chain reaction (PCR) technology in microfluidic systems often results in bubble formation post-amplification, leading to microdroplet fragmentation and compromised detection accuracy. To solve this issue, this study introduces a method based on the constant pressure regulation of microdroplets during PCR within microfluidic chips. An ideal pressure reference value for continuous pressure control was produced by examining air solubility in water at various pressures and temperatures as well as modeling air saturation solubility against pressure for various temperature scenarios. Employing a high-efficiency constant pressure device facilitates precise modulation of the microfluidic chip’s inlet and outlet pressure. This ensures that air solubility remains unsaturated during PCR amplification, preventing bubble precipitation and maintaining microdroplet integrity. The device and chip were subsequently utilized for quantitative analysis of the human epidermal growth factor receptor (EGFR) exon 18 gene, with results indicating a strong linear relationship between detection signal and DNA concentration within a range of 10¹–10⁵ copies/μL (R² = 0.999). By thwarting bubble generation during PCR process, the constant pressure methodology enhances microdroplet stability and PCR efficiency, underscoring its significant potential for nucleic acid quantification and trace detection.

In addressing the detection of drug resistance in Helicobacter pylori, we have successfully developed an efficient and highly accurate detection methodology. Initially, we designed and fabricated a microarray chip, which underwent finite element analysis for its optical and thermal characteristics. Ultimately, COC material was chosen as the processing material for the chip, ensuring superior performance. Subsequently, we established a comprehensive detection system and validated its performance. Following that, comparative experiments were conducted for detecting drug resistance in H. pylori. The experimental results indicate that our established methodology aligns with the results obtained using the E-test detection kit, achieving a concordance rate of 100%. In comparison to the E-test detection kit, our methodology reduces the detection time to 1.5 h and provides a more extensive coverage of detection sites.

To overcome the limitations of repeated GH injection therapy, gelatin methacryloyl (GelMA)/hyaluronic acid (HA) composite hydrogel microneedles were fabricated for potential non-invasive transdermal delivery of recombinant human growth hormone (rhGH) and long-acting human growth hormone Fc fusion protein (hGH-Fc). All-atom molecular dynamics simulations combined with morphological, mechanical, and scanning electron microscopy (SEM) characterization were applied to screen optimal microneedle formulations, with Fourier transform infrared spectroscopy (FTIR) analysis and Parafilm M penetration tests verifying physicochemical compatibility and insertion capability. MD simulations revealed that rhGH stabilized the matrix via hydrogen bonds and electrostatic interactions, while hGH-Fc bound mainly through van der Waals forces. Through orthogonal screening of GelMA (5%, 10%, 15%) and HA (0.5%, 1%, 2%, 5%), the 10% GelMA with 0.5–2% HA was identified as the optimal range; the representative 10% GelMA/1% HA formulation exhibited a single-tip force exceeding 1.40 N and ~500 μm penetration in Parafilm M. This study offers a theoretical and technical basis for the design and optimization of hydrogel microneedle carriers for short- and long-acting GH delivery.