Various retinal vessel segmentation methods based on convolutional neural networks were proposed recently, and Dense U-net as a new semantic segmentation network was successfully applied to scene segmentation. Retinal vessel is tiny, and the features of retinal vessel can be learned effectively by the patch-based learning strategy. In this study, we proposed a new retinal vessel segmentation framework based on Dense U-net and the patch-based learning strategy. In the process of training, training patches were obtained by random extraction strategy, Dense U-net was adopted as a training network, and random transformation was used as a data augmentation strategy. In the process of testing, test images were divided into image patches, test patches were predicted by training model, and the segmentation result can be reconstructed by overlapping-patches sequential reconstruction strategy. This proposed method was applied to public datasets DRIVE and STARE, and retinal vessel segmentation was performed. Sensitivity (Se), specificity (Sp), accuracy (Acc), and area under each curve (AUC) were adopted as evaluation metrics to verify the effectiveness of proposed method. Compared with state-of-the-art methods including the unsupervised, supervised, and convolutional neural network (CNN) methods, the result demonstrated that our approach is competitive in these evaluation metrics. This method can obtain a better segmentation result than specialists, and has clinical application value.

The integration of flexible materials with optical sensing technologies has advanced wearable optical biosensors, offering significant potential in personalized medicine, health monitoring, and disease prevention. This review summarizes the recent advancements in flexible materials for wearable optical biosensors, with a focus on materials such as polymer substrates, nanostructured materials, MXenes, hydrogels, and textile-based integrated platforms. These materials enhance the functionality, sensitivity, and adaptability of sensors, particularly in wearable applications. The review also explores various optical sensing mechanisms, including surface plasmon resonance (SPR), optical fiber sensing, fluorescence sensing, chemiluminescence, and surface-enhanced Raman spectroscopy (SERS), emphasizing their role in improving the detection capabilities for biomarkers, physiological parameters, and environmental pollutants. Despite significant advancements, critical challenges remain in the fabrication and practical deployment of flexible optical biosensors, particularly regarding the long-term stability of materials under dynamic environments, maintaining reliable biocompatibility during prolonged skin contact, and minimizing signal interference caused by motion artifacts and environmental fluctuations. Addressing these issues is vital to ensure robustness and accuracy in real-world applications. Looking forward, future research should emphasize the development of multifunctional and miniaturized devices, the integration of wireless communication and intelligent data analytics, and the improvement of environmental resilience. Such innovations are expected to accelerate the transition of flexible optical biosensors from laboratory research to practical clinical and consumer healthcare applications, paving the way for intelligent health management and early disease diagnostics. Overall, flexible optical biosensors hold great promise in personalized health management, early disease diagnosis, and continuous physiological monitoring, with the potential to revolutionize the healthcare sector.

Hydrogen peroxide (H₂O₂) is a signaling molecule that has the capacity to control a variety of biological processes in organisms. Cancer cells release more H₂O₂ during abnormal tumor growth. There has been a considerable amount of interest in utilizing H₂O₂ as a biomarker for the diagnosis of cancer tissue. In this study, an electrochemical sensor for H₂O₂ was constructed based on 3D reduced graphene oxide (rGO), MXene (Ti₃C₂), and multi-walled carbon nanotubes (MWCNTs) composite. Three-dimensional (3D) rGO–Ti₃C₂–MWCNTs sensor showed good linearity for H₂O₂ in the ranges of 1–60 μM and 60 μM–9.77 mM at a working potential of −0.25 V, with sensitivities of 235.2 µA mM⁻¹ cm⁻² and 103.8 µA mM⁻¹ cm⁻², respectively, and a detection limit of 0.3 µM (S/N = 3). The sensor exhibited long-term stability, good repeatability, and outstanding immunity to interference. In addition, the modified electrode was employed to detect real-time H₂O₂ release from cancer cells and cancer tissue ex vivo.

Dopamine (DA) and uric acid (UA) are essential for many physiological processes in the human body. Abnormal levels of DA and UA can lead to multiple diseases, such as Parkinson’s disease and gout. In this work, a three-dimensional reduced graphene oxide–MXene (3D rGO-Ti₃C₂) composite electrode was prepared using a simple one-step hydrothermal reduction process, which could separate the oxidation potentials of DA and UA, enabling the simultaneous detection of DA and UA. The 3D rGO-Ti₃C₂ electrode exhibited excellent electrocatalytic activity towards both DA and UA. In 0.01 M PBS solution, the linear range of DA was 0.5–500 µM with a sensitivity of 0.74 µA·µM⁻¹·cm⁻² and a detection limit of 0.056 µM (S/N = 3), while the linear range of UA was 0.5–60 µM and 80–450 µM, with sensitivity of 2.96 and 0.81 µA·µM⁻¹·cm⁻², respectively, and a detection limit of 0.086 µM (S/N = 3). In 10% fetal bovine serum (FBS) solution, the linear range of DA was 0.5–500 µM with a sensitivity of 0.41 µA·µM⁻¹·cm⁻² and a detection limit of 0.091 µM (S/N = 3). The linear range of UA was 2–500 µM with a sensitivity of 0.11 µA·µM⁻¹·cm⁻² and a detection limit of 0.6 µM (S/N = 3). The modified electrode exhibited advantages such as high sensitivity, a strong anti-interference capability, and good repeatability. Furthermore, the modified electrode was successfully used for DA measurement in vivo. This could present a simple reliable route for neurotransmitter detection in neuroscience.

An enzyme-based amperometric biosensor was fabricated for detecting hydrogen peroxide (H₂O₂). Horseradish peroxidase (HRP) was modified using functionalized fluorescent gold nanoclusters (AuNCs) via biomineralization. HRP-AuNCs were successfully immobilized on multiwalled carbon nanotube- (MWCNT-) coated carbon fiber ultramicroelectrodes (CFUMEs). The AuNCs, which act as molecular electric wires, effectively promote the electron transfer between the enzyme active center and the electrode. Additionally, the HRP conjugated with the AuNCs retains its biological activity, which enables the catalytic reaction of H₂O₂. The HRP-AuNCs/MWCNTs/CFUMEs have been proven as excellent amperometric sensors for H₂O₂. The sensitivity of the H₂O₂ biosensor is 3.0×10−4A/M, and the detection limit is estimated to be 443 nM. Furthermore, the biosensor exhibited long-term stability and good reproducibility. Moreover, the biosensor has been tested by determining the H₂O₂ concentration in calf serum samples.