The objective of this study was to investigate the feasibility of ultrasonic monitoring for the transient depth-dependent osmotic swelling and solute diffusion in normal and degenerated articular cartilage (artC) tissues. Full-thickness artC specimens were collected from fresh bovine patellae. The artC specimens were continuously monitored using a focused beam of 50 MHz ultrasound (US) during sequential changes of the bathing solution from 0.15 mol/L to 2 mol/L saline, 0.15 mol/L saline, 1 mg/mL trypsin solution, 0.15 mol/L saline, 2 mol/L saline and back to 0.15 mol/L saline. The transient displacements of US echoes from the artC tissues at different depths were used to represent the tissue deformation and the NaCl diffusion. The trypsin solution was used selectively to digest the proteoglycans in artC. It was demonstrated that high-frequency US was feasible for monitoring the transient osmotic swelling, solute transport and progressive degeneration of artC in real-time. Preliminary results showed that the normal bovine patellar artC shrank during the first several minutes and then recovered to its original state in approximately 1 h when the solution was changed from 0.15 mol/L to 2 mol/L saline. Degenerated artC showed neither shrinkage nor recovery during the same process. In addition, a dehydrated-hydrated artC specimen showed much stronger shrinkage and it resumed the original state when the solution was changed from 2 mol/L back to 0.15 mol/L saline. The diffusion of NaCl and the digestion process of proteoglycans induced by trypsin were also successfully monitored in real-time.
Accurate vessel segmentation is essential for diagnosing and managing vascular and ophthalmic diseases. Traditional learning-based vessel segmentation methods heavily rely on high-quality, pixel-level annotated datasets. However, segmentation performance suffers significantly when applied in federated learning settings due to vessel morphology inconsistency and vessel-background imbalance. The former limits the ability of models to capture fine-grained vessels, while the latter overemphasizes background pixels and biases the model towards them. To address these challenges, we propose a novel method named Federated Vessel-Aware Calibration (FVAC), which leverages global uncertainty to provide differentiated guidance for clients, focusing on pixels of various morphologies that are difficult to distinguish. Furthermore, we introduce a foreground-background decoupling alignment strategy that utilizes more stable and balanced global features to mitigate semantic drift caused by vessel-background imbalance in local clients. Comprehensive experiments confirm the effectiveness of our method