A pH/Viscosity Dual-Responsive Fluorescent Probe with Machine Learning Enables Quantitative Prediction of Thrombus Age.

Abstract

Accurate determination of thrombus age is critical in determining the appropriate treatment strategies. In cases of anastomotic site thrombosis, the decision between thrombolytic therapy and flap replacement surgery depends on the time since thrombus formation. However, clinically applicable methods for reliably estimating the time since thrombus formation remain lacking. This study aims to develop a reliable strategy to accurately determine the thrombus age. Specifically, a reliable rat flap model for slow-progression thrombosis has been successfully established. The vascular obstruction state remains stable within 20 min after stimulation. As the thrombus develops, it is accompanied by an increase in local viscosity and acidity. We have designed and synthesized a probe (Cy). It exhibits dual responses to viscosity and pH value. In this animal model, Cy is able to precisely illuminate the thrombus area. Its fluorescence intensity shows a significant correlation with the thrombus age. The logistic regression model based on fluorescence intensity could effectively identify thrombi older than 6 h in vitro. To our knowledge, this is the first study to quantitatively predict thrombus age using a fluorescent probe combined with machine learning. This strategy also shows great potential for guiding personalized treatment for thrombotic diseases.