Evaluation of the Lyophilized Platelets Function by Thromboelastography in an In Vitro Platelet Transfusion Model.

BACKGROUND: At present, liquid conserved platelets (PLTs) can only be stored at 22°C for up to 5 days. Waste of outdated PLTs and short supply of fresh PLTs were both seen in blood banks. Lyophilized PLTs were considered as one of the candidate replacements for liquid conserved PLTs. It is important to evaluate the function before it can be used in clinical trial. METHODS: In this study, an in vitro platelet transfusion model was established to evaluate the function of rehydrated lyophilized platelets (RLPs) by thromboelastography (TEG). Blood samples from 11 patients were spiked with 3 preparations of specific donors' apheresis platelets (stored at room temperature, frozen, and lyophilized) to an increment equivalent to transfusion with 3x10^11 platelets. Whole blood TEG assay was performed and the maximum amplitude (MA) value was used to evaluate the function of "transfused" platelets. RESULTS: The recovery of rehydrated lyophilized platelets (RLPs) in our study was 81.38% ± 2.38, and mean platelet volume (MPV) was 9.02 ± 0.54 fL. MA was significantly enhanced in the three different groups after the addition of PLTs when compared with the whole blood (WB) group. CONCLUSIONS: Our findings suggest that RLPs are capable of enhancing the MA value as well as fresh and frozen PLTs in vitro. The clinical significance of this remains to be determined.

Adaptive soft sensor using stacking approximate kernel based BLS for batch processes.

To deal with the highly nonlinear and time-varying characteristics of Batch Process, a model named adaptive stacking approximate kernel based broad learning system is proposed in this paper. This model innovatively introduces the approximate kernel based broad learning system (AKBLS) algorithm and the Adaptive Stacking framework, giving it strong nonlinear fitting ability, excellent generalization ability, and adaptive ability. The Broad Learning System (BLS) is known for its shorter training time for effective nonlinear processing, but the uncertainty brought by its double random mapping results in poor resistance to noisy data and unpredictable impact on performance. To address this issue, this paper proposes an AKBLS algorithm that reduces uncertainty, eliminates redundant features, and improves prediction accuracy by projecting feature nodes into the kernel space. It also significantly reduces the computation time of the kernel matrix by searching for approximate kernels to enhance its ability in industrial online applications. Extensive comparative experiments on various public datasets of different sizes validate this. The Adaptive Stacking framework utilizes the Stacking ensemble learning method, which integrates predictions from multiple AKBLS models using a meta-learner to improve generalization. Additionally, by employing the moving window method-where a fixed-length window slides through the database over time-the model gains adaptive ability, allowing it to better respond to gradual changes in industrial Batch Process. Experiments on a substantial dataset of penicillin simulations demonstrate that the proposed model significantly improves predictive accuracy compared to other common algorithms.