Enhancing the vision-language foundation model with key semantic knowledge-emphasized report refinement.

Recently, vision-language representation learning has made remarkable advancements in building up medical foundation models, holding immense potential for transforming the landscape of clinical research and medical care. The underlying hypothesis is that the rich knowledge embedded in radiology reports can effectively assist and guide the learning process, reducing the need for additional labels. However, these reports tend to be complex and sometimes even consist of redundant descriptions that make the representation learning too challenging to capture the key semantic information. This paper develops a novel iterative vision-language representation learning framework by proposing a key semantic knowledge-emphasized report refinement method. Particularly, raw radiology reports are refined to highlight the key information according to a constructed clinical dictionary and two model-optimized knowledge-enhancement metrics. The iterative framework is designed to progressively learn, starting from gaining a general understanding of the patient's condition based on raw reports and gradually refines and extracts critical information essential to the fine-grained analysis tasks. The effectiveness of the proposed framework is validated on various downstream medical image analysis tasks, including disease classification, region-of-interest segmentation, and phrase grounding. Our framework surpasses seven state-of-the-art methods in both fine-tuning and zero-shot settings, demonstrating its encouraging potential for different clinical applications.

Highly Stable Silicon Anode Enabled by a Water-Soluble Tannic Acid Functionalized Dual-Network Binder.

Silicon (Si), a high-capacity electrode material, is crucial for achieving high-energy-density lithium-ion batteries. However, Si suffers from poor cycling stability due to its significant volume changes during operation. In this work, a tannic acid functionalized aqueous dual-network binder with an intramolecular tannic acid functionalized network has been synthesized, which is composed of covalent-cross-linked polyamide and ionic-cross-linked alginate (Alg(Ni)-PAM-TA), and employed as an advanced binder for stabilizing Si anodes. The resultant Alg(Ni)-PAM-TA binder, incorporating diverse functional groups including amide, carboxylic acid, and dynamic hydrogen bonds, can easily interact with both Si nanoparticles and the Cu foil, thereby facilitating the formation of a highly resilient network characterized by exceptional adhesion strength. Moreover, molecular dynamics (MD) simulations indicate that the Alg(Ni)-PAM-TA network shows an increased intramolecular hydrogen bond number with increasing concentration of TA and a decreased intramolecular hydrogen bond between PAM and Alg as a result of the aggregation behavior of tannic acids themselves. Consequently, the binder significantly enhances the Si electrode's integrity throughout repeated charge/discharge cycles. At a current density of 0.84 A g-1, the Si electrode retains a capacity of 1863.4 mAh g-1 after 200 cycles. This aqueous binder functionalized with the intramolecular network via the incorporation of TA molecules holds great promise for the development of high-energy-density lithium-ion batteries.