Sepsis is a critical condition. The key factor affecting the survival of patient is whether standard treatment can be obtained timely. Because of the complexity of its pathogenesis and high heterogeneity, there is no special diagnosis method currently. Early identification is difficult. Delayed diagnosis and treatment is closely related to the mortality of patients. With the continuous updating of the guidelines, sepsis has been included in the “time window” disease, putting forward a great challenge to the early screening and evaluation of sepsis. This article aims to review the application of Sepsis-Related Organ Failure Assessment, sepsis biomarkers and artificial intelligence algorithms in early screening and evaluation of sepsis, so as to provide guidance tools for timely starting standardized treatment of sepsis.
The precision diagnosis and treatment of chest wall tumors heavily rely on imaging assessment, which is limited by the subjectivity of conventional methods and their insensitivity to microscopic features. Artificial intelligence (AI) technologies, particularly radiomics and deep learning, offer new opportunities to overcome these bottlenecks. It is important to note, however, that direct AI research focused specifically on chest wall tumors remains scarce. Most current insights are derived from methodological borrowing and conceptual extrapolation of studies on other solid tumors, such as lung, breast, and esophageal cancers. This review systematically outlines the potential application framework of AI in the management of chest wall tumors, encompassing benign-malignant differentiation, non-invasive pathological subtyping, early treatment response prediction, and prognosis stratification based on multi-modal imaging. It discusses the core model-building strategies and validation processes centered on data fusion, and critically analyzes the unique challenges in this field, including data scarcity, model interpretability, and clinical translation. Moving forward, key directions for translating the potential of AI into clinical reality include fostering the development of dedicated chest wall tumor datasets, conducting prospective validation studies, and exploring "imaging-genomics" integration as well as dynamic decision-support systems.
ObjectiveTo assess the clinical value of liquid biopsy in breast cancer care, with a focus on its potential to complement tissue biopsy in diagnosis, treatment monitoring, and disease management. MethodA narrative review of recent literature was performed, examining the biological basis and clinical relevance of key liquid biopsy markers, including circulating tumor DNA (ctDNA), circulating tumor cells, and extracellular vesicles, in the context of breast cancer. ResultsAmong the markers reviewed, ctDNA can non-invasively detect mutations in key genes, such as ESR1, PIK3CA, HER2, and BRCA1/2, with results that are largely consistent with those from tissue samples. In addition to mutation testing, ctDNA can be measured repeatedly over time, making it useful for early screening, monitoring of high-risk individuals, efficacy evaluation, evolution of drug-resistant clones, and monitoring of minimal residual disease. Circulating tumor cells count and phenotypic analysis provide prognostic information regarding recurrence risk and survival, and single cell analysis helps reveal tumor spatiotemporal heterogeneity. Extracellular vesicles carry nucleic acids and proteins of tumor origin and have shown potential value in early diagnosis and disease monitoring. Liquid biopsy has shown clinical value in the diagnosis of advanced breast cancer, including cases with central nervous system involvement. However, current limitations in detection sensitivity and specificity, combined with the lack of standardized testing procedures, continue to restrict its clinical use. As a result, tissue biopsy remains the gold standard for initial pathological diagnosis and molecular classification. ConclusionsLiquid biopsy and tissue biopsy have different but complementary roles in breast cancer care. Tissue biopsy remains the basis for diagnosis and molecular subtyping, while liquid biopsy is more suitable for continuous monitoring and molecular stratified guidance over the course of disease. With advances in detection methods, combined use of multiple markers, improved standardization, and the integration of artificial intelligence for multi-omics data analysis, liquid biopsy holds promise for advancing more personalized and dynamic management for breast cancer patients.
