Yixin Wei

This study aims to develop a comprehensive nomogram to accurately predict differentiated thyroid cancer (DTC) involved lymph node, enabling the personalization of treatment plans and optimizing patient outcomes. The nomogram was developed through an analysis of data obtained from derivation cohorts. To identify independent predictors of involved lymph node, we utilized the Least Absolute Shrinkage and Selection Operator (LASSO) method in conjunction with multivariable logistic regression. These predictive factors were subsequently integrated into the nomogram's design. To ensure the tool's robustness and clinical applicability, we assessed its discriminative ability, calibration, and clinical utility in both the derivation cohort and an independent validation cohort. The identification of four key parameters - Thyroglobulin, Lymphocyte count, BRAF^V600E, and the number of tumor foci - emerged as robust predictors of involved lymph node. The nomogram demonstrated outstanding performance across calibration, discrimination, and clinical applicability, with predictive accuracies of 0.992 in the derivation cohort and 0.920 in the validation cohort. Peripheral lymphocyte counts have been found to correlate with involved lymph node in DTC patients. A nomogram tailored specifically for predicting involved lymph node in DTC guides personalized postoperative radioactive iodine (RAI) and plays a critical role in the preoperative assessment.

BACKGROUND AND PURPOSE: Exploring the directionality of neural information in the brain is important for understanding brain mechanisms and neurodisease development. Granger causality analysis and the metabolic connectivity map can be used to investigate directional transmission of information between brain regions, but their differences in depicting functional effective connectivity are not clear. MATERIALS AND METHODS: Using the Monash rs-PET/MR imaging data set, we conducted Granger causality and metabolic connectivity map analyses of the dopamine reward circuit in the brain. The dopamine reward circuit is a well-known system consisting primarily of the bilateral orbital frontal cortex, caudate, nucleus accumbens, thalamus, and substantia nigra. We validated these circuit pathways using Granger causality and the metabolic connectivity map for identifying effective connectivities against a priori knowledge by testing the significance of directed pathways (P < .05, false discovery rate–corrected). RESULTS: We found 3 types of effective connectivities in the dopamine reward circuit: long-range, neighborhood, and symmetric. Granger causality analysis revealed long-range connections in the orbital frontal cortex–caudate and orbital frontal cortex–nucleus accumbens regions. Metabolic connectivity map analysis revealed neighborhood connections in the nucleus accumbens–caudate, substantia nigra–thalamus, and thalamus-caudate regions. Metabolic connectivity map analysis also found symmetric connections in each of the bilateral nucleus accumbens, caudate, thalamus, and orbital frontal cortex–caudate regions. Different patterns in directional networks of the dopamine reward circuit were revealed by Granger causality and metabolic connectivity map analyses. CONCLUSIONS: Granger causality analysis primarily identified bidirectional cortico-nucleus connections, while the metabolic connectivity map primarily identified direct connections among neighborhood and symmetric regions. The results of this study indicated that investigations of effective connectivities should use an appropriate analysis method depending on the purpose of the study.