Academic radiology · 2026
Optimization of Multiparametric MRI Sequence Combinations for Early Detection of Chronic Kidney Disease.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Liu Zhenni, Gong Yunan, Liu Zhonglin, Wu Xiaoqing, Zhou Zhiqiang, Shen Daming, Shang Song'an, Liu Wenjin, Lu Siyuan, Ye Jing
Аннотация
RATIONALE AND OBJECTIVES: To evaluate renal multiparametric MRI sequence combinations for early chronic kidney disease (CKD) detection and identify an approach balancing performance and clinical feasibility.
METHODS: This prospective study included 105 patients with early CKD and 127 healthy controls. All participants underwent renal MRI with T1WI, DWI, T1 mapping, and BOLD imaging. Twenty four features were extracted from T1WI, ADC, T1, and R2* maps. Participants were divided into training and temporally independent test cohorts by enrollment date. Fifteen predefined models were developed using elastic net regularized logistic regression. Representative models were selected by fivefold cross validation and evaluated using ROC analysis, calibration analysis, decision curve analysis, and fixed threshold diagnostic metrics.
RESULTS: Feature reproducibility was acceptable to good, with intraclass correlation coefficients of 0.731 to 0.891. In training, the model using all four sequences achieved the highest AUC of 0.903. DWI was the best single sequence model (AUC, 0.874), T1WI+DWI was the best two sequence model (AUC, 0.884), and T1WI+DWI+T1 mapping was selected as the representative three sequence model (AUC, 0.895). Cortical ADC variability, medullary ADC variability, and ADC corticomedullary difference were consistently retained. In the test cohort, the model using all four sequences showed the highest AUC of 0.866 and the best calibration. T1WI+DWI achieved the highest accuracy and F1 score, at 0.844 and 0.815.
CONCLUSION: Renal multiparametric MRI supports noninvasive early CKD detection. DWI was the most important single sequence. Multi sequence integration improved diagnostic performance, while T1WI+DWI offered a practical balance between performance and feasibility.
METHODS: This prospective study included 105 patients with early CKD and 127 healthy controls. All participants underwent renal MRI with T1WI, DWI, T1 mapping, and BOLD imaging. Twenty four features were extracted from T1WI, ADC, T1, and R2* maps. Participants were divided into training and temporally independent test cohorts by enrollment date. Fifteen predefined models were developed using elastic net regularized logistic regression. Representative models were selected by fivefold cross validation and evaluated using ROC analysis, calibration analysis, decision curve analysis, and fixed threshold diagnostic metrics.
RESULTS: Feature reproducibility was acceptable to good, with intraclass correlation coefficients of 0.731 to 0.891. In training, the model using all four sequences achieved the highest AUC of 0.903. DWI was the best single sequence model (AUC, 0.874), T1WI+DWI was the best two sequence model (AUC, 0.884), and T1WI+DWI+T1 mapping was selected as the representative three sequence model (AUC, 0.895). Cortical ADC variability, medullary ADC variability, and ADC corticomedullary difference were consistently retained. In the test cohort, the model using all four sequences showed the highest AUC of 0.866 and the best calibration. T1WI+DWI achieved the highest accuracy and F1 score, at 0.844 and 0.815.
CONCLUSION: Renal multiparametric MRI supports noninvasive early CKD detection. DWI was the most important single sequence. Multi sequence integration improved diagnostic performance, while T1WI+DWI offered a practical balance between performance and feasibility.