Neuro-degenerative diseases · 2025
A Radiomics-Based Analysis of Functional Dopaminergic Scintigraphic Imaging for the Diagnosis of Dementia with Lewy Bodies.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
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.
Perriraz Jérémy, Abler Daniel, Salvioni Chiabotti Paolo, Hall Caroline, Lejay Noemie, Kurian George K, Vetterli Thomas, Rouaud Olivier, Nicod Lalonde Marie, Schaefer Niklaus, Allali Gilles, Depeursinge Adrien, Prior John O, Jreige Mario
Аннотация
INTRODUCTION: Radiomics features, a technique based on quantitative image analysis, can be used to capture tissue and lesion characteristics, such as heterogeneity and shape. Using functional dopaminergic scintigraphy, we aimed to study the value of radiomics features in predicting the diagnosis of dementia with Lewy bodies (DLB).
MATERIALS AND METHODS: We retrospectively analyzed 74 patients (29 F and 45 M, mean age 71.6 ± 9.2) investigated in the Leenaards Memory Center (Lausanne University Hospital) for DLB who underwent quantitative I-123-ioflupane single positron emission computed tomography (SPECT)/CT (DaTscan). All scanned examinations had xSPECT reconstruction, allowing standard uptake value (SUV) quantification. We segmented the right and left striatum using a 3D Slicer and performed radiomics feature extraction and analysis using the QuantImage v2 platform. The dataset was divided into training (80%) and test (20%) sets, and various classification algorithms were used to predict the definitive clinical diagnosis of DLB using xSPECT and/or clinical features. Receiver operating characteristic (ROC) curve analysis was performed to characterize the performance of the obtained models.
RESULTS: Thirty-three of 74 patients (45%) were diagnosed with DLB. The xSPECT radiomics models showing the highest diagnostic performance were developed based on nine non-correlated features from both striatal regions and a support vector classifier (SVC) algorithm. The xSPECT radiomics models demonstrated superior performance compared to models based on SUV intensity features alone (p = 0.001) or clinical features alone (p = 0.001), with area under the ROC curve (AUC) values of 0.932 (0.920-0.944), 0.856 (0.840-0.875), and 0.793 (0.770-0.815), respectively. The combined model, incorporating both clinical and xSPECT features, achieved the highest overall performance with a sensitivity of 100% (95% confidence interval: 100-100), specificity of 89.7% (87.6-91.4), and an AUC of 0.956 (0.945-0.964).
CONCLUSION: The radiomics model based on quantitative I-123-ioflupane xSPECT/CT showed high diagnostic accuracy in predicting the diagnosis of DLB using diverse features derived from striatal analysis. This tool may improve the diagnostic accuracy of I-123-ioflupane, which is of major importance for DLB diagnosis.
MATERIALS AND METHODS: We retrospectively analyzed 74 patients (29 F and 45 M, mean age 71.6 ± 9.2) investigated in the Leenaards Memory Center (Lausanne University Hospital) for DLB who underwent quantitative I-123-ioflupane single positron emission computed tomography (SPECT)/CT (DaTscan). All scanned examinations had xSPECT reconstruction, allowing standard uptake value (SUV) quantification. We segmented the right and left striatum using a 3D Slicer and performed radiomics feature extraction and analysis using the QuantImage v2 platform. The dataset was divided into training (80%) and test (20%) sets, and various classification algorithms were used to predict the definitive clinical diagnosis of DLB using xSPECT and/or clinical features. Receiver operating characteristic (ROC) curve analysis was performed to characterize the performance of the obtained models.
RESULTS: Thirty-three of 74 patients (45%) were diagnosed with DLB. The xSPECT radiomics models showing the highest diagnostic performance were developed based on nine non-correlated features from both striatal regions and a support vector classifier (SVC) algorithm. The xSPECT radiomics models demonstrated superior performance compared to models based on SUV intensity features alone (p = 0.001) or clinical features alone (p = 0.001), with area under the ROC curve (AUC) values of 0.932 (0.920-0.944), 0.856 (0.840-0.875), and 0.793 (0.770-0.815), respectively. The combined model, incorporating both clinical and xSPECT features, achieved the highest overall performance with a sensitivity of 100% (95% confidence interval: 100-100), specificity of 89.7% (87.6-91.4), and an AUC of 0.956 (0.945-0.964).
CONCLUSION: The radiomics model based on quantitative I-123-ioflupane xSPECT/CT showed high diagnostic accuracy in predicting the diagnosis of DLB using diverse features derived from striatal analysis. This tool may improve the diagnostic accuracy of I-123-ioflupane, which is of major importance for DLB diagnosis.