Department of Functional Imaging and Artificial Intelligence, Kanazawa University, 13-1 Takara-machi, Kanazawa, 920-8640, Japan.
Department of Nuclear Medicine, Kanazawa University, Kanazawa, Japan.
Ann Nucl Med. 2024 Oct;38(10):814-824. doi: 10.1007/s12149-024-01950-4. Epub 2024 Jun 13.
We aimed to establish a practical diagnostic index for Lewy body diseases (LBD), such as Parkinson's disease and dementia, with Lewy bodies in outpatient settings and criteria for exempting patients from late imaging.
We acquired early and late I-metaiodobenzylguanidine (MIBG) images from 108 consecutive patients with suspected LBD and standardized heart-to-mediastinum (H/M) ratios for collimator conditions. Exclusions included young-onset Parkinson's disease (age < 50 years) and genetic transthyretin-type amyloidosis. We developed logistic models incorporating H/M ratios with or without age (n = 92). The sympathetic MIBG index for LBD (SMILe index), categorized LBD likelihood from 0 (lowest) to 1 (highest). Diagnostic accuracy was assessed as the area under the receiver operating characteristic (ROC) curve (AUC). The characteristics of the new index were compared with H/M ratios. The need for late imaging was explored using the SMILe index.
Early or late SMILe indexes using a single H/M ratio variable discriminated LBD from non-LBD. The AUC values for early and late SMILe indexes were 0.880 and 0.894 (p < 0.0001 for both), identical to those for early and late H/M ratios. The sensitivity and the specificity of early SMILe indexes with a 0.5 threshold were 76% and 90%, achieving accuracy of accuracy 86%. Similarly, the late SMILe index demonstrated a sensitivity of 76% and specificity of 87%, with an accuracy of 84%. Early SMILe indexes < 0.3 or > 0.7 (representing 84% patients) indicated a diagnosis without a late MIBG study.
The I-MIBG-derived SMILe indexes provide likelihood of LBD, and those with a 50% threshold demonstrated optimal diagnostic accuracy for LBD. The index values of either < 0.3 or > 0.7 accurately selected patients who do not need late imaging.
我们旨在建立一个实用的诊断指标，用于在门诊环境中诊断路易体疾病（LBD），如帕金森病和痴呆症，以及免除患者进行晚期成像的标准。
我们从 108 例疑似 LBD 患者中获取了早期和晚期 I-间碘苄胍（MIBG）图像，并为准直器条件获得了标准化的心脏与纵隔（H/M）比值。排除标准包括年轻起病的帕金森病（年龄<50 岁）和遗传性转甲状腺素蛋白型淀粉样变性。我们开发了包含 H/M 比值和年龄的逻辑模型（n=92）。用于 LBD 的交感神经 MIBG 指数（SMILe 指数），将 LBD 的可能性从 0（最低）到 1（最高）进行分类。诊断准确性作为接收者操作特征（ROC）曲线下的面积（AUC）进行评估。该新指数的特征与 H/M 比值进行了比较。使用 SMILe 指数探索了晚期成像的必要性。
使用单一 H/M 比值变量的早期或晚期 SMILe 指数可区分 LBD 与非 LBD。早期和晚期 SMILe 指数的 AUC 值分别为 0.880 和 0.894（两者均 p<0.0001），与早期和晚期 H/M 比值相同。早期 SMILe 指数 0.5 阈值的灵敏度和特异性分别为 76%和 90%，准确率为 86%。同样，晚期 SMILe 指数的灵敏度为 76%，特异性为 87%，准确率为 84%。早期 SMILe 指数<0.3 或>0.7（代表 84%的患者）表明无需进行晚期 MIBG 研究即可诊断。
I-MIBG 衍生的 SMILe 指数提供了 LBD 的可能性，50%的阈值对 LBD 的诊断准确性最佳。指数值<0.3 或>0.7 可以准确选择不需要晚期成像的患者。
