- 期刊
整合FDG-PET和MRI以提高失智症的檢測和診斷
Integrate FDG-PET and MRI to Improve Detection and Diagnosis of Dementia
摘要
失智症（Dementia），主要是大腦細胞因病變而急劇退化甚至死亡，從而影響失智症患者的腦功能，逐漸喪失記憶力、語言能力、視覺空間判斷力等。最近的文獻研究中指出了關於不同類型中的神經影像學以及各種異常的生物標記物（biomarker），這些生物標記物有助於顯著改善各種早期診斷和分類不同種類的失智症，在本研究中我們利用PET/MR－體化機器，同時掃描患者[F18]氟脫氧葡萄糖正子掃描（18F-fluorodeoxyglucose, positron emission tomography, FDG-PET）與磁振造影（MRI）的T1結構影像，並以體素為基礎（voxel-based）的影像分析方法，找出失智症患者PET與MRI的共通性與差異性，以增加失智症種類的診斷正確率。臨床診斷為阿茲海默病患者六位，伴有額顳葉失智患者九位和健康受試者十七位同時接受FDG-PET、MRI，以及臨床和行為評估。每位受試者的FDG-PET和MRI影像都經由相關PET與MRI數據後處理軟體以獲得所有區域的更精準的對位。本影像處理步驟都是採用可在Matlab7.6平台上（MathWorks Inc.）的SPM12軟體進行。組間統計分析比較則使用方差分析（ANOVA, analyses of variance），統計分析軟體工具則使用SPSS 17.0。以MRI 大腦的灰質體積大小或影像強度來代表神經元是否有萎縮，MRI在分辨阿茲海默症（Alzheimer's disease, AD）患者在海馬迴區域具有高度的專一性，但前額葉與顳葉腦區也只能分辨健康受試者與AD患者之間的差異，這表明MRI結構影像對檢測失智症患者的分類度較低。但是以FDG-PET比較失智症患者的葡萄糖代謝低下，在後顳頂葉與後扣帶迴腦區也只能分辨健康受試者與額顳葉失智症（Frontotemporal Dementia, FTD）、AD患者之間的差異，但在大腦的前扣帶迴區域，可以明確的區分健康受試者、FTD患者與AD患者三組之間的差異。我們的研究結果顯示，以FDG-PET與MRI兩種不同成像方式的結果和整合，可能提供一種新的方法來提高這些失智症的診斷準確性。
並列摘要
Dementia is mainly caused by brain cell degradation and death due to pathological changes, which affects the brain function of patients with dementia, and gradually loses memory, language ability, visual spatial judgment and so on. Biomarkers for different types of neuroimaging and various abnormalities have been reported in recent literature studies. These biomarkers help to significantly improve various early diagnoses and classifications of different types of dementia. In this study we used PET/MR machines and collected [F18] Fluorodeoxyglucose positron scanning (FDG-PET) and magnetic resonance imaging (MRI) T1 structural images simultaneously. A voxel-based image analysis method is used to find out the commonality and difference of PET and MRI in patients with dementia and increase the accuracy of diagnosis. We collected Alzheimer's disease (AD: n = 6), patients with frontotemporal dementia (FTD: n = 9) and healthy controls (n = 17) with both FDG-PET、MRI, and clinical behavioral assessment. Each subject's FDG-PET and MRI images were post-processed to obtain a more accurate alignment of all regions. Image processing steps were used the SPM12 toolbox and performed on the Matlab 7.6 platform (MathWorks Inc.). ANOVA (analysis of variance) was performed between groups and statistical analysis software uses SPSS 17.0. Brain gray matter volume or image intensity to represent whether the neurons have atrophy. MRI in the differentiation of AD patients in the hippocampus region has a high degree of specificity, but the prefrontal and temporal lobe brain regions can only distinguish healthy subjects. The difference between patients with dementia, indicates that MRI structural images have a low classification information related to the detection of dementia patients. However, FDG-PET compares the patients with dementia, the posterior parietal lobe and posterior cingulate gyrus can only distinguish the difference between healthy subjects and patients with FTD and AD. But the anterior cingulate gyrus in front of the brain can clearly distinguish the difference between healthy subjects、FTD and AD patients. Our results suggest that the results and integration of FDG-PET and MRI imaging methods may provide a new way to improve the diagnostic accuracy of these dementia.