AUTHOR=Zhang Wei , Zhang Tianhao , Pan Tingting , Zhao Shilun , Nie Binbin , Liu Hua , Shan Baoci ,  Alzheimer’s Disease Neuroimaging Initiative 

TITLE=Deep Learning With 18F-Fluorodeoxyglucose-PET Gives Valid Diagnoses for the Uncertain Cases in Memory Impairment of Alzheimer’s Disease

JOURNAL=Frontiers in Aging Neuroscience

VOLUME=Volume 13 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/aging-neuroscience/articles/10.3389/fnagi.2021.764272

DOI=10.3389/fnagi.2021.764272

ISSN=1663-4365

ABSTRACT=Objectives: Neuropsychological tests are an important basis for memory impairment diagnosis in Alzheimer’s disease (AD). However, multiple memory tests might be conflicting within subjects and lead to uncertain diagnoses in some cases. This study proposed a framework to give diagnoses for the uncertain cases in memory impairment.

Methods: We collected 2386 samples including AD, mild cognitive impairment (MCI) and cognitive normal (CN) with 18F-Fluorodeoxyglucose (FDG) PET and three different neuropsychological tests (Mini-mental State Examination, Alzheimer's Disease Assessment Scale-cognitive subscale and Clinical Dementia Rating) from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). A deep learning (DL) framework with FDG-PET was proposed to give a diagnosis to uncertain memory impairment cases that were conflicting between tests. Subsequent ANOVA, chi-squared and t-test were applied to explain the potential causes of uncertain cases.

Results: For the certain cases in testing set, the proposed DL framework outperformed other methods with 95.65% accuracy. For the uncertain cases, its positive diagnoses had significant (p < .001) worse decline in memory function than negative diagnoses, in a longitudinal study of 40 months on average. In memory impaired group, uncertain cases were mainly explained by an AD metabolism pattern but mild in extent (p < .05). In healthy group, uncertain cases were mainly explained by a not-energetic mental state (p < .001) measured by global deterioration scale (GDS), with a significant depression-related metabolism pattern detected (p < .05).

Conclusions: A DL framework for diagnosing uncertain cases of memory impairment is proposed. Proved by longitudinal tracing of its diagnoses, it showed clinical validity and had application potential. Its valid diagnoses also gave evidence and explanation of uncertain cases based on neurodegeneration and depression mental state.