A 3D convolutional neural network to classify subjects as Alzheimer's disease, frontotemporal dementia or healthy controls using brain 18F-FDG PET

• Published in: NeuroImage (Orlando, Fla.) Vol. 288; p. 120530
• Main Authors: Rogeau, Antoine, Hives, Florent, Bordier, Cécile, Lahousse, Hélène, Roca, Vincent, Lebouvier, Thibaud, Pasquier, Florence, Huglo, Damien, Semah, Franck, Lopes, Renaud
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.03.2024; Elsevier Limited; Elsevier
• Series: NeuroImage
• Subjects: Alzheimer Disease - diagnostic imaging; Alzheimer's disease; Artificial intelligence; Autopsies; Biomarkers; Brain - diagnostic imaging; Brain mapping; Brain research; Cerebral blood flow; Classification; Cognitive science; Convolutional neural network; Deep learning; Dementia; Dementia disorders; FDG PET; Fluorodeoxyglucose F18; Frontotemporal dementia; Frontotemporal Dementia - diagnostic imaging; Glucose metabolism; Humans; Image processing; Ischemia; Magnetic resonance imaging; Medical imaging; Metabolism; Neural networks; Neural Networks, Computer; Neurodegeneration; Neurodegenerative diseases; Neuroimaging; Positron emission tomography; Positron-Emission Tomography - methods; Retrospective Studies; United States > US; Deep learning; Frontotemporal dementia; Alzheimer's disease; Convolutional neural network; FDG PET
• DOI: 10.1016/j.neuroimage.2024.120530
• ISSN: 1053-8119; 1095-9572

•Visual interpretation of [18F]-FDG-PET scans remains challenging and with the advent of new treatments, accurate diagnosis is more important than ever.•A tailor-made 3D VGG16-like network outperforms clinical interpretation by specialist physicians, achieving an overall accuracy of 89.8 % in predicting the class of test scans.•The posterior cingulate cortex (PCC) for AD and anterior regions for FTD were key regions in the classification process.•The findings suggest the potential for integrating deep learning tools into clinical practice for more accurate and objective neurodegenerative disease diagnosis using [18F]-FDG-PET scans. With the arrival of disease-modifying drugs, neurodegenerative diseases will require an accurate diagnosis for optimal treatment. Convolutional neural networks are powerful deep learning techniques that can provide great help to physicians in image analysis. The purpose of this study is to introduce and validate a 3D neural network for classification of Alzheimer's disease (AD), frontotemporal dementia (FTD) or cognitively normal (CN) subjects based on brain glucose metabolism. Retrospective [18F]-FDG-PET scans of 199 CE, 192 FTD and 200 CN subjects were collected from our local database, Alzheimer's disease and frontotemporal lobar degeneration neuroimaging initiatives. Training and test sets were created using randomization on a 90 %-10 % basis, and training of a 3D VGG16-like neural network was performed using data augmentation and cross-validation. Performance was compared to clinical interpretation by three specialists in the independent test set. Regions determining classification were identified in an occlusion experiment and Gradient-weighted Class Activation Mapping. Test set subjects were age- and sex-matched across categories. The model achieved an overall 89.8 % accuracy in predicting the class of test scans. Areas under the ROC curves were 93.3 % for AD, 95.3 % for FTD, and 99.9 % for CN. The physicians' consensus showed a 69.5 % accuracy, and there was substantial agreement between them (kappa = 0.61, 95 % CI: 0.49–0.73). To our knowledge, this is the first study to introduce a deep learning model able to discriminate AD and FTD based on [18F]-FDG PET scans, and to isolate CN subjects with excellent accuracy. These initial results are promising and hint at the potential for generalization to data from other centers.