A 3D convolutional neural network to classify subjects as Alzheimer’s disease, frontotemporal dementia or healthy controls using brain 18F-FDG PET
Open accessAbstract
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.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
CNNs on MRI and FDG-PET data reported 86.0-94.6% classification accuracy for differentiating FTD, AD, and healthy controls.
“Studies utilizing CNNs on MRI and FDG-PET data have reported high classification accuracies (86.0–94.6%) for differentiating FTD, AD, and healthy controls”
From Multimodal neuroimaging and AI integration in cognitive disorders: advances, challenges, and future directions for precision medicine (2026) · cites this paper as Rogeau et al., 2024
Cites
Nichols, E., Steinmetz, J. D., Vollset, S. E., Fukutaki, K., Chalek, J., Abd-Allah, F., Abdoli, A., Abualhasan, A., Abu-Gharbieh, E., Akram, T. T., et al. (2022) · The Lancet Public Health · doi:10.1016/s2468-2667(21)00249-8 · Cited by 5,575 · see citing works
Hu, J., Qing, Z., Liu, R., Zhang, X., Lv, P., Wang, M., Wang, Y., He, K., Gao, Y., & Zhang, B. (2021) · Frontiers in Neuroscience · doi:10.3389/fnins.2020.626154 · Cited by 70 · cited 1× · see citing works
“Artificial intelligence (AI) techniques have been successfully applied in radiology and are increasingly popular in neuroimaging, especially for structural imaging (Hu et al., 2020; Jo et al., 2019; Nemoto et al., 2021).”
Differentiating Dementia with Lewy Bodies and Alzheimer's Disease by Deep Learning to Structural MRI
Nemoto, K., Sakaguchi, H., Kasai, W., Hotta, M., Kamei, R., Noguchi, T., Minamimoto, R., Arai, T., & Asada, T. (2021) · Journal of Neuroimaging · doi:10.1111/jon.12835 · Cited by 37 · see citing works
Deep grading for MRI-based differential diagnosis of Alzheimer’s disease and frontotemporal dementia
Nguyen, H.-D., Clément, M., Planche, V., Mansencal, B., & Coupé, P. (2023) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2023.102636 · Cited by 35 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Deep learning techniques for Alzheimer's disease detection in 3D imaging: A systematic review
Zia‐ur‐Rehman, Awang, M. K., Ali, G., & Faheem, M. (2024) · Health Science Reports · doi:10.1002/hsr2.70025 · Cited by 23 · see citing works
Cerebral glucose metabolism in Alzheimer's disease
Salmon, É., Collette, F., & Bastin, C. (2024) · Cortex · doi:10.1016/j.cortex.2024.07.004 · Cited by 33 · see citing works
Ahmed, R., Fahad, N., Miah, M. U., Hossen, M. J., Morol, M., Mahmud, M., & Rahman, M. M. (2024) · Healthcare Analytics · doi:10.1016/j.health.2024.100362 · Cited by 29 · see citing works
Mahamud, E., Assaduzzaman, M., Islam, J., Fahad, N., Hossen, M. J., & Ramanathan, T. T. (2025) · Intelligence-Based Medicine · doi:10.1016/j.ibmed.2025.100240 · Cited by 24 · see citing works
Use of Artificial Intelligence in Imaging Dementia
Aljuhani, M., Ashraf, A., & Edison, P. (2024) · Cells · doi:10.3390/cells13231965 · Cited by 21 · see citing works
Hagos, M. T., Curran, K. M., Namee, B. M., & Initiative, A. D. N. (2025) · Scientific Reports · doi:10.1038/s41598-025-86110-y · Cited by 14 · see citing works
Dattola, S., Ielo, A., Varone, G., Cacciola, A., Quartarone, A., & Bonanno, L. (2025) · Frontiers in Aging Neuroscience · doi:10.3389/fnagi.2025.1547727 · Cited by 14 · see citing works
Akan, T., Akan, S., Alp, S., Ledbetter, C. R., Tafti, A. P., Arevalo, O., & Bhuiyan, M. A. N. (2025) · Journal of the Neurological Sciences · doi:10.1016/j.jns.2025.123735 · Cited by 11 · see citing works
Alzheimer's disease recognition via long-range state space model using multi-modal brain images
Ren, Z., Zhou, M., Shakil, S., & Tong, R. K.-Y. (2025) · Frontiers in Neuroscience · doi:10.3389/fnins.2025.1576931 · Cited by 8 · see citing works
Classifying cognitive impairment based on FDG-PET and combined T1-MRI and rs-fMRI: An ADNI study
Jahani, I., Jahani, A., Delrobaei, M., Khadem, A., & Macintosh, B. J. (2024) · Journal of Alzheimer's Disease · doi:10.1177/13872877241302493 · Cited by 8 · see citing works
“27 Their study showed that 3D CNNs have great potential to distinguish various kinds of dementia from normal aging.”
Dang, M., Liu, B., Chen, Y., & Zhang, Z. (2026) · Psychoradiology · doi:10.1093/psyrad/kkag007 · Cited by 7 · see citing works
Hussain, S. S., Xu, D., Shah, P. M., Khan, H., & Zeb, A. (2025) · Computerized Medical Imaging and Graphics · doi:10.1016/j.compmedimag.2025.102638 · Cited by 7 · see citing works
MRI-Based Biomarker in the Diagnosis of Alzheimer’s Disease Using Attention-UNet
Arumugam, J., Venkatesan, V., & Beigh, T. (2025) · SN Computer Science · doi:10.1007/s42979-025-03686-y · Cited by 5 · see citing works
“Brain glucose metabolism-based 3D Neural Network was designed by Antoine Rogeau et al [21].”
Revolutionizing Brain Disease Diagnosis: The Convergence of AI, Genetic Screening, and Neuroimaging
Wang, L.-C., Li, S., & Jin, X. (2024) · Proceedings of the 2024 International Conference on Smart Healthcare and Wearable Intelligent Devices · doi:10.1145/3703847.3703850 · Cited by 5
“This model, surpassing clinical interpretations by specialists, marks a pioneering success in discriminating between AD, FTD, and CN subjects through deep learning, promising enhanced patient outcomes through precise diagnosis and treatment [63.].”
Lightweight 3D CNN for MRI Analysis in Alzheimer’s Disease: Balancing Accuracy and Efficiency
Cao, K., Lu, Z., Zhao, C., Du, J.-M., Li, L.-L., Jung, H.-K., & Geng, M.-H. (2025) · Journal of Imaging · doi:10.3390/jimaging11120426 · Cited by 4
“For instance, multi-scale feature extraction and attention mechanisms have been incorporated to enhance feature representation and robustness in AD, MCI, and NC classification tasks [12].”
Shakir, M. H., Seemanta, S., Shimu, S. Z., Salekin, S., Arman, A. N. M., Miah, M. U., Rahman, M. M., & Mahmud, M. (2025) · IEEE International Joint Conference on Neural Network · doi:10.1109/ijcnn64981.2025.11228086 · Cited by 4 · see citing works
“Cohen’s Kappa Score Calculation: Cohen’s kappa statistic is a widely used measure of agreement between two raters, particularly in assessing interrater reliability [19].”
Neuroimaging in the era of biologically defined Alzheimer's disease and disease-modifying therapy
Benzinger, T. L. S., Cogswell, P. M., Arani, A., Chaney, A. M., Hutson, K., Glasser, M. F., Graff‐Radford, J., McConathy, J., Raji, C. A., Teunissen, C. E., et al. (2026) · The Lancet Neurology · doi:10.1016/s1474-4422(26)00285-1 · Cited by 4 · see citing works
Tsougos, I., Valotassiou, V. J., Tsivaka, D., Σάτρα, Μ., Angelidis, G., Papatriantafyllou, J. D., Panagiotidis, E., Dardiotis, E., Hadjigeorgiou, G. M., & Georgoulias, P. (2026) · Annals of Nuclear Medicine · doi:10.1007/s12149-026-02194-0 · Cited by 2 · see citing works
“Early DL work has also explored 3D CNNs for multi-diagnosis classifica - tion including AD vs frontotemporal dementia (FTD) using FDG volumes, highlighting volumetric context for pattern recognition [39 ].”
A Low Cost CNN-Based Method for Differential Diagnosis of Alzheimer’s and Frontotemporal Dementia
Jha, B. K., Siddiqui, M. F., & Pandey, A. (2024) · 2024 IEEE International Conference on Computer Vision and Machine Intelligence (CVMI) · doi:10.1109/cvmi61877.2024.10782448 · Cited by 2 · see citing works
“The authors of [11], used a 3D neural network for the classification of AD, FTD, and cognitively normal (CN) subjects using Positron Emission Tomography (PET) scans for training a model similar to 3D-VGG16.”
Patient Diagnosis Alzheimer’s Disease with Multi-Stage Features Fusion Network and Structural MRI
Nguyen, T. M. T., & Bui, N. (2025) · Journal of Dementia and Alzheimer's Disease · doi:10.3390/jdad2040035 · Cited by 1 · see citing works
AI-aided neuroimaging for the diagnosis of Alzheimer's disease: A concise review of recent studies
Lee, Y., Nadalin, P., Onemli, E., Hadad, R., Yu, F., Greenbaum, L., Greenbaum, A., & Ahn, J. (2026) · Heliyon · doi:10.1016/j.heliyon.2026.e45308
Lv, C., Liu, T., Chen, H., Huang, W., Huang, M., Zang, L., Shen, C., Guo, Y., & Chen, F. (2026) · iScience · doi:10.1016/j.isci.2026.117313
Zhou, G.-T. (2026) · Edelweiss Applied Science and Technology · doi:10.55214/2576-8484.v10i8.13501
“For PET imaging, Rogeau et al. [18] devised a VGG16-like network to distinguish 18F-FDG PET into AD, Frontotemporal Dementia, and Healthy Control groups.”
Manouvriez, D., Kuchcinski, G., Lecerf, S., Lahousse, H., Rogeau, A., Villain, N., Kas, A., Pyatigorskaya, N., Nguyen, M., Petrovic, S., et al. (2026) · NeuroImage · doi:10.1016/j.neuroimage.2026.122136
Self-citationZhao, J.-W., Fang, F.-X., Mai, D.-M., Lin, J.-W., Li, F., Wei, L.-Y., Wei, Z., Feng, C., Cao, H.-L., Liao, N.-K., et al. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.109882
A Comprehensive Study of Deep Learning Techniques for Early Detection of Frontotemporal Dementia
M, A., S, R., & Kumar, G. K. (2026) · 2026 6th International Conference on Pervasive Computing and Social Networking (ICPCSN) · doi:10.1109/icpcsn68523.2026.11543797
“Rogeau et al. [2] constructed an AD, FDG-PET-based 3D CNN model, which classified AD, FDG-PET, and healthy controls as interpretable by Grad-CAM.”
Influential
- Authors
- Rogeau, A., Hives, F., Bordier, C., Lahousse, H., Roca, V., Lebouvier, T., Pasquier, F., Huglo, D., Semah, F., & Lopes, R.
- Affiliations
- University College London Hospitals NHS Foundation Trust (GB)
- Centre Hospitalier Universitaire de Lille (FR)
- Centre National de la Recherche Scientifique (FR)
- Inserm (FR)
- Corresponding author
- Rogeau, Antoine
- ORCID
- Venue
- NeuroImage
- Publisher
- Elsevier BV
- Published
- 2024-03
- Volume
- 288
- Pages
- 120530
- ISSN
- 1053-8119 (ISSN-L)
- OpenAlex
- W4391512868
- Cited by
- 29
- References
- 45
- Citations by year
- 2026: 9 · 2025: 12 · 2024: 8
- FWCI
- 8.52
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- NeuroImage (publishedVersion, cc-by)
- PubMed (publishedVersion)
- HAL (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- LillOA (Université de Lille (University Of Lille)) (submittedVersion, cc-by)
- LillOA (Université de Lille (University Of Lille)) (submittedVersion, other-oa)
- Funders
- National Institute on Aging
- National Institutes of Health (U01 AG024904)
- U.S. Department of Defense (W81XWH-12-2-0012)
- National Institute of Biomedical Imaging and Bioengineering
- Alzheimer's Disease Neuroimaging Initiative
- Topics
- Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
- Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Medical Imaging Techniques and Applications
- Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
- Keywords
- Alzheimer's disease, frontotemporal dementia, 18F-FDG PET, brain glucose metabolism, 3D convolutional neural networks, VGG16, neurodegenerative disease classification, occlusion analysis, Grad-CAM, deep learning diagnosis
- MeSH terms
- Alzheimer Disease, Brain, Humans, Retrospective Studies, Neural Networks, Computer, Fluorodeoxyglucose F18, Positron-Emission Tomography, Frontotemporal Dementia
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
