Early and Late Fusion for Multimodal Aggression Prediction in Dementia Patients: A Comparative Analysis

Topic: Amyloid, tau and blood-test diagnosis

Aggression in patients with dementia poses significant caregiving and clinical issues. In this work, fusion approaches—Early Fusion and Late Fusion—were compared to classify aggression using audio and visual signals. Early Fusion integrates the extracted features of the two modalities into one dataset before classification, while Late Fusion integrates the prediction probabilities of standalone audio and visual classifiers with a meta-classifier. Both models were tested using a Random Forest classifier with five-fold cross-validation, and the performance was compared on accuracy, precision, recall, F1-score, ROC-AUC, and inference time. The results showcase that Late Fusion is superior to Early Fusion in terms of accuracy (0.876 vs. 0.828), recall (0.914 vs. 0.818), F1-score (0.867 vs. 0.835), and ROC-AUC score (0.970 vs. 0.922), proving more suitable for high-sensitivity use cases like healthcare and security. However, Early Fusion exhibited higher precision (0.852 vs. 0.824), indicating that in cases when false positives are a requirement, Early Fusion is preferable. Paired t-tests were applied for statistical comparison and indicate that precision alone is significantly different, with the advantage of Early Fusion. Late Fusion also performs slightly less in inference time, which makes it suitable for use in real-time systems. These findings provide significant information on multimodal fusion strategies and their applicability in the detection of aggressive behavior, which can contribute to the development of efficient monitoring systems for dementia care.

Key findings

• Aggression in patients with dementia poses significant caregiving and clinical issues.
• In this work, fusion approaches—Early Fusion and Late Fusion—were compared to classify aggression using audio and visual signals.
• The results showcase that Late Fusion is superior to Early Fusion in terms of accuracy (0.876 vs. 0.828), recall (0.914 vs. 0.818), F1-score (0.867 vs. 0.835), and ROC-AUC score (0.970 vs. 0.922), proving more suitable for high-sensitivity use cases like healthcare and security.
• Paired t-tests were applied for statistical comparison and indicate that precision alone is significantly different, with the advantage of Early Fusion.
• These findings provide significant information on multimodal fusion strategies and their applicability in the detection of aggressive behavior, which can contribute to the development of efficient monitoring systems for dementia care.

Why it matters

This work compares two ways of combining audio and visual signals to classify aggression in dementia patients, and found that Late Fusion gave better accuracy, recall, F1-score and ROC-AUC, while Early Fusion gave slightly better precision. It is a technical classification study and does not test any treatment, cure or prevention for Alzheimer's disease.

Study details

• Evidence strengthLow

Read the original

• DOI: 10.3390/app15115823
• Original source →
• Free full text: Applied Sciences →

Full text: free to read — OpenAlex lists this copy as free to read: the link above (checked 6 Feb 1020).

@article{galanakis2025-e7e8a3b1,
  author  = {Galanakis, Ioannis and Soldatos, Rigas Filippos and Karanikolas, Nikitas Ν. and Voulodimos, Athanasios S. and Voyiatzis, Ioannis and Σαμαράκου, Μαρία},
  title   = {{Early and Late Fusion for Multimodal Aggression Prediction in Dementia Patients: A Comparative Analysis}},
  journal = {Applied Sciences},
  year    = {2025},
  doi     = {10.3390/app15115823},
  url     = {https://doi.org/10.3390/app15115823}
}

Studies this one cites

Referenced by this study and also in the alzhub library, so you can read them here.

• Mini-Mental State Examination (MMSE) for the early detection of dementia in people with mild cognitive impairment (MCI) The Cochrane database of systematic reviews · 2021-07-27 · usefulness score 75/100 · Getting a diagnosis
• Interpretable machine learning for dementia: A systematic review Alzheimer's & dementia : the journal of the Alzheimer's Association · 2023-02-03 · usefulness score 66/100 · Digital health tools
• Machine Learning for Dementia Prediction: A Systematic Review and Future Research Directions Journal of medical systems · 2023-02-01 · usefulness score 62/100 · Digital health tools

Related studies in Brain tissue and pathology

• Comparative lipidomic analysis of mouse and human brain with Alzheimer disease The Journal of biological chemistry · 2011-12-01 · usefulness score 52/100
• Psychosocial burdens in early- versus late-onset dementia: analysis of discrimination, stress, and loneliness in the All of Us Research Program Innovation in Aging · 2025-01-01 · usefulness score 40/100
• Cholinergic markers in elderly patients with early signs of Alzheimer disease JAMA · 1999-04-01 · usefulness score 52/100
• Updating the Evidence on the Association between Serum Cholesterol and Risk of Late-Life Dementia: Review and Meta-Analysis Journal of Alzheimer's disease : JAD · 2017-01-01 · usefulness score 73/100
• Communicating palliative hope in late-stage dementia: thematic analysis of hope work in care plan meetings with nursing home residents’ families The Journals of Gerontology Series B · 2025-06-06 · usefulness score 37/100
• Developing and validating a prediction tool for cerebral amyloid angiopathy neuropathological severity Alzheimer s & Dementia · 2025-03-01 · usefulness score 49/100

Similar studies

• Combining early markers strongly predicts conversion from mild cognitive impairment to Alzheimer's disease 2008 · Hippocampus size on brain scans
• Research progress in predicting the conversion from mild cognitive impairment to Alzheimer’s disease via multimodal MRI and artificial intelligence 2025 · Digital health tools
• Neuroimaging markers for the prediction and early diagnosis of Alzheimer's disease dementia 2011 · Mild cognitive impairment
• Predicting Progression from Mild Cognitive Impairment to Alzheimer's Dementia Using Clinical, MRI, and Plasma Biomarkers via Probabilistic Pattern Classification 2016 · Mild cognitive impairment
• Multimodal Detection of Agitation in People With Dementia in Clinical Settings: Observational Pilot Study 2025 · Digital health tools
• Multi-modal data analysis for early detection of alzheimer’s disease and related dementias 2025 · Getting a diagnosis

More on Brain tissue and pathology

• A review that covers this ground Prognosis of depression in old age compared to middle age: a systematic review of comparative studies The American journal of psychiatry · 2005-09-01 · usefulness score 53/100
• Newest in this subject Inflammatory biomarker candidates in dementia with Lewy bodies: diagnostic classification, clinical associations, and survival Research Square · 2026-11-01 · Preprint — not yet peer-reviewed · usefulness score 16/100
• Newest in this subject TMEM106B C-Terminal fragment aggregation exacerbates α-synuclein and TDP-43 proteinopathy phenotypes bioRxiv · 2026-10-30 · Preprint — not yet peer-reviewed · usefulness score 16/100

See all studies in Brain tissue and pathology →