• DR.KANIGIRI BHARATHI

Neuro-degenerative disorders such as Alzheimer’s, Parkinson’s, and ALS are characterized by progressive and irreversible cognitive and motor decline. Early-stage detection remains a significant clinical challenge due to heterogeneous symptoms, gradual disease progression, lack of biomarkers, and limited availability of labeled medical data. Existing machine learning models typically rely on single-modal sources (e.g., MRI or clinical records), resulting in reduced sensitivity to early pathological changes and poor real-world generalization. This research proposes a multi-modal deep learning framework that integrates structural MRI, genetic data, speech/audio biomarkers, and clinical history to enable accurate and interpretable early detection of neuro-degenerative disorders. The framework utilizes heterogeneous data fusion techniques, combining 3D-CNN based visual encoders, transformer-based sequence encoders, and graph neural networks to learn complementary representations from multiple modalities. Further, novel cross-modal attention and missing-modality compensation mechanisms are proposed to ensure robustness when one or more data sources are incomplete, noisy, or unavailable. To address the scarcity of labeled data, the research incorporates self-supervised pretraining, generative augmentation (e.g., diffusion models), and transfer learning. Predictive uncertainty estimation and explainability are integrated into the model to support clinically trustworthy decision-making. The proposed system will be evaluated on large-scale dementia datasets, clinical speech corpora, and genomics repositories, benchmarking against state-of-the-art uni-modal and multi-modal baselines. The expected outcome is a computationally efficient, clinically interpretable, and generalizable deep learning framework capable of detecting early neuro-degenerative changes years before traditional diagnosis, enabling timely intervention and improved patient outcomes. KEYWORDS: Neuro-degenerative disorders, multi-modal learning, deep learning, MRI, speech, genomics, graph neural networks, transformers, early diagnosis.

KEYWORDS: Neuro-degenerative disorders, multi-modal learning, deep learning, MRI, speech, genomics, graph neural networks, transformers, early diagnosis.

"MULTI-MODAL DEEP LEARNING FRAME WORK FOR EARLY DETECTION OF NEURO-DEGENERATIVE DISORDERS", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 12, page no.d51-d59, December-2025, Available :http://www.jetir.org/papers/JETIR2512309.pdf

"MULTI-MODAL DEEP LEARNING FRAME WORK FOR EARLY DETECTION OF NEURO-DEGENERATIVE DISORDERS", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 12, page no. ppd51-d59, December-2025, Available at : http://www.jetir.org/papers/JETIR2512309.pdf