ORIGINAL RESEARCH article
Front. Aging Neurosci.
Sec. Parkinson’s Disease and Aging-related Movement Disorders
This article is part of the Research TopicImaging and Electrophysiology in the Diagnosis and Treatment of Parkinson's DiseaseView all articles
MultimodalCNN-PD: A Parkinson's Disease Diagnostics Framework Using Multimodal Convolutional Neural Network
Provisionally accepted
- 1Huazhong University of Science and Technology, Wuhan, China
- 2Yancheng No 1 People's Hospital, Yancheng, China
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Parkinson's disease (PD) is a prevalent neurodegenerative disorder that severely affects motor and cognitive functions, with early diagnosis, particularly during the prodromal phase, being critical for effective intervention. This study presents MultimodalCNN-PD++, a deep learning model that integrates Magnetic Resonance Imaging (MRI) with clinical metadata, including motor and cognitive assessments, demographic data, and genetic biomarkers, to enhance PD classification. The model employs a lightweight EfficientNet-B0 backbone, Mobile Convolutional Block Attention Modules (Mobile CBAM), and an enhanced Meta-Guided Cross-Attention (MGCA++) mechanism to improve classification accuracy and computational efficiency. A three-stage hierarchical feature selection method identifies the most discriminative clinical features, while clinical metadata is processed with BioClinicalBERT using Low-Rank Adaptation (LoRA). Validated on the Parkinson's Progression Markers Initiative (PPMI) dataset, the model achieved a 97.5% accuracy in distinguishing Normal Control, prodromal PD, and diagnosed PD cases. It also demonstrated reduced parameters and computational costs, making it more efficient for clinical use. External validation on the OASIS-3 dataset confirmed robust generalizability with 96.2% accuracy, despite variations in demographics and data acquisition protocols. Ablation studies highlighted the contributions of Mobile CBAM, MGCA++, hierarchical feature selection, and BioClinicalBERT-LoRA. This framework sets a new benchmark for multiclass PD diagnosis, showing strong potential as a clinically deployable AI tool for early detection and personalized management of neurodegenerative diseases.
Keywords: Clinical metadata, deep learning, early diagnosis, machine learning, MRI, multimodal CNN, neural networks, Parkinson's disease
Received: 27 Oct 2025; Accepted: 31 Jan 2026.
Copyright: © 2026 Muhammad Ibrahim, Zhi, Liu, Wang and Meng. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence:
Umar Muhammad Ibrahim
Chengjie Meng
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.