تشخیص بیماری پارکینسون با استفاده از شبکههای عصبی پیچیده عمیق با یادگیری انتقالی و افزایش دادهها
کد: G-1953
نویسندگان: Mohadeseh Montazeri ℗, Mahdieh Montazeri *
زمان بندی: زمان بندی نشده!
برچسب: پردازش سیگنال های پزشکی
دانلود: دانلود پوستر
خلاصه مقاله:
خلاصه مقاله
Background and Aims: Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterized by motor impairments such as tremors, rigidity, and bradykinesia. Early diagnosis of PD is essential for slowing the disease's progression and improving patient outcomes. Traditional diagnostic methods rely heavily on clinical evaluations, which are often subjective and time-consuming. This study aims to present a more accurate and efficient method for diagnosing PD by utilizing deep learning techniques and MRI scans, focusing on overcoming the limitations of existing diagnostic approaches. Method: This paper proposes a hybrid deep learning approach combining transfer learning and Generative Adversarial Networks (GANs) for PD diagnosis. We used MRI images from the Parkinson’s Progression Markers Initiative (PPMI) database, which contains 1,700+ MRI scans from over 500 PD patients and healthy controls, to train and evaluate the model. Transfer learning helps improve classification performance by leveraging knowledge from large-scale datasets despite limited labeled data for PD. Additionally, GAN-based data augmentation is applied to generate synthetic MRI images, which alleviates the issue of overfitting caused by small sample sizes. Results: Our method achieved 89.23% classification accuracy, with improvements in both sensitivity and specificity. The results indicate that combining transfer learning with GAN-based data augmentation outperforms traditional diagnostic methods, providing a more reliable and efficient solution for early-stage PD diagnosis. Conclusion: This approach could be implemented in clinical practice to assist healthcare professionals in diagnosing Parkinson’s disease earlier, potentially improving patient outcomes. Future work will explore expanding the dataset with more diverse MRI images and applying this method to other neurodegenerative diseases.
کلمات کلیدی
Parkinson, Convolutional Neural Networks, Transfer Learning
