سامانه همایش‌ها TSTA | ثبت‌نام و ارسال مقاله کنفرانس‌ها و کنگره‌های ملی

صفحه اصلی / سی و دومین کنفرانس ملی و دهمین کنفرانس بین المللی مهندسی زیست پزشکی ایران

Mitigating MRI Domain Shift in Sex Classification: A Deep Learning Approach with ComBat Harmonization

نویسندگان :

Peyman Sharifian¹ Mohammad Saber Azimi² Masoud Noroozi³ Alireza Karimian⁴ Hossein Arabi⁵

1- دانشگاه اصفهان 2- دانشگاه شهید بهشتی 3- دانشگاه اصفهان 4- دانشگاه اصفهان 5- دانشگاه ژنو

کلمات کلیدی :

Deep Learning،Sex Classification،Combat Harmonization،Domain Adaptation،Magnetic Resonance Imaging

چکیده :

Deep learning models for medical image analysis often suffer from performance degradation when applied to data from different scanners or protocols, a phenomenon known as domain shift. This study investigates this challenge in the context of sex classification from 3D T1-weighted brain magnetic resonance imaging (MRI) scans using the IXI and OASIS3 datasets. While models achieved high within-domain accuracy (around 0.95) when trained and tested on a single dataset (IXI or OASIS3), we demonstrate a significant performance drop to chance level (about 0.50) when models trained on one dataset are tested on the other, highlighting the presence of a strong domain shift. To address this, we employed the ComBat harmonization technique to align the feature distributions of the two datasets. We evaluated three state-of-the-art 3D deep learning architectures (3D ResNet18, 3D DenseNet, and 3D EfficientNet) across multiple training strategies. Our results show that ComBat harmonization effectively reduces the domain shift, leading to a substantial improvement in cross-domain classification performance. For instance, the cross-domain balanced accuracy of our best model (ResNet18 3D with Attention) improved from approximately 0.50 (chance level) to 0.61 after harmonization. t-SNE visualization of extracted features provides clear qualitative evidence of the reduced domain discrepancy post-harmonization. Cross-domain balanced accuracy improved from ~0.50 to 0.61 after ComBat, a modest yet meaningful gain that moves the model from chance-level failure toward more reliable generalization while remaining below clinical utility. This work underscores the critical importance of domain adaptation techniques for building robust and generalizable neuroimaging AI models.

لیست مقالات

لیست مقالات بایگانی شده

بررسی دیدگاه ها و سیر تاریخی پیرامون حسابداری دارایی های موروثی

فاطمه رفیعی - مهدی ناظمی اردکانی

Recovery of Hand Motor Function in Children with ‎Hemiparetic Cerebral Palsy using an Interactive ‎Computer Game Combined with Mechanoreceptor ‎Stimulation: A Pilot Study

Mobina Ramezani - Shahed Salehzehi - Parisa Hosseini - Zahra Alizadeh Sarborj - Ghazal Mohammadzadeh - Hamid Reza Kobravi - Narges Hashemi - Mehran Beiraghi Toosi - Javad Akhondian

Design and Biomechanical Comparison of a Patient-Specific Anatomical Plate Versus Conventional Plate for Distal Humerus Fractures: A Finite Element Analysis

Fahime Rezazade - Azadeh Ghouchani - Maryam Amoochi