Technology Category
Computer Vision: Multi-modal VisionMachine Learning: Deep Neural Architectures and Foundation ModelsKnowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
Placenta Accreta Spectrum Detection Using an MRI-based Hybrid CNN-Transformer Model
🤖 AI Summary
MRI diagnosis of placenta accreta spectrum (PAS) disorders suffers from substantial inter-rater variability among radiologists, resulting in low diagnostic consistency and significant clinical challenges. To address this, we propose the first dual-path 3D deep learning architecture that synergistically integrates a 3D DenseNet121 branch—capturing fine-grained local texture features—with a 3D Vision Transformer branch—modeling global spatial contextual relationships—thereby enhancing discriminative robustness. The model is trained on a large-scale, retrospective, multicenter dataset of 3D MRI volumes and rigorously evaluated via five-fold cross-validation. On an independent external test set, it achieves a mean classification accuracy of 84.3%, outperforming standalone 3D CNN and 3D ViT baselines by 4.7% and 3.9%, respectively. The framework demonstrates strong generalizability and holds promise for clinical decision support. This work establishes a novel, interpretable, and highly robust paradigm for intelligent PAS imaging diagnosis.
Technology Category
Application Category
📝 Abstract
Placenta Accreta Spectrum (PAS) is a serious obstetric condition that can be challenging to diagnose with Magnetic Resonance Imaging (MRI) due to variability in radiologists' interpretations. To overcome this challenge, a hybrid 3D deep learning model for automated PAS detection from volumetric MRI scans is proposed in this study. The model integrates a 3D DenseNet121 to capture local features and a 3D Vision Transformer (ViT) to model global spatial context. It was developed and evaluated on a retrospective dataset of 1,133 MRI volumes. Multiple 3D deep learning architectures were also evaluated for comparison. On an independent test set, the DenseNet121-ViT model achieved the highest performance with a five-run average accuracy of 84.3%. These results highlight the strength of hybrid CNN-Transformer models as a computer-aided diagnosis tool. The model's performance demonstrates a clear potential to assist radiologists by providing a robust decision support to improve diagnostic consistency across interpretations, and ultimately enhance the accuracy and timeliness of PAS diagnosis.
Problem
Research questions and friction points this paper is trying to address.
Automated detection of Placenta Accreta Spectrum from MRI scans
Addresses diagnostic inconsistency due to radiologist interpretation variability
Improves accuracy and timeliness of PAS diagnosis using deep learning
Innovation
Methods, ideas, or system contributions that make the work stand out.
Hybrid 3D DenseNet121 and Vision Transformer model
Automated detection from volumetric MRI scans
Combines local features with global spatial context
🔎 Similar Papers
No similar papers found.
