Rethinking Medical Anomaly Detection in Brain MRI: An Image Quality Assessment Perspective

2024-08-15

arXiv.org

AI Summary

To address the limitation of conventional reconstruction losses (e.g., L1) in capturing subtle pathological deviations in brain MRI anomaly detection, this work pioneers the integration of image quality assessment (IQA) principles into anomaly discrimination. We propose a SSIM-L1 hybrid loss to jointly optimize structural fidelity and pixel-level accuracy; introduce an average intensity ratio (AIR)-enhanced preprocessing strategy to improve contrast between normal and abnormal regions; and develop an IQA-driven anomaly localization mechanism. Evaluated on BraTS21 (T2/FLAIR) and MSULB datasets, our method achieves over 10% improvement in Dice score and area under the precision-recall curve (AUPRC) over state-of-the-art approaches, notably enhancing detection sensitivity for small lesions. This work establishes a novel IQA-informed paradigm for medical image anomaly detection.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Reconstruction-based methods, particularly those leveraging autoencoders, have been widely adopted to perform anomaly detection in brain MRI. While most existing works try to improve detection accuracy by proposing new model structures or algorithms, we tackle the problem through image quality assessment, an underexplored perspective in the field. We propose a fusion quality loss function that combines Structural Similarity Index Measure loss with l1 loss, offering a more comprehensive evaluation of reconstruction quality. Additionally, we introduce a data pre-processing strategy that enhances the average intensity ratio (AIR) between normal and abnormal regions, further improving the distinction of anomalies. By fusing the aforementioned two methods, we devise the image quality assessment (IQA) approach. The proposed IQA approach achieves significant improvements (>10%) in terms of Dice coefficient (DICE) and Area Under the Precision-Recall Curve (AUPRC) on the BraTS21 (T2, FLAIR) and MSULB datasets when compared with state-of-the-art methods. These results highlight the importance of invoking the comprehensive image quality assessment in medical anomaly detection and provide a new perspective for future research in this field.

Problem

• Proposing fusion quality metric for medical anomaly detection
• Amplifying divisive discrepancies between normal and abnormal regions
• Enhancing brain MRI anomaly detection via image quality assessment

Innovation

• Proposes fusion quality metric combining SSIM and l1
• Designs AIR-based data transformation amplifying divisive discrepancies
• Devises IQA approach integrating metric and transformation components