This release introduces significant enhancements to the ASL (Arterial Spin Labeling) processing pipeline, focusing on improved image rescaling, memory management, and expanded support for calibration options. These updates are particularly relevant for high-precision neuroimaging workflows like those utilized by the LOAMRI group and similar academic research environments.
🚀 New Features & Enhancements
ASL Mapping & Data Handling
- Average Support: Introduced the
average M0parameter in theASLDataclass. This allows for more flexible CBF (Cerebral Blood Flow) quantification by enabling the use of averaged equilibrium magnetization data. - Ultra-Long TE ASL Improvements: Enhanced the
UltraLongTE_ASLMappingmodule with updated parameter initialization and more robust image limit adjustments. - Memory Efficiency: Refactored output map creation across
CBFMappingand other modules to useclone_image. This improves memory handling and ensures that metadata and header information are preserved accurately during map generation.
Image Processing Refinement
- Enhanced Outlier Removal: Improved the
_adjust_image_limitsmethod to specifically rescale values. This results in cleaner data by better identifying and handling statistical outliers in the neuroimaging volumes.
🧪 Testing & Quality Assurance
- Validation Updates: Added new test cases for averaging in both
MultiTE_ASLMappingandASLData. - Stricter Assertions: Updated
test_multite_asl_object_create_map_successto ensure that output values are strictly greater than zero, preventing the generation of non-physical negative mapping values. - Refactored Tests: Renamed test functions for
UltraLongTE_ASLMappingto follow consistent naming conventions, making the CI/CD pipeline easier to debug.
🛠️ Maintenance & Documentation
- Version Bump: Significant update from
v1.0.1throughv1.0.2to the currentv1.1.0. - Documentation: Fixed typos in
TODOcomments regarding map dimensions and updatedFUNDING.ymlto reflect current project support.
📦 Technical Summary
- Tag:
v1.1.0 - Repository:
LOAMRI - Primary Focus: ASL quantification, memory optimization, and outlier robustification.