Early Knee Osteoarthritis Detection by Multi-Component T₂ Mapping

Area of Science:

• Biomedical imaging
• Osteoarthritis research
• Quantitative MRI

Background:

• Early detection of knee osteoarthritis (OA) is crucial for timely intervention.
• Conventional mono-exponential T2 mapping has limitations in identifying subtle early cartilage changes.
• Multi-component T2 models offer potential for improved sensitivity to tissue alterations.

Purpose of the Study:

• To evaluate the efficacy of multi-component T2 mapping (bi-exponential and stretched-exponential models) for early knee OA detection.
• To compare the diagnostic performance of these models against conventional mono-exponential T2 mapping.
• To assess the impact of sub-regional analysis on diagnostic accuracy.

Main Methods:

• T2 relaxation maps were generated using mono-exponential, bi-exponential, and stretched-exponential models.
• Data were acquired from 26 early-stage knee OA patients and 26 healthy controls.
• Quantitative T2 parameters were extracted from six cartilage sub-regions and analyzed using linear discriminant analysis after age adjustment.

Main Results:

• Global whole-cartilage analysis showed limited diagnostic performance across all models (AUC ≤ 0.65).
• Sub-regional analysis significantly improved classification accuracy, highlighting regional heterogeneity in early OA.
• The bi-exponential T2 model achieved the highest AUC (0.68) in sub-regional analysis, outperforming SE (0.60) and ME (0.51) models.

Conclusions:

• Multi-component T2 mapping, especially with bi-exponential modeling at the sub-regional level, enhances sensitivity to early knee OA.
• Sub-regional analysis is critical for detecting localized cartilage degeneration in early OA.
• This advanced T2 mapping technique shows potential as a noninvasive imaging biomarker for early knee OA detection.