Role of Apparent Diffusion Coefficient Map-Based First- and High-Order Radiomic Features for the Discrimination of...

Area of Science:

• Radiology
• Oncology
• Medical Imaging

Background:

• Chondrosarcomas typically arise from the lateral pelvis.
• Midline chondrosarcomas (10%) present diagnostic challenges due to imaging similarities with chordoma.
• Accurate differentiation is crucial for appropriate treatment planning.

Purpose of the Study:

• To evaluate the diagnostic accuracy of apparent diffusion coefficient (ADC)-based radiomic features for differentiating sacral chordomas and chondrosarcomas.
• To assess the utility of two novel diffusion indices: restriction index (RI) and facilitation index (FI).

Main Methods:

• Retrospective review of 82 pelvic MRIs (42 chordomas, 40 chondrosarcomas) from December 2014 to September 2021.
• Segmentation of lesions on ADC maps, extraction of 8 first-order ADC metrics, RI, and FI.
• Analysis of 107 radiomic features, including first-order and high-order features, for group differences.

Main Results:

• Good to excellent interobserver reliability for 8 of 10 ADC metrics.
• Significant differences (P < .005) observed for RI, FI, median ADC, mean ADC, and skewness.
• Optimal cutpoints identified for diagnosis: RI > 0.015, FI < 0.25, mean ADC < 1.7 × 10^-3 mm²/s, skewness > 0.177. FI was optimal for a decision tree.
• Secondary analysis revealed 18 significant radiomic features differentiating the two tumors.

Conclusions:

• The novel ADC index, FI, along with mean ADC, skewness, and 12 high-order radiomic features, can aid in differentiating sacral chordomas from chondrosarcomas.
• These quantitative imaging biomarkers offer potential for improved diagnostic accuracy in challenging cases.
• Further validation in larger cohorts is warranted.