SPECT/CT Radiomics for Differentiating between Enchondroma and Grade I Chondrosarcoma

Area of Science:

• Medical Imaging
• Radiology
• Oncology

Background:

• Differentiating enchondromas from atypical cartilaginous tumors (ACTs) is crucial for appropriate patient management.
• Current diagnostic methods may have limitations in distinguishing these bone lesions.
• Radiomics analysis of SPECT/CT scans offers a novel approach for quantitative assessment.

Purpose of the Study:

• To evaluate the efficacy of SPECT/CT radiomics parameters in differentiating enchondromas from ACTs in long bones.
• To identify specific radiomics features predictive of ACTs.

Main Methods:

• Retrospective analysis of quantitative HDP SPECT/CT data from 49 patients with enchondromas or ACTs.
• Patients were divided into training (n=32) and testing (n=17) datasets.
• LASSO regression was used for feature selection, followed by multivariate analysis to assess predictive value.

Main Results:

• Two radiomics features, zone-length non-uniformity (ZLNU_GLZLM) and coarseness (Coarseness_NGLDM), were selected.
• Higher ZLNU_GLZLM was identified as the sole significant independent predictor of ACTs.
• In the test dataset, ZLNU_GLZLM demonstrated high sensitivity (83.3%) and specificity (90.9%) for ACT diagnosis.

Conclusions:

• HDP SPECT/CT radiomics, particularly ZLNU_GLZLM, shows significant potential in differentiating enchondromas from ACTs.
• This quantitative imaging approach may enhance diagnostic accuracy for bone tumors.
• Further validation in larger cohorts is warranted.