EBONI: A tool for automated quantification of bone metastasis load in PSMA PET/CT

Objectives: PSMA PET/CT has a high diagnostic accuracy for lesion detection in metastatic prostate cancer (MPC), including bone metastases. Novel therapeutic approaches require valid biomarkers for standardized disease staging, progression and therapy response. Here, we present EBONI (Evaluation of Bone Involvement), a software to automatically quantify bone metastasis load in PSMA PET/CT. Lesion quantity, mean and maximum lesional standardized uptake value (SUV), Z-Score and percentage of affected bone volume are determined. Furthermore, SUV-weighted tumor volume is calculated for each single lesion. Methods: EBONI was developed in the MATLAB environment to ensure platform independence. The code basis is open source and the tool is made freely available. To validate EBONI, results of automated quantification of 38 PSMA PET/CT scans with different levels of bone involvement were compared to visual expert reading. Influence of SUV threshold (SUVT) and Hounsfield unit thresholds (HUT) were analyzed. Results: A very high correlation between bone lesion quantity as determined visually and automatically was found (SUV_max: r2=0.97, SUV_mean: r²=0.88, lesion count: r²=0.97). HUT had no significant influence while a SUVT of 2.5 proved optimal for automated lesion quantification. Non-specific uptake e.g. in abdominal organs and the urinary tract are eliminated systematically (Fig. 1). The amount of false positive tissue misclassification is low, occurred mainly around the salivary and lacrimal gland and could easily be corrected for. There were no false negative ratings

Conclusions: EBONI-analysis is robust, quick (<3 minutes per scan) and 100% reproducible. It allows rater independent quantification of bone metastasis in MPC. It provides lesion quantification equivalent to visual assessment and complementary information. It can be easily implemented as an add-on to visual analysis of PSMA PET/CT scans and has the potential of reducing turnaround time.

• Download figure
• Open in new tab
• Download powerpoint