Software-based de-filtering restores quantitative accuracy in Clarity2D-enhanced whole-body bone scintigraphy.

Area of Science:

• Nuclear Medicine
• Medical Imaging Analysis
• Computational Pathology

Background:

• Whole-body scintigraphy is crucial for quantitative analysis of bone metastases.
• Noise-reduction filters like Clarity2D can improve image quality but may distort quantitative metrics.
• Restoring quantitative accuracy after filtering is essential for reliable clinical interpretation.

Purpose of the Study:

• To evaluate the efficacy of a software-based de-filtering algorithm in restoring quantitative accuracy of bone scan index (BSI) and hot spot number (HSn).
• To assess the impact of de-filtering on lesion detection accuracy and spatial concordance in whole-body scintigraphy images.
• To determine if de-filtering can mitigate distortions caused by the Clarity2D noise-reduction filter.

Main Methods:

• Retrospective analysis of 101 patients undergoing 99mTc-HMDP whole-body scintigraphy on a CZT SPECT/CT system.
• Comparison of unfiltered, Clarity2D-filtered, and software de-filtered images using quantitative indices (BSI, HSn) and lesion masks.
• Evaluation of agreement using Pearson correlation, Bland-Altman analysis, Dice coefficient, and Hausdorff distance.

Main Results:

• Clarity2D filtering significantly impaired quantitative concordance for BSI (r=0.23) and HSn (r=0.23).
• De-filtering successfully restored quantitative concordance (BSI r=0.99, HSn r=0.98) and improved spatial overlap (Dice 0.40 to 0.82).
• De-filtered images showed superior lesion detection accuracy (F1-score 0.78) and stability compared to Clarity2D-filtered images.

Conclusions:

• Software-based de-filtering reliably reverses Clarity2D-induced quantitative distortions in whole-body scintigraphy.
• This technique enables accurate BSI and HSn measurements and robust lesion detection without additional radiation or acquisition time.
• The de-filtering algorithm preserves the integrity of quantitative analyses, potentially broadening the clinical use of noise-reduction filters.