Comparison of Two Mathematical Models of Cellularity Calculation

Area of Science:

• Oncology
• Radiology
• Medical Imaging

Background:

• Functional MRI, including diffusion-weighted imaging (DWI) and dynamic-contrast enhanced MRI (DCE MRI), shows promise in characterizing tumor cellularity and vascularity.
• Previous research proposed two formulas based on a logistic tumor growth model to predict tumor cellularity using DWI and DCE MRI data.

Purpose of the Study:

• To validate two proposed formulas for predicting tumor cellularity in head and neck squamous cell carcinomas.
• To assess the predictive accuracy of these formulas against histopathological findings.

Main Methods:

• The study included 16 patients with head and neck squamous cell carcinomas (14 men, 2 women; mean age 57.0 ± 7.5 years).
• Tumor cellularity was calculated using two distinct formulas: one based on apparent diffusion coefficient (ADC) values alone, and another incorporating ADC and V_e values.
• Histopathological analysis provided an average cell count per high-power field for comparison.

Main Results:

• No significant correlation was found between tumor cellularity calculated using the ADC-only formula and histopathological cell counts.
• A moderate positive correlation (r=0.515, P=.041) was observed between cellularity predicted by the formula including ADC and V_e values and the histopathological estimates.
• This suggests the combined formula is more sensitive in predicting tumor cellularity.

Conclusions:

• The formula incorporating both apparent diffusion coefficient (ADC) and V_e values demonstrates superior sensitivity in predicting tumor cellularity compared to the formula relying solely on ADC values.
• These findings support the potential of combined DWI and DCE MRI parameters for non-invasively assessing tumor cellularity in head and neck cancers.