# Generalization of the Conformity Index for Multi-Target Radiotherapy Plans

## TL;DR

This paper introduces a new way to calculate the Conformity Index for radiotherapy plans involving multiple targets, improving accuracy by avoiding dose spillover effects.

## Contribution

A redefined Target Volume parameter that eliminates dose spillover interference in multi-target radiotherapy conformity index calculations.

## Key findings

- The new VTV calculation produced significantly lower values compared to traditional methods in breast cancer and NPC plans.
- The new CI values were higher and more reflective of true dose conformity in multi-target scenarios.
- For PTV1 in NPC, the new and traditional formulas produced identical results.

## Abstract

This study proposes a generalized method to calculate the Conformity Index (CI) for multi-target radiotherapy plans (e.g., breast or nasopharyngeal cancer). Standard CI formulas are often distorted in these complex scenarios because they erroneously include dose spillover from adjacent targets. To address this, we redefined the Target Volume (VTV) parameter to mathematically isolate the prescription dose region of each specific target. Validation on clinical plans demonstrated that the new formula effectively eliminates interference from neighboring targets, providing CI values that accurately reflect the true dose conformity. This improved calculation is recommended for the objective evaluation of multi-target radiotherapy plans.

Background and Purpose: Based on the distortion of the current conformity index (CI) formula in handling multi-target plans, the VTV parameter in the current CI formula has been redefined to more accurately calculate the CI value of multi-target plans, providing a reference for clinical applications. Methods and Materials: Considering the limitations of the current VTV calculation formula in CI proposed by van’t Riet and Paddick, a new VTV has been defined to better reflect the true conformity of the target volume in multi-target planning. We selected 15 breast cancer (BC) plans with PTVsc and PTVcw as the target volumes, and 15 nasopharyngeal carcinoma (NPC) plans with PTVp, PTVn, PTVrpn, PTV1, and PTV2 as target volumes. VTVnew and CInew were calculated using the proposed formulas, while VTVold and CIold were calculated using traditional formulas based on van’t Riet and Paddick. A paired, two-tailed Wilcoxon signed-rank test was conducted to compare VTVnew with VTVold, and CInew with CIold across all target volumes. Pearson’s correlation analysis was performed between CInew and CIold. Results: For BC, the VTV values calculated by the two methods for PTVsc and PTVcw showed statistically significant differences; the values calculated in this study were significantly lower than those calculated by van’t Riet and Paddick (p < 0.05). Consequently, the CInew values for BC were significantly higher than CIold. These results were consistent with those for PTVp, PTVn, PTVrpn, and PTV2 in NPC. For PTV1 in NPC, the results calculated by the two formulas were identical. Conclusions: The new VTV calculation formula eliminates the influence of dose spillage from adjacent targets, retaining only the prescription dose range of the specific target under analysis. This makes the calculated CI more reflective of true conformity compared to traditional formulas. We recommend using the proposed formula to calculate CI values for multi-target plans such as those for BC and NPC.