Analysis of inequalities and correction model of the Goh method used in quantifying interstitial lung disease

Abstract

Introduction

Interstitial lung disease is a significant complication in autoimmune diseases. The extent of involvement is crucial in determining both onset and treatment response. While there are computational methods for this calculation, manual methods, such as the Goh method, remain the most commonly used. However, it has some disadvantages, such as presuming that areas across the lung are equivalent and thus influence the calculation equally. The aim of this study is to analyse these difficulties and find a mathematical correction method.

Materials and methods

High-resolution chest tomography images were obtained from patients with and without ILD. Pulmonary areas were analysed and compared in the slices referenced in the Goh method. Through geometric analysis of lung structure, the slice closest to the average of its areas was identified for comparison with other measurements. Based on this, a correction equation was formulated, and through simulation of various interstitial involvement patterns, the degree of variation was determined.

Results

Images from 80 patients were analysed, mostly women diagnosed with systemic sclerosis. It was found that apical areas (slices 1 and 2) and the sum of left-sided pulmonary areas were 20% smaller compared to basal and right-sided areas, respectively. Simulation with different involvement components found that apical or left-sided disease overestimates extension by 5% to 10%, especially when it exceeds 15%. The third slice was determined to be closest to the average of the areas, and a correction formula was devised based on this.

Conclusions

The Goh method incorrectly assumes the equivalence of pulmonary areas, which can impact calculations. We propose a mathematical correction for this purpose.