Software comparison for the analysis of cheese eyes in X-ray computed tomography

Cheese eyes in traditional Swiss-type cheese are an 'eye-catcher' and therefore a key factor for quality. In contrast to the widespread X-ray radiography, in the computed tomography (CT) system, the density difference between gas and the cheese body is imaged three-dimensionally. To quantify the eye volume and distribution in the cheese matrix, three major commercial software tools and a Python script were compared by measuring 12 cheeses with increasing numbers of cheese eyes. The number of cheese eyes was controlled by producing the 12 hard cheeses using an increasing number of plastic balls of 10 and 20 mm diameter. Although the agreement among the different software products was very high (correlation between void volume of the introduced balls and the calculated eye volume ranges between 0.9990 and 0.9999), the accessibility of different properties varies among the different software products, and systematic trends in accuracy could be identified.