Digital image analysis is a reliable methodology to assess starch pattern index in commercial apple cultivars

Assessment of starch pattern index (SPI) by visual comparison of iodine-stained transversely cut fruit with reference charts is widely accepted as an effective method for estimating apple fruit ripening stage, crucial for predicting the optimal harvest window for long-term storage. While the method itself is very simple and cost efficient, it has an obvious drawback of being subjective and requiring a prolonged training time of the personnel. A method based on digital-image-analysis of the iodine treated apple tissue for an automated estimation of SPI might be a viable and economical alternative for practical application. Researchers at the Research Center Laimburg in Italy have studied the variability, i.e., the subjectivity as well as the repeatability, of visual SPI assessment done by an expert panel was observed, aiming to obtain a valid reference for evaluating the acceptable deviation boundaries of the subsequent, instrumentally obtained SPI. The assessments were done with both 5- and 10-point SPI scale. The results revealed a high consistency and agreement between different panellists (ICC 2 > 0.9), with the responses deviating by 0.2 and 0.6 points on a 5-point and 10-point SPI scale, respectively, when considering mean values obtained by Bland-Altman analysis of agreement between measurement methods. The variability of the panellists when repeating own SPI estimation of the same disks after a period of 2 up to 10 months revealed an average repeatability within a range ± 0.3 units, with the discrepancies for individual samples amounting up to 1.1 units. The automated digital image analysis readings (Amilon, Automatic starch meter, Isolcell) were comparable to visual assessments, revealing high absolute agreement (ICC > 0.9), thereby confirming the interchangeability of both methods for SPI estimation. A user-friendly technique for objective determination of starch clearing index could potentially result in rapidization, simplification and cost-reduction of the starch index determination process.   SourceSadar, N.; Stürz, S.; Ebner, I.; Zanella, A. (2023). Digital image analysis-based assessment of starch pattern index in commercial apple cultivars: reliability and applicability. VII International Conference Postharvest Unlimited & XII International Symposium on Postharvest Quality of Ornamental Plants, Wageningen University & Research, May 14 to 17 2023