Eating quality evaluation of Khao Dawk Mali 105 rice using near-infrared spectroscopy

• Published in: Food science & technology 2017-06, Vol.79, p.70-77
• Main Authors: Siriphollakul, Pornarree, Nakano, Kazuhiro, Kanlayanarat, Sirichai, Ohashi, Shintaroh, Sakai, Ryosuke, Rittiron, Ronnarit, Maniwara, Phonkrit
• Format: Article
• Language: English
• Subjects: Amylose; Jasmine rice; NIR; Oryza sativa
• ISSN: 0023-6438; 1096-1127
• DOI: 10.1016/j.lwt.2017.01.014

Eating quality evaluation of Khao Dawk Mali 105 rice (KDML105) based on near infrared spectroscopy (NIRS) of single kernels was developed to measure the amylose content of uncooked rice, and texture of cooked rice. The rice samples were scanned using near infrared transmittance spectrometry over the wavelengths of 940–2222 nm before cooking. Calibration models of amylose content and cooked rice texture were generated by partial least squares (PLS) regression based on first derivative upon logarithms of transmittance. The PLS regression for amylose content (AC) which were expressed as coefficients of determination (R2) were 0.95 and 0.92 for calibration and prediction, respectively. Root mean square error of prediction (RMSEP) was 9.9 g/kg, dry weight. The texture of cooked rice was expressed in springiness (H1), resilience (A1), deformation (H2) and cohesiveness (A2) from low and high compression tests. The PLS prediction results (R2pre) for H1, A1, H2 and A2 were 0.61, 0.86, 0.87 and 0.91, respectively. The RMSEP (and bias) were 0.03 (0.004), 0.01 (0.001), 0.02 (0.005) and 0.01 (0.000), correspondingly. The validity of each calibration model was statistically evaluated. The use of NIRS was feasible to predict amylose content of uncooked rice, and eating quality (texture) of cooked rice before cooking. •Amylose content and texture of Thai rice are nondestructively determined.•Single kernel transmission measurement gives the best result for amylose prediction.•T-squared score and variable importance projection effectively describe PLS modeling.•Eating quality of cooked rice is feasibly predicted using NIRS prior to cooking.