Relationship between sensory textural attributes and near infrared spectra of cooked rice. An Abstract from Journal of Near Infrared Spectroscopy

Weena Srisawas,^a Vinod K. Jindal^a,c,* and Warunee Thanapase^b
^aFood Engineering and Bioprocess Technology, School of Environment, Resources and Development, Asian Institute of Technology, P.O. Box 4, Kloung Luang, Pathumthani 12120, Thailand. Email: jindal@ait.ac.t
^bKasetsart Agriculture and Agro-Industrial Product Improvement Institute, Kasetsart University, 50 Paholyothin Rd., Chatuchak, Bangkok 10900, Thailand

Fourteen varieties of Thai indica rice, cooked with five water-to-rice ratios ranging from 1.3 to 2.5 on a weight basis, were characterised by sensory and instrumental texture profile analysis. The potential of near infrared (NIR) reflectance spectroscopy was investigated as an alternative tool for evaluating eating quality attributes of cooked rice by developing predictive models for sensory hardness, stickiness and glossiness. Partial least squares regression models were developed which predicted sensory hardness and stickiness slightly better than the glossiness with r²_v values ranging from 0.88 to 0.91 and standard errors of prediction (SEP) lower than 0.4 unit score on nine-point sensory intensity scales. Results indicated that NIR spectroscopy-based models could be used for estimating the sensory hardness, stickiness and glossiness scores of cooked rice with higher accuracy (lower SEP) compared to the instrumental texture profile analysis based-models.