Predicting obesity using longitudinal near infra-red spectroscopy (NIRS) data

Citation

Abdullah A, Hussain A & Khan IH (2017) Predicting obesity using longitudinal near infra-red spectroscopy (NIRS) data. In: ICCDA '17 Proceedings of the International Conference on Compute and Data Analysis. ICCDA '17: International Conference on Compute and Data Analysis, Lakeland, FL, USA, 19.05.2017-23.05.2017. New York: ACM, pp. 123-128. https://doi.org/10.1145/3093241.3093286

Abstract
Globally there has been a dramatic increase in obesity [1]. Thus understanding, predicting and managing obesity has the potential to save lives and billions. Behavioral studies suggest that binging by obese persons is prompted by inflated brain reward center activity to stimuli linked with high-calorie foods [2], but there are hardly any data-analytic calorie-based cognitive studies using non-invasive Near-Infrared Spectroscopy (NIRS) data that predict obesity using predictive data mining. In this paper, details of a novel research methodology are presented for a 24-month longitudinal NIRS study in natural subject environments. The proposed methodology is based on brain reward center activation mapping, simulated results of Naïve Bayes modeling using these activation maps demonstrate how cerebral functional activity data can be used to predict obesity in the non-obese.

Keywords
Prediction; data mining, obesity, NIRS, Naïve Bayes; paired t-test