Non-invasive protein and free fatty acid prediction in pulse flours using hyperspectral imaging and PLSR modelling
A non-destructive analytical technique combining hyperspectral imaging with partial least squares regression to predict protein and free fatty acid content in yellow pea and chickpea flours. Eliminates the need for chemical reagents, sample destruction, and intensive labor while delivering rapid, accurate nutritional assessment for pulse-based food products.
Overview
This research proposes a non-invasive analytical approach for predicting protein and free fatty acid content in yellow pea and chickpea flours using hyperspectral imaging combined with partial least squares regression (PLSR) modelling. The technology addresses key limitations of conventional nutritional analysis methods, which typically require sample destruction, chemical reagents, and significant labor investment. By capturing spectral fingerprints in the near infrared range, the technique enables rapid nutritional characterization of pulse flours without altering the sample.
The approach builds on prior success with rocket leaves, where hyperspectral imaging achieved prediction accuracies of r² = 0.76–0.78 for phytonutrient mapping with low root mean square errors. Spectral analysis of pulse flour samples revealed dominant protein peaks, indicating strong potential for protein content prediction. Data preprocessing techniques are expected to enable accurate free fatty acid prediction as well.
Technical specifications
Key features:
- Hyperspectral imaging captures spectral data in the near infrared range across 150 samples per pulse variety (chickpea and yellow pea)
- Protein content validated using an FP628 LECO nitrogen/protein analyzer as the reference method
- Free fatty acid content validated using Soxhlet extraction
- PLSR modelling performed in MATLAB 2019b to correlate spectral information with analytical measurements
- External prediction validation ensures calibration model robustness
- Final model selection based on lowest RMSE in both calibration and external prediction
Technology readiness level
The research is at an early-to-mid stage of development. Prior proof-of-concept work on rocket leaves demonstrated feasibility of hyperspectral imaging for phytonutrient prediction. The current proposal focuses on extending this approach to pulse flours, with systematic validation planned using 300 total flour samples. Once calibration and external validation are complete, the resulting models will be ready for application in non-invasive quality assessment of pulse flour products.
About University of Manitoba
The University of Manitoba is a comprehensive public research university in Winnipeg with a broad portfolio spanning discovery to translation. Industry engages through an on-campus research and technology park that colocates companies with university labs, and through a health sciences campus embedded in the city’s hospital district for clinical studies and clinical-grade prototyping. A well-established co-op and internship ecosystem and the region’s logistics hub position make collaboration, field trials, and pilot-scale testing accessible. Research is supported by competitive funding from NSERC, CIHR, SSHRC, and the Canada Foundation for Innovation. A dedicated technology transfer office provides IP strategy, licensing, and startup support with flexible agreements for corporate partners.
delivered to your inbox.