Non-essential element concentrations in brown grain rice: Assessment by advanced data mining techniques

Area of Science:

• Environmental Chemistry
• Food Safety
• Analytical Chemistry

Background:

• Non-essential elements in food can pose health risks.
• Rice is a global staple, making its elemental composition critical for safety.
• Geographical origin can significantly impact crop elemental profiles due to soil and environmental factors.

Purpose of the Study:

• To quantify 17 non-essential elements in brown grain rice from Corrientes, Argentina.
• To assess the influence of geographical origin on non-essential element profiles in rice.
• To evaluate the effectiveness of various chemometric methods for classifying rice by origin.

Main Methods:

• Inductively coupled plasma mass spectrometry (ICP-MS) was used for elemental quantification.
• A validated analytical method ensured accurate and reliable measurements.
• Machine learning algorithms including LDA, k-NN, PLS-DA, SVM, and RF were employed for classification.

Main Results:

• Concentrations of most non-essential elements were low or undetectable.
• Detected levels were comparable to international rice studies.
• Random Forests (RF) and Support Vector Machine (SVM) achieved 96% accuracy in classifying rice by geographical origin.

Conclusions:

• Geographical origin is a significant factor in the non-essential element composition of rice.
• Advanced classification methods, particularly RF and SVM, are effective for determining rice provenance based on elemental data.
• This research contributes to understanding rice elemental variability and potential traceability applications.