Systematic evaluation of one-dimensional-to-two-dimensional near-infrared spectroscopy transformations with deep learning for quantifying coconut sap adulteration.

Ravipat Lapcharoensuk; Agustami Sitorus

doi:10.1016/j.foodchem.2026.150310

Home/Archive/PubMed

PubMed · 42424738

Systematic evaluation of one-dimensional-to-two-dimensional near-infrared spectroscopy transformations with deep learning for quantifying coconut sap adulteration.

Ravipat Lapcharoensuk·Agustami Sitorus

View original source ↗Export metadata ↓

Abstract

Near-infrared (NIR) spectroscopy have limitations when combined with deep learning (DL) algorithms because they rely on low-dimensional datasets. Therefore, we investigated the potential of transforming one-dimensional (1D) NIR spectra into two-dimensional (2D) spectrograms using synchronous and asynchronous techniques and the continuous wavelet transform (CWT) and their effectiveness by integrating with DL for detecting adulteration in coconut sap. NIR spectra (12,500-4000 cm-1) were collected from binary mixtures (0%-100%;w/w). The performance of all DL (convolutional neural networks-CNN, AlexNet and ResNet) models was compared with that of partial least squares (PLS). The models were ranked in the mentioned order based on their performances: 2D-CWT > 2D-asynchronous > 2D-synchronous > 1D/2D-PLS. The important features of the best model can be explained and visualized using gradient-weighted-class-activation-mapping. The findings highlight that the 1D-to-2D NIR data transformation combined with DL is a highly robust approach because it addresses the feature representation gap in NIR data and effectively captures the spatial-spectral correlations.

Explore related subjects

Spectroscopy, Near-Infrared Deep Learning Cocos Food Contamination Convolutional Neural Networks Artificial intelligence Deep learning Food adulteration High-dimensional Model interpretability Journal Article Evaluation Study

Keep this discovery

BibTeX RIS

Ravipat Lapcharoensuk, Agustami Sitorus. 2026-07-04. Systematic evaluation of one-dimensional-to-two-dimensional near-infrared spectroscopy transformations with deep learning for quantifying coconut sap adulteration.. https://doi.org/10.1016/j.foodchem.2026.150310

Cite the original work for its findings. Save a collection to share your selection of sources.

Discover connections

Artificial Intelligence subject Deep Learning subject Food Contamination subject Journal Article subject

Connections use source metadata and explicit phrase matches, not verified experimental comparisons.

KEEP EXPLORING

Related citations

PubMed2025

Influence of soil types with different soil-forming process on the qualitative and quantitative detection of microplastics by near-infrared spectroscopy.

Microplastics (MPs) have become a pressing global environmental threat, with soils-acting as sinks for MPs from multiple sources-gaining increasing attention. Near-infrared (NIR) spectroscopy offers a promising tool for MPs detection due to its rapid, non-destructive, and field-applicable features. Although previous studies have focused on the effects of individual soil components on the NIR detection performance of MPs, there is still a lack of systematic research on how the complex background-formed by the coupling of multiple physicochemical properties in natural soils-affects detection performance. This study focuses on soil types with different soil-forming processes, selected five representative agricultural soils to systematically evaluate how the combinations of physicochemical properties they represented affect the performance of NIR-based qualitative and quantitative analysis of MPs in soils. The results demonstrated that soil type significantly affected both the spectral response and detection performance of MPs. Brown Pedocals and Brown Earth exhibited clearer characteristic absorption and stronger linear responses, achieving higher identification accuracy under low (<1.5 %) or zero MPs concentrations and the best quantitative performance (R2 ≥ 0.988, prediction set root mean square error (RMSEP) ≤ 0.110 %). In contrast, Phaeozem and Red Soil were more prone to misclassification at low concentrations, while Fluvo-aquic Soil showed the poorest quantitative performance. This study is the first to reveal, at a holistic level, the critical constraints posed by natural soil complexity on the NIR detection of MPs, offering targeted empirical evidence to support the application of NIR technology in real-world soil environments.

Soil↗

PubMed2026

Clinical efficacy and brain mechanism characteristics of guide chi and regulate spirit tuina therapy in the treatment of post-stroke walking dysfunction: A randomized controlled trial based on fNIRS.

BACKGROUND: This study aims to preliminarily evaluate the role of Guide Chi and Regulate Spirit(GCRS) Tuina in enhancing walking function in post-stroke patients with walking dysfunction; secondly, by using functional near-infrared spectroscopy (fNIRS), it investigates the effect of GCRS Tuina on the restoration of brain function in this patient population. METHODS: Participants in the control group received 4-week rehabilitation treatment, while those in the Combined Tuina Group (CTG) additionally received GCRS Tuina therapy for another 4 weeks on this basis. Functional Ambulation Category (FAC), Fugl - Meyer Assessment Scale for Lower Extremity Motor Function (FMA - LE), and Modified Barthel Index (MBI) were evaluated at the baseline and after 4 treatment weeks. A gait and motion analysis system was used to measure step length, stride, walking speed, and step frequency. FNIRS was used to measure the resting-state functional connectivity(FC) strength, as well as the β - value and HbO2 concentration mean during the walking task. RESULTS: A total of 60 participants completed the randomized, and 53 completed the trial and entered the statistical analysis. Compared with the Single Rehabilitation Group(SRG), the CTG group had higher FAC, FMA-LE, and MBI scores after 4 weeks. After the treatment course, the standardized step length, stride, walking speed, and step frequency of the CTG were higher than SRG. At the Region of Interest(ROI) level, the CTG exhibited 13 inter-ROI FC strengths that were higher than SRG, and the differences could survive the FDR correction (PFDR<0.05). Under the walking task, the CTG group had higher β values in 18 channels and higher Oxyhemoglobin(HbO2) concentrations in 19 channels than the SRG (PFDR <0.05). CONCLUSION: GCRS Tuina therapy can significantly improve patients' walking function, enhance lower limb motor ability and daily living ability. It can improve walking efficiency. Tuina can significantly increase the FC and enhance the activation levels and HbO2 concentrations. The stimulation of Tuina may help reconstruct the brain's motor control network, restore impaired motor function, promote the occurrence of neural plasticity, strengthen the neural circuits in the cognitive-motor-sensory cortex to improve walking function. TRIAL REGISTRATION: This study has been registered with the International Traditional Medicine Clinical Trial Registry (ITMCTR2024000654).

Humans↗

PubMed2026

Effects of acute resistance exercise on prefrontal oxygenation and task-switching performance: Considerations of loading strategies and blood flow restriction.

Although acute resistance exercise (RE) has been proposed to influence cognitive flexibility and underlying neural mechanisms, it remains unclear whether these effects vary across loading strategies and whether exercise-induced prefrontal hemodynamic responses translate into cognitive outcomes. The present study examined (1) prefrontal cortex (PFC) oxygenated hemoglobin (O2Hb) responses across exercise sets and conditions, (2) the effects of low-load (LL), LL with blood flow restriction (BFR), and high-load (HL) RE on task-switching performance, and (3) whether exercise-related PFC O2Hb responses were associated with pre- to post-exercise changes in task-switching performance. Thirty physically active adults completed three randomized, counterbalanced RE conditions consisting of four sets of barbell squats. LL was performed at 30% one-repetition maximum (1RM) with and without BFR, whereas HL was performed at 70% 1RM. Cognitive flexibility was assessed pre- and post-exercise using a modified Stroop task, indexed by switch-cost reaction time (RT) and accuracy. PFC O2Hb was assessed using functional near-infrared spectroscopy during exercise and expressed as changes from the resting baseline for each set (Sets 1-4). PFC O2Hb increased across sets, rising from Set 1 to Set 3 before plateauing, with no differences observed across conditions. Switch cost RT and accuracy did not improve from pre- to post-exercise, and no differences across conditions were detected. PFC O2Hb during the final set was not associated with changes in switch cost. These findings suggest that although acute RE elicits robust increases in prefrontal hemodynamic activity, such responses may not translate into acute improvements in cognitive flexibility.

Humans↗