Standard Error Of Prediction Values Research Articles

Development and validation of near-infrared spectroscopy procedures for prediction of cassava root dry matter and amylose contents in Ugandan cassava germplasm.

Cassava utilization for food and/or industrial products depends on inherent properties of root dry matter content (DMC) and the starch fraction of amylose content (AC). Accordingly, in the present study, near-infrared reflectance spectroscopy (NIRS) models were developed to aid breeding and selection of DMC and AC as critical industrial traits taking care of root sample preparation and cassava germplasm diversity available in Uganda. Upon undertaking calibrations and cross-validations, best models were adopted for validation. DMC in calibration samples ranged from 20 to 45 g 100g-1, whereas, for amylose content, it ranged from 14 to 33 g 100g-1. In the validation set, average DMC was 29.5 g 100g-1, whereas, for amylose content, it was 24.64 g 100g-1. For DMC, a modified partial least square regression model had regression coefficients (R2) of 0.98 and 0.96, respectively, in the calibration and validation set. These were also associated with low bias (-0.018) and ratio of performance deviation that ranged from 4.7 to 5.0. In addition, standard error of prediction values ranged from 0.9 g 100g-1 to 1.06 g 100g-1. For AC, the regression coefficient was 0.91 for the calibration set and 0.94 for the validation set. A bias equivalent to -0.03 and a ratio of performance deviation of 4.23 were observed. These findings confirm the robustness of NIRS in the estimation of dry matter content and amylose content in cassava roots and thus justify its use in routine cassava breeding operations. © 2023 The Authors. Journal of The Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

Near-Infrared Reflectance Spectrophotometry (NIRS) Application in the Amino Acid Profiling of Quality Protein Maize (QPM).

The accurate quantification of amino acids in maize breeding programs is challenging due to the high cost of analysis using High-Performance Liquid Chromatography (HPLC) and other conventional methods. Using the Near-Infrared Spectroscopic (NIRS) method in breeding to screen many genotypes has proven to be a fast, cost-effective, and non-destructive method. Thus, this study aimed to develop and apply the NIRS prediction models for quantifying amino acids in biofortified quality protein maize (QPM). Sixty-three (63) QPM maize genotypes were used as the calibration set, and another twenty (20) genotypes were used as the validation set. The microwave hydrolysis system coupled with post-column derivatization with 6-amino-quinoline-succinimidyl-carbamate as the derivatization reagent and the HPLC method were used to generate the reference data set used for the calibration development. The calibration models were developed for essential and non-essential amino acids using WINSI Foss software. Good coefficients of determination in calibration (R2cal) of 0.91, 0.93, 0.93, and 0.91 and low standard errors in calibrations (SEC) of 0.62, 0.71, 0.26, and 1.75 were obtained for glutamic acids, alanine, proline, and leucine, respectively, while aspartic acids, serine, glycine, arginine, tyrosine, valines, and phenylalanine had fairly good R2Cal values of 0.86, 0.71, 0.81, 0.78, 0.68, 0.79, and 0.75. In contrast, poor (R2cal) was obtained for histidine (0.07), cystine (0.09), methionine (0.09), lysine (0.20), threonine (0.51), and isoleucine (0.09), respectively. The models’ prediction performances (R2pred) and standard error of prediction (SEP) were reasonably good for certain amino acids such as aspartic acid (0.90), glycine (0.80), arginine (0.94), alanine (0.90), proline (0.80), tyrosine (0.83), valine (0.82), leucine (0.90), and phenylalanine (0.88) with SEP values of 0.24, 0.39,0.24, 0.93, 0.47,0.34, 0.78, 2.20, and 0.77, respectively. However, certain amino acids had their R2pred below 0.50, which could be improved to become useful for screening purposes for those amino acids. Further work is recommended by including a training set representing the sample population’s variance to improve the model’s performance.

Intelligent estimation of blood glucose level using wristband PPG signal and physiological parameters

The inconvenience and risk associated with the regular use of invasive blood glucose measurements has led to tremendous research in this area. This paper proposes the design of a non-invasive blood glucose estimation system using novel Mel frequency cepstral coefficients features of wristband photoplethysmogram signal and physiological parameters. A dataset from 217 participants of a hospital in Cuenca Ecuador is used to validate the proposed model. The support vector regression (SVR) and extreme gradient boost regression (XGBR) techniques are used to estimate blood glucose levels (BGL). The XGBR technique achieves the least value for the standard error of prediction (SEP), 9.78 mg/dL. Further, 5 features are selected from the feature set based on the feature importance in XGBR. The XGBR model with the reduced feature set results in further reduction of SEP value (5.53 mg/dL) with a correlation coefficient of 0.99. Standard Clarke error grid analysis and Bland-Altman analysis shows that the predicted glucose values are in the clinically acceptable region. The results of the proposed model demonstrate the potential of wearable BGL monitoring technology.

Longitudinal Study Comparing Orthogonal Signal Correction Algorithms Coupled with Partial Least-Squares for Quantitative Near-Infrared Spectroscopy

Several orthogonal signal correction (OSC) methods coupled with partial least-squares (PLS) were applied to near-infrared spectra in a longitudinal study for the prediction of glucose, urea, and triacetin in liquid samples. The OSC methods included Wold’s OSC, Fearn’s OSC, orthogonal projection to latent structures (OPLS), and direct orthogonal signal correction (DOSC). These preprocessing methods are designed to simplify the spectra and remove information that is orthogonal to the analyte. Samples consisted of various concentrations of glucose, urea, and triacetin in the range from 1 to 19 mM. Calibration and prediction data were collected over 1.7 years to test how well each of the OSC methods helped to maintain analyte prediction performance over time. When comparing the OSC methods, it was found that all of the OSC methods improved the predictions in most data sets throughout the study but that DOSC clearly outperformed the others. On the basis of 18 prediction sets collected across 617 days after the collection of the calibration data, the average percent improvement in the standard error of prediction (SEP) when applying DOSC was 4.6%, 83.1%, and 43.0%, respectively, for glucose, urea, and triacetin. These comparisons were made relative to the SEP values obtained with unprocessed spectra.

Near-Infrared Hyperspectral Imaging (NIR-HSI) for Nondestructive Prediction of Anthocyanins Content in Black Rice Seeds

Anthocyanins are an important micro-component that contributes to the quality factors and health benefits of black rice. Anthocyanins concentration and compositions differ among rice seeds depending on the varieties, growth conditions, and maturity level at harvesting. Chemical composition-based seeds inspection on a real-time, non-destructive, and accurate basis is essential to establish industries to optimize the cost and quality of the product. Therefore, this research aimed to evaluate the feasibility of near-infrared hyperspectral imaging (NIR-HSI) to predict the content of anthocyanins in black rice seeds, which will open up the possibility to develop a sorting machine based on rice micro-components. Images of thirty-two samples of black rice seeds, harvested in 2019 and 2020, were captured using the NIR-HSI system with a wavelength of 895–2504 nm. The spectral data extracted from the image were then synchronized with the rice anthocyanins reference value analyzed using high-performance liquid chromatography (HPLC). For comparison, the seed samples were ground into powder, which was also captured using the same NIR-HSI system to obtain the data and was then analyzed using the same method. The model performance of partial least square regression (PLSR) of the seed sample developed based on harvesting time, and mixed data revealed the model consistency with R2 over 0.85 for calibration datasets. The best prediction models for 2019, 2020, and mixed data were obtained by applying standard normal variate (SNV) pre-processing, indicated by the highest coefficient of determination (R2) of 0.85, 0.95, 0.90, and the lowest standard error of prediction (SEP) of 0.11, 0.17, and 0.16 mg/g, respectively. The obtained R2 and SEP values of the seed model were comparable to the result of powder of 0.92–0.95 and 0.09–0.15 mg/g, respectively. Additionally, the obtained beta coefficients from the developed model were used to generate seed chemical images for predicting anthocyanins in rice seed. The root mean square error (RMSE) value for seed prediction evaluation showed an acceptable result of 0.21 mg/g. This result exhibits the potential of NIR-HSI to be applied in a seed sorting machine based on the anthocyanins content.

Development of ternary models for prediction of biogas yield in a novel modular biodigester: a case of fuzzy Mamdani model (FMM), artificial neural network (ANN), and response surface methodology (RSM)

Fuzzy Mamdani Model (FMM), Artificial Neural Network (ANN), and Response Surface Methodology (RSM) models were adopted to model and optimise biogas production on assorted substrates of poultry wastes (PW) and cow dung in a modular biodigester system. The gas produced at retention time of 42 days was within the range of 50 to 52 ml/day. The temperature range was between the range of 20 and 40 °C, while the pH values were stable at optimal range of 8.0. The statistical analysis of the experimental result obtained from methane yield, using different computational technique, reveals that fuzzy Mamdani model reflect zero (0) values for average absolute deviation (AAD), mean average error (MAE), standard error of prediction (SEP), and root-mean-square error (RMSE) and average coefficient of determination (R2) of one (1.0). ANN result obtain showed AAD value of 0.00011991, MAE value of 0.000061538, SEP value of 0.0000044369, and RMSE value of 0.00022188 and R2 value of 1.0, and RSM result reveals absolute AAD of 0.0030, MAE of 0.0015, SEP of 0.00011092, RMSE of 0.0055, and R2 of 0.9998. It can be concluded from the result of the analysis that RSM gave a good prediction of the model while ANN and FMM models produced slightly better results. The result obtained from the FMM model seems marginally better and superior in terms of forecasting of biogas and methane production rate with satisfactory outcome.

Estimation of Chemical Composition of Pork Trimmings by Use of Density Measurement-Hydrostatic Method.

This study aims to determine the possibility of using density measurements by using the hydrostatic method for the estimation of the chemical composition of pork. The research material included 75 pork samples obtained during industrial butchering and cutting. The density measurements were performed using the hydrostatic method based on Archimedes’ principle. The meat samples were minced, and the content of the basic chemical components in them was determined. The usefulness of density measurement using the hydrostatic method in chemical composition estimation was determined by analyzing the correlation for the entire population, and after grouping the samples with a low (<15%), medium (15–25%), and high (>25%) fat content. High (in absolute value) coefficients of correlation between the meat density and the content of water (0.96), protein (0.94), and fat (−0.96) were found based on the results obtained. In order to achieve higher accuracy of the estimation, the applied regression equations should be adjusted to the presumed fat content in the meat. The standard error of prediction (SEP) values ranged from 0.67% to 2.82%, which indicates that the calculated estimation accuracy may be sufficient for proper planning of the production. Higher SEP values were found in fat content estimation and the lowest ones were found in protein content estimation.

Inline control of yoghurt fermentation process using a near infrared light backscatter sensor

In industrial yoghurt production, continuous process monitoring using pH meters is a cumbersome and non-appropriate technique. The objective of this study was to evaluate the suitability of an optical sensor as a new method for the inline control of yoghurt fermentation. Tests were conducted with three fat and three inulin concentrations and fermentation was monitored simultaneously using an inline light backscatter sensor, pH-meters and a rheometer. A mathematical model that correlates the near infrared light backscatter ratio with the pH, at real time during milk fermentation, was developed. The model was calibrated and successfully validated at the different experimental conditions, predicting the evolution of pH from 5.2 to 4.6 with determination coefficient values higher than 0.993 and standard error of prediction values between 0.02 and 0.11 pH units. The results demonstrated that this is a promising method for the inline control of pH in industrial yoghurt production.

Advanced process design and understanding of continuous twin-screw granulation via implementation of in-line process analytical technologies

Process analytical technologies (PAT) are identified as an essential element in the Quality by Design framework, providing the cornerstone to implement continuous pharmaceutical manufacturing. This study is concerned with employing three in-line PATs: Eyecon™ 3D imaging system, Near-infrared spectroscopy (NIRS) and Raman spectroscopy (RS), in a continuous twin-screw granulation process to enable real-time monitoring and prediction of critical quality attributes of granules. The Thermo Scientific™ Pharma 11 twin-screw granulator was used to manufacture granules from a low-dose formulation with caffeine anhydrous as the model drug. A 30-run full factorial design including three critical process parameters (liquid to solid ratio, barrel temperature and throughput) was conducted to evaluate the performance of each analytical tool. Eyecon™ successfully captured the granule size and shape variation from different experimental conditions and demonstrated sufficient sensitivity to the fluctuation of size parameter D10 in the presence of process perturbations. The partial least square regression (PLSR) models developed using NIRS showed small relative standard error of prediction values (less than 5%) for most granule physical properties. In contrast, the RS-based PLSR models revealed higher prediction errors towards granule drug concentration, potentially due to the inhomogeneous premixing of raw materials during calibration model development.

Near infrared spectroscopic analysis of total alkaloids as nicotine, total nitrogen and total ash in Cuban cigar tobacco

Three models using partial least squares regression of near infrared reflectance spectroscopy were developed for the first time to evaluate the concentration of total alkaloids as nicotine, total nitrogen and total ash in powdered samples of Cuban cigar tobacco. They had lower standard error of cross-validation (SECV) and standard error of prediction (SEP) than those using: (1) principal components regression; (2) multiple linear regression and (3) other partial least squares models previously published for this and other tobacco types. SECV values in mass fraction units were: 0.0010, 0.0011 and 0.0048, respectively. The coefficients of determination (R2) in calibration were: 0.99, 0.94 and 0.96, respectively. Standard error of prediction values were: 0.0011, 0.0012 and 0.0049, respectively. The r2 obtained when comparing the predicted values using the validation set with the reference data was nearly the same as in calibration. The equation for total alkaloids as nicotine was good for quantification, and for process control, development and applied research according to the quotients range/standard error of prediction and standard deviation/standard error of prediction, respectively. Equations for the other analytes were as good as the one for alkaloids, but suitable only for screening purposes. Joint F-tests on the regression coefficients from ordinary least squares and iterative reweighted least squares fittings among predicted and reference data indicated slopes and intercepts enclosing 1 and 0, respectively. The relative systematic errors between the near infrared spectroscopy predicted values and the reference methods are close to the same figures reported previously for traditional analytical procedures under intermediate precision and reproducibility conditions; hence, the models are suitable for this industrial context.

Non-destructive assessment of the oxidative stability of intact macadamia nuts during the drying process by near-infrared spectroscopy

We have developed a rapid non-destructive method to assess the oxidative stability of intact macadamia nuts using near-infrared spectroscopy (NIRS). Intact macadamia nuts of the cultivars HAES 344 ‘Kau’, HAES 660 ‘Keaau’, IAC 4–12 B, and IAC Campinas B were harvested and immediately oven-dried for 4 days at 30 °C, 2 days at 40 °C, and 1 day at 60 °C to achieve 1.5% kernel moisture content. At each drying step nuts were withdrawn and their moisture content, peroxide value (PV), and acidity index (AI) determined. The best partial least square model for PV prediction was obtained using the Savitzky-Golay (SG) second derivative resulting in a standard error of prediction (SEP) of 0.55 meq·kg−1 and a coefficient of determination (R2C) of 0.57. The best AI prediction-model result was obtained using the SG second derivative (SEP = 0.14%, R2C = 0.29). Based on the maximum quality limits of 3 meq·kg−1 for PV and 0.5% for AI, the SEP values represented 18% and 28%, respectively. Therefore, the prediction method can be considered useful since the errors are lower than the quality limits. Thus, NIRS can be used to assess the oxidative stability of intact macadamia kernels.

At‐line monitoring of industrial frying processes using ATR‐FTIR‐PLS method

AbstractThis study investigated the application of infrared spectroscopy with multivariate calibration methods for at‐line monitoring of the degradation of soybean oil in industrial frying processes by determining when the acidity index and total polar materials (TPM). The infrared spectra (650–3,200 cm−1) were acquired using the attenuated total reflection accessory (ATR‐FTIR), with a resolution of 4 cm−1, and 16 scans. Partial least‐squares regression (PLS) models were evaluated for individual and simultaneous determination, and results were compared with reference methods. The individual calibration model showed standard error of prediction values of 0.09% (w/w) and 1.6% (w/w) for the acidity index and TPM, respectively. The simultaneous determination of the acidity index and TPM showed SEP values of 0.17% (w/w) and 1.6% (w/w), respectively. The results demonstrate that infrared spectroscopy combined with multivariate calibration techniques can be servant used in routine soybean oil quality control in industrial frying processes.Practical applicationsFoods prepared by the frying process are strongly influenced by the oil used in the process, since the product tends to absorb part of the oil during its cooking. This oil when used for a long period and exposed to high temperatures, end up suffering oxidation reactions, and can greatly reduce your shelf life. In this way the industries that use these processes for the manufacture of their products, apply a series of chemical analyzes in order to validate the quality of the oil used, these analyzes use a high amount of solvents and time. As an alternative to the use of these traditional analyzes, methods have been developed through the use of infrared spectroscopy combined with multivariate analysis tools, for the creation of partial least squares calibration models for the prediction of quality parameters of oils used in frying.

Determination of nutritional parameters of yoghurts by FT Raman spectroscopy

FT-Raman quantitative analysis of nutritional parameters of yoghurts was performed with the help of partial least squares models. The relative standard errors of prediction for fat, lactose and protein determination in the quantified commercial samples equalled to 3.9, 3.2 and 3.6%, respectively. Models based on attenuated total reflectance spectra of the liquid yoghurt samples and of dried yoghurt films collected with the single reflection diamond accessory showed relative standard errors of prediction values of 1.6−5.0% and 2.7−5.2%, respectively, for the analysed components. Despite a relatively low signal-to-noise ratio in the obtained spectra, Raman spectroscopy, combined with chemometrics, constitutes a fast and powerful tool for macronutrients quantification in yoghurts. Errors received for attenuated total reflectance method were found to be relatively higher than those for Raman spectroscopy due to inhomogeneity of the analysed samples.

A method for the inline measurement of milk gel firmness using an optical sensor

At present, selection of cutting time during cheesemaking is made based on subjective methods, which has effects on product homogeneity and has prevented complete automation of cheesemaking. In this work, a new method for inline monitoring of curd firmness is presented. The method consisted of developing a model that correlates the backscatter ratio of near infrared light during milk coagulation with the rheological storage modulus. The model was developed through a factorial design with 2 factors: protein concentration (3.4 and 5.1%) and coagulation temperature (30 and 40°C). Each treatment was replicated 3 times; the model was calibrated with the first replicate and validated using the remaining 2 replicates. The coagulation process was simultaneously monitored using an optical sensor and small-amplitude oscillatory rheology. The model was calibrated and successfully validated at the different protein concentrations and coagulation temperatures studied, predicting the evolution of storage modulus during milk coagulation with coefficient of determination values >0.998 and standard error of prediction values <3.4 Pa. The results demonstrated that the proposed method allows inline monitoring of curd firming in cheesemaking and cutting the curd at a proper firmness to each type of cheese.

Quality Analysis of Stored Bell Peppers Using Near-Infrared Hyperspectral Imaging

Abstract. The objective of this study was to predict the moisture content (MC), soluble solids content (SSC), and titratable acidity (TA) content in bell peppers during storage (18°C, 85% relative humidity) over 12 days, based on near-infrared hyperspectral imaging (NIR-HSI) in the 1000-1500 nm wavelength range. The mean spectra of 148 mature bell peppers were extracted from the hyperspectral images, and multivariate calibration models were built using partial least squares (PLS) regression with different preprocessing spectra techniques. The most effective wavelengths were selected using the variable importance in projection (VIP) technique, which selected optimal variables for the target quality parameters of bell peppers from a full set of variables. Subsequently the selected variables were used to develop a PLS-VIP model for simplifying the prediction model. The MC, SSC, and TA content in bell peppers during storage changed from 90.7% to 93.0%, from 6.1%Brix to 7.3%Brix, and from 0.222% to 0.334%, respectively. The PLS regression model with MC, SSC, and TA content resulted in coefficients of determination (R2pred) of 0.83, 0.85, and 0.7, with standard errors of prediction (SEP) of 0.08%, 0.075%Brix, and 0.013%, respectively, using SNV preprocessed spectra for MC and TA content and Savitzky-Golay (S-G) second-order derivatives preprocessed spectra for SSC of bell peppers. By contrast, the prediction results yielded R2pred of 0.69, 0.75, and 0.68, respectively, with SEP values of 0.103%, 0.107%Brix, and 0.011% when the PLS-VIP model was employed. The PLS-VIP model simplified the calibration model by selecting the most important variables in terms of their responsiveness to bell pepper quality properties. The results revealed that HSI coupled with multivariate analysis can be used successfully to predict the MC, SSC, and TA content in bell peppers. Keywords: Fruit quality, Hyperspectral imagery, Image analysis, Spectral analysis, Stored bell pepper.

Determination of fat, protein, moisture, and salt content of Cheddar cheese using mid-infrared transmittance spectroscopy

The objective of our work was to develop and evaluate the performance of a rapid method for measuring fat, protein, moisture, and salt content of Cheddar cheese using a combination mid-infrared (MIR) transmittance analysis and an in-line conductivity sensor in an MIR milk analyzer. Cheddar cheese was blended with a dissolving solution containing pentasodium triphosphate and disodium metasilicate to achieve a uniform, particle-free dispersion of cheese, which had a fat and protein content similar to milk and could be analyzed using a MIR transmittance milk analyzer. Annatto-colored Cheddar cheese samples (34) from one cheese factory were analyzed using reference chemistry methods for fat (Mojonnier ether extraction), crude protein (Kjeldahl), moisture (oven-drying total solids), and salt (Volhard silver nitrate titration). The same 34 cheese samples were also dissolved using the cheese dissolver solution, and then run through the MIR and used for calibration. The reference testing for fat and crude protein was done on the cheese after dispersion in the dissolver solution. Validation was done using a total of 36 annatto-colored Cheddar cheese samples from 4 cheese factories. The 36 validation cheese samples were also analyzed using near-infrared spectroscopy for fat, moisture, and the coulometric method for salt in each factory where they were produced. The validation cheeses were also tested using the same chemical reference methods that were used for analysis of the calibration samples. Standard error of prediction (SEP) values for moisture and fat on the near-infrared spectroscopy were 0.30 and 0.45, respectively, whereas the MIR produced SEP values of 0.28 and 0.23 for moisture (mean 36.82%) and fat (mean 34.0%), respectively. The MIR also out-performed the coulometric method for salt determination with SEP values of 0.036 and 0.139 at a mean level of salt of 1.8%, respectively. The MIR had an SEP value of 0.19 for estimation at a mean level of 24.0% crude protein, which suggests that MIR could be an easy and effective way for cheese producers to measure protein to determine protein recovery in cheese making.

Quantification of active ingredients in pharmaceutical suspensions by FT Raman spectroscopy

A method for quantitative determination of nifuroxazide, acetaminophen, ibuprofen and naproxen in commercial oral suspensions containing active compounds in the 1.7–4.0% (w/w) range is described. Partial least squares regression and counter-propagation artificial neural network models were built and validated. The lowest relative standard error of prediction values was found for the strongest scattering nifuroxazide suspensions. Commercial suspensions were quantified with recovery in the 99.3–101.2% and the 98.2–100.2% range, applying partial least squares and artificial neural network modeling techniques, respectively.

Evaluation of a newly developed mid-infrared sensor for real-time monitoring of yeast fermentations

A mid-infrared (MIR) sensor using the attenuated total reflection (ATR) technique has been developed for real-time monitoring in biotechnology. The MIR-ATR sensor consists of an IR emitter as light source, a zinc selenide ATR prism as boundary to the process, and four thermopile detectors, each equipped with an optical bandpass filter. The suitability of the sensor for practical application was tested during aerobic batch-fermentations of Saccharomyces cerevisiae by simultaneous monitoring of glucose and ethanol. The performance of the sensor was compared to a commercial Fourier transform mid-infrared (FT-MIR) spectrometer by on-line measurements in a bypass loop. Sensor and spectrometer were calibrated by multiple linear regression (MLR) in order to link the measured absorbance in the transmission ranges of the four optical sensor channels to the analyte concentrations. For reference analysis, high-performance liquid chromatography (HPLC) was applied. Process monitoring using the sensor yielded in standard errors of prediction (SEP) of 6.15g/L and 1.36g/L for glucose and ethanol. In the case of the FT-MIR spectrometer the corresponding SEP values were 4.34g/L and 0.61g/L, respectively. The advantages of optical multi-channel mid-infrared sensors in comparison to FT-MIR spectrometer setups are the compactness, easy process implementation and lower price.

Optimisation of the Quantification of Total Soluble Solids, pH and Titratable Acidity in South African Grape Must using Fourier Transform Mid-infrared Spectroscopy

Calibration models for Fourier transform mid-infrared (FT-MIR) spectroscopy were developed for the simultaneous quantification of total soluble solids (TSS, measured as °Brix), pH and titratable acidity (TA, expressed as g/L tartaric acid) in South African (SA) grape must. An exploratory data analysis of the FT-MIR spectra of 1170 grape must samples (647 for °Brix, 252 for pH and 271 for TA) was done by principal component analysis, and partial least squares regression was used for the computation of the regression models. The prediction errors for TSS (0.34 °Brix), pH (0.04 units) and TA (0.51 g/L) provided analytical data of satisfactory accuracy. The evaluation of readyto- use global calibrations to quantify these three parameters in SA samples presented standard error of prediction (SEP) values of 0.46°Brix, 0.10 pH units and 3.13 g/L for TA. After slope and intercept adjustments of the original global calibration algorithms, the SEP values were reduced to 0.38 °Brix, 0.05 pH units and 0.49 g/L for TA. These results show the necessity for optimisation of the global FT-MIR WineScan calibrations to provide a better fit to samples of South African origin. The results demonstrate that FT-MIR spectroscopy is a useful technique for the rapid quantification of major grape must parameters and for quality control purposes in an industrial cellar.

Validation and Comparison of Formol and FT-IR Methods for Assimilable Nitrogen in Vine Grapes

In this study, two methods used to evaluate assimilable nitrogen in grape juice were compared, namely: the formol method and the Fοurier-transform mid-infrared (FT-MIR) spectroscopy. At the outset the formol method was validated in the laboratory by determining parameters such as precision, accuracy and trueness. Parallel recovery studies were conducted with a model solution and natural grape juice for ammonium and arginine. Average recoveries of ammonium for the formol titration ranged from 77% to 88% and for arginine from 90% to 97%. The repeatability and the intra laboratory reproducibility were ±6.4 and ±31.7 (mg/L of nitrogen) respectively. Formol titration was used as the reference method for the calibration of FT-MIR spectroscopy to determine the assimilable nitrogen in grape juice. Grape juice samples (n=71) for 7 red and 7 white varieties from the Lisbon region in Portugal were analyzed simultaneously by using the formol method and Fοurier-transform mid-infrared (FT-MIR) spectroscopy. The results of this study demonstrated that it is possible to explore the applicability of FT-ΜIR technology to detect the assimilable nitrogen in grape juice for quantification purposes. The correlation coefficient (R2=0. 993), standard prediction error (SEP; 6.4 mg/L) and the RPD or standard deviation of the data/ standard error of prediction value (Sd/Se; 7.8) proved satisfactory for the calibration of the instrument. In addition, the results obtained by FT-MIR spectroscopy were comparable to those obtained when using the reference method.