Search

Long-term studies have shown a bias drift over time in the prediction performance of near-infrared spectroscopy measurement systems. This bias drift generally requires extra laboratory reference measurements to detect and correct for this bias. Since these reference measurements are expensive and time consuming, there is a need for advanced methodologies for bias drift monitoring and correction without the need for taking extra samples. In this study, we propose and validate a method to monitor the bias drift and two methods to tackle it. The first method requires no extra measurements and uses a modified version of Partial Least Squares Regression to estimate and correct the bias. This method is based on the assumption that the mean concentration of the predicted component remains constant over time. The second method uses regular bulk milk measurements as a reference for bias correction. This method compares the measured concentrations of the bulk milk to the volume-weighted average concentrations of individual milk samples predicted by the sensor. Any difference between the actual and calculated bulk milk composition is then used to perform a bias correction on the predictions by the sensor system. The effectiveness of these methods to improve the component prediction was evaluated on data originating from a custom-built sensor that automatically measures the NIR reflectance and transmittance spectra of raw milk on the farm. We evaluate the practical use case where models for predicting the milk composition are trained upon installation of the sensor at the farm, and later used to predict the composition of subsequent samples over a period of more than 6 months. The effectiveness of the fully unsupervised method was confirmed when the mean concentration of the milk samples remained constant, while the effectiveness reduced when this was not the case. The bulk milk correction method was effective when all relevant samples for the component were measured by the sensor and included in the analyzed bulk milk, but is less effective when samples included in the bulk which are not measured by the sensor system. When the necessary conditions are met, these methods can be used to extend the lifetime of deployed prediction models by significantly reducing the bias on the predicted values., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

In high-yielding dairy cattle, severe postpartum negative energy balance is often associated with metabolic and infectious disorders that negatively affect production, fertility, and welfare. Mobilization of adipose tissue associated with negative energy balance is reflected through an increased level of nonesterified fatty acids (NEFA) in the blood plasma. Earlier, identification of negative energy balance through detection of increased blood plasma NEFA concentration required laborious and stressful blood sampling. More recently, attempts have been made to predict blood NEFA concentration from milk samples. In this study, we aimed to develop and validate a model to predict blood plasma NEFA concentration using the milk mid-infrared (MIR) spectra that are routinely measured in the context of milk recording. To this end, blood plasma and milk samples were collected in wk 2, 3, and 20 postpartum for 192 lactations in 3 herds. The blood plasma samples were taken in the morning, and representative milk samples were collected during the morning and evening milk sessions on the same day. To predict plasma NEFA concentration from the milk MIR spectra, partial least squares regression models were trained on part of the observations from the first herd. The models were then thoroughly validated on all other observations of the first herd and on the observations of the 2 independent herds to explore their robustness and wide applicability. The final model could accurately predict blood plasma NEFA concentrations <0.6 mmol/L with a root mean square error of prediction of <0.143 mmol/L. However, for blood plasma with >1.2 mmol/L NEFA, the model clearly underestimated the true level. Additionally, we found that morning blood plasma NEFA levels were predicted with significantly higher accuracy using MIR spectra of evening milk samples compared with MIR spectra of morning samples, with root mean square error of prediction values of, respectively, 0.182 and 0.197 mmol/L, and R 2 values of 0.613 and 0.502. These results suggest a time delay between variations in blood plasma NEFA and related milk biomarkers. Based on the MIR spectra of evening milk samples, cows at risk for negative energy status, indicated by detrimental morning blood plasma NEFA levels (>0.6 mmol/L), could be identified with a sensitivity and specificity of, respectively, 0.831 and 0.800. As this model can be applied to millions of historical and future milk MIR spectra, it opens an opportunity for regular metabolic screening and improved resilience phenotyping., (Copyright © 2020 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.)

Automated monitoring of fertility in dairy cows using milk progesterone is based on the accurate and timely identification of luteolysis. In this way, well-adapted insemination advice can be provided to the farmer to further optimize fertility management. To properly evaluate and compare the performance of new and existing data-processing algorithms, a test data set of progesterone time-series that fully covers the desired variability in progesterone profiles is needed. Further, the data should be measured with a high frequency to allow rapid onset events, such as luteolysis, to be precisely determined. Collecting this type of data would require a lot of time, effort, and budget. In the absence of such data, an alternative was developed using simulated progesterone profiles for multiple cows and lactations, in which the different fertility statuses were represented. To these, relevant variability in terms of cycle characteristics and measurement error was added, resulting in a large cost-efficient data set of well-controlled but highly variable and farm-representative profiles. Besides the progesterone profiles, information on (the timing of) luteolysis was extracted from the modeling approach and used as a reference for the evaluation and comparison of the algorithms. In this study, 2 progesterone monitoring tools were compared: a multiprocess Kalman filter combined with a fixed threshold on the smoothed progesterone values to detect luteolysis, and a progesterone monitoring algorithm using synergistic control, PMASC, which uses a mathematical model based on the luteal dynamics and a statistical control chart to detect luteolysis. The timing of the alerts and the robustness against missing values of both algorithms were investigated using 2 different sampling schemes: one sample per cow every 8 h versus 1 sample per day. The alerts for luteolysis of the PMASC algorithm were on average 20 h earlier compared with the ones of the multiprocess Kalman filter, and their timing was less sensitive to missing values. This was shown by the fact that, when 1 sample per day was used, the Kalman filter gave its alerts on average 24 h later, and the variability in timing of the alerts compared with simulated luteolysis increased with 22%. Accordingly, we postulate that implementation of the PMASC system could improve the consistency of luteolysis detection on farm and lower the analysis costs compared with the current state of the art., (Copyright © 2019 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.)

The progesterone (P4) monitoring algorithm using synergistic control (PMASC) uses luteal dynamics to identify fertility events in dairy cows. This algorithm employs a combination of mathematical functions describing the increasing and decreasing P4 concentrations during the development and regression of the corpus luteum and a statistical control chart that allows identification of luteolysis. The mathematical model combines sigmoidal functions from which the cycle characteristics can be calculated. Both the moment at which luteolysis is detected and confirmed by PMASC, as well as the model features themselves, can be used to inform the farmer on the fertility status of the cows., (Copyright © 2019 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.)

Udder health problems are often associated with milk losses. These losses are different between quarters, as infected quarters are affected both by systemic and pathogen-specific local effects, whereas noninfected quarters are only subject to systemic effects. To gain insight in these losses and the milk yield dynamics during disease, it is essential to have a reliable reference for quarter-level milk yield in an unperturbed state, mimicking its potential yield. We developed a novel methodology to predict this quarter milk yield per milking session, using an historical data set of 504 lactations collected on a test farm by an automated milking system from DeLaval (Tumba, Sweden). Using a linear mixed model framework in which covariates associated with the linearized Wood model and the milking interval are included, we were able to describe quarter-level yield per milking session with a proportional error below 10%. Applying this model enables us to predict the milk yield of individual quarters 1 to 50 d ahead with a mean prediction error ranging between 8 and 20%, depending on the amount of historical data available to estimate the random effect covariates for the predicted lactation. The developed methodology was illustrated using 2 examples for which quarter-level milk losses are calculated during clinical mastitis. These showed that the quarter-level mixed model allows us to gain insight in quarter lactation dynamics and enables to calculate milk losses in different situations., (Copyright © 2018 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.)

Timely identification of a cow's reproduction status is essential to minimize fertility-related losses on dairy farms. This includes optimal estrus detection, pregnancy diagnosis, and the timely recognition of early embryonic death and ovarian problems. On-farm milk progesterone (P4) analysis can indicate all of these fertility events simultaneously. However, milk P4 measurements are subject to a large variability both in terms of measurement errors and absolute values between cycles. The objective of this paper is to present a newly developed methodology for detecting luteolysis preceding estrus and give an indication of its on-farm use. The innovative monitoring system presented is based on milk P4 using the principles of synergistic control. Instead of using filtering techniques and fixed thresholds, the present system employs an individually on-line updated model to describe the P4 profile, combined with a statistical process control chart to identify the cow's fertility status. The inputs for the latter are the residuals of the on-line updated model, corrected for the concentration-dependent variability that is typical for milk P4 measurements. To show its possible use, the system was validated on the P4 profiles of 38 dairy cows. The positive predictive value for luteolysis followed by estrus was 100%, meaning that the monitoring system picked up all estrous periods identified by the experts. Pregnancy or embryonic mortality was characterized by the absence or detection of luteolysis following an insemination, respectively. For 13 cows, no luteolysis was detected by the system within the 25 to 32 d after insemination, indicating pregnancy, which was confirmed later by rectal palpation. It was also shown that the system is able to cope with deviating P4 profiles having prolonged follicular or luteal phases, which may suggest the occurrence of cysts. Future research is recommended for optimizing sampling frequency, predicting the optimal insemination window, and establishing rules to detect problems based on deviating P4 patterns., (Copyright © 2018 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.)

Reproductive performance is an important factor affecting the profitability of dairy farms. Optimal fertility results are often confined by the time-consuming nature of classical heat detection, the fact that high-producing dairy cows show estrous symptoms shorter and less clearly, and the occurrence of ovarian problems. Today's commercially available solutions for automatic estrus detection include monitoring of activity, temperature and progesterone. The latter has the advantage that, besides estrus, it also allows to detect pregnancy and ovarian problems. Due to the large variation in progesterone profiles, even between cycles within the same cow, the use of general thresholds is suboptimal. To this end, an intelligent and individual interpretation of the progesterone measurements is required. Therefore, an alternative solution is proposed, which takes individual and complete cycle progesterone profiles into account for reproduction monitoring. In this way, profile characteristics can be translated into specific attentions for the farmers, based on individual rather than general guidelines. To enable the use of the profile and cycle characteristics, an appropriate model to describe the milk progesterone profile was developed. The proposed model describes the basal adrenal progesterone production and the growing and regressing cyclic corpus luteum. To identify the most appropriate way to describe the increasing and decreasing part of each cycle, three mathematical candidate functions were evaluated on the increasing and decreasing parts of the progesterone cycle separately: the Hill function, the logistic growth curve and the Gompertz growth curve. These functions differ in the way they describe the sigmoidal shape of each profile. The increasing and decreasing parts of the P4 cycles were described best by the model based on respectively the Hill and Gompertz function. Combining these two functions, a full mathematical model to characterize the progesterone cycle was obtained. It was shown that this approach retains the flexibility to deal with both varying baseline and luteal progesterone values, as well as prolonged or delayed cycles., (Copyright © 2017 Elsevier Inc. All rights reserved.)

A global optimizer has been developed, capable of computing the optimal configuration in a probe for spatially resolved reflectance spectroscopy (SRS). The main objective is to minimize the number of detection fibers, while maintaining an accurate estimation of both absorption and scattering profiles. Multiple fibers are necessary to robustify the estimation of optical properties against noise, which is typically present in the measured signals and influences the accuracy of the inverse estimation. The optimizer is based on a robust metamodel-based inverse estimation of the absorption coefficient and a reduced scattering coefficient from the acquired SRS signals. A genetic algorithm is used to evaluate the effect of the fiber placement on the performance of the inverse estimator to find the bulk optical properties of raw milk. The algorithm to find the optimal fiber placement was repeatedly executed for cases with a different number of detection fibers, ranging from 3 to 30. Afterwards, the optimal designs for each considered number of fibers were compared based on their performance in separating the absorption and scattering properties, and the significance of the differences was tested. A sensor configuration with 13 detection fibers was found to be the combination with the lowest number of fibers which provided an estimation performance which was not significantly worse than the one obtained with the best design (30 detection fibers). This design resulted in the root mean squared error of prediction (RMSEP) of 1.411 cm(-1) (R(2) = 0.965) for the estimation of the bulk absorption coefficient values, and 0.382 cm(-1) (R(2) = 0.996) for the reduced scattering coefficient values., (Copyright © 2016 Elsevier B.V. All rights reserved.)

The size of colloidal particles in food products has a considerable impact on the product's physicochemical, functional and sensory characteristics. Measurement techniques to monitor the size of suspended particles could, therefore, help to further reduce the variability in production processes and promote the development of new food products with improved properties. Visible and near-infrared (Vis/NIR) spectroscopy is already widely used to measure the composition of agricultural and food products. However, this technology can also be consulted to acquire microstructure-related scattering properties of food products. In this study, the effect of the fat globule size on the Vis/NIR bulk scattering properties of milk was investigated. Variability in fat globule size distribution was created using ultrasonic homogenization of raw milk. Reduction of the fat globule size resulted in a higher wavelength-dependency of both the Vis/NIR bulk scattering coefficient and the scattering anisotropy factor. Moreover, the anisotropy factor and the bulk scattering coefficients for wavelengths above 600 nm were reduced and were dominated by Rayleigh scattering. Additionally, the bulk scattering properties could be well (R(2) ≥ 0.990) estimated from measured particle size distributions by consulting an algorithm based on the Mie solution. Future research could aim at the inversion of this model to estimate the particle size distributions from Vis/NIR spectroscopic measurements., (Copyright © 2015 Elsevier B.V. All rights reserved.)

Milk production is a dominant factor in the metabolism of dairy cows involving a very intensive interaction with the blood circulation. As a result, the extracted milk contains valuable information on the metabolic status of the cow. On-line measurement of milk components during milking two or more times a day would promote early detection of systemic and local alterations, thus providing a great input for strategic and management decisions. The objective of this study was to investigate the potential of mid-infrared (mid-IR) spectroscopy to measure the milk composition using two different measurement modes: micro attenuated total reflection (μATR) and high throughput transmission (HTT). Partial least squares (PLS) regression was used for prediction of fat, crude protein, lactose and urea after preprocessing IR data and selecting the most informative wavenumber variables. The prediction accuracies were determined separately for raw and homogenized copies of a wide range of milk samples in order to estimate the possibility for on-line analysis of the milk. In case of fat content both measurement modes resulted in an excellent prediction for homogenized samples (R(2)>0.92) but in poor results for raw samples (R(2)<0.70). Homogenization was however not mandatory to achieve good predictions for crude protein and lactose with both μATR and HTT, and urea with μATR spectroscopy. Excellent results were obtained for prediction of crude protein, lactose and urea content (R(2)>0.99, 0.98 and 0.86 respectively) in raw and homogenized milk using μATR IR spectroscopy. These results were significantly better than those obtained by HTT IR spectroscopy. However, the prediction performance of HTT was still good for crude protein and lactose content (R(2)>0.86 and 0.78 respectively) in raw and homogenized samples. However, the detection of urea in milk with HTT spectroscopy was significantly better (R(2)=0.69 versus 0.16) after homogenization of the milk samples. Based on these observations it can be concluded that μATR approach is most suitable for rapid at line or even on-line milk composition measurement, although homogenization is crucial to achieve good prediction of the fat content., (Copyright © 2011 Elsevier B.V. All rights reserved.)

Today, measurement of raw milk quality and composition relies on Fourier transform infrared spectroscopy to monitor and improve dairy production and cow health. However, these laboratory analyzers are bulky, expensive and can only be used by experts. Moreover, the sample logistics and data transfer delay the information on product quality, and the measures taken to optimize the care and feeding of the cattle render them less suitable for real-time monitoring. An on-farm spectrometer with compact size and affordable cost could bring a solution for this discrepancy. This paper evaluates the performance of microelectromechanical system (MEMS)-based near-infrared (NIR) spectrometers as on-farm milk analyzers. These spectrometers use Fabry–Pérot interferometers for wavelength tuning, giving them the advantage of very compact size and affordable price. This study discusses the ability of MEMS spectrometers to reach the accuracy limits set by the International Committee for Animal Recording (ICAR) for at-line analyzers of the milk content regarding fat, protein and lactose. According to the achieved results, the transmission measurements with the NIRONE 2.5 spectrometer perform best, with an acceptable root mean squared error of prediction (RMSEP = 0.21% w/w) for the measurement of milk fat and excellent performance (RMSEP ≤ 0.11% w/w) for protein and lactose. In addition, the transmission measurements using the NIRONE 2.0 module give similar results for fat and lactose (RMSEP of 0.21 and 0.10% w/w respectively), while the prediction of protein is slightly deteriorated (RMSEP = 0.15% w/w). These results show that the MEMS spectrometers can reach sufficient prediction accuracy compared to ICAR standard values for at-line and in-line fat, protein and lactose prediction. [ABSTRACT FROM AUTHOR]

• A fiber-optic spatially-resolved spectroscopy (SRS) setup measured milk NIR spectra. • The setup evaluated 30 different illumination-to-detection distances (1.1–4 mm) • Shorter illumination-to-detection distances were optimal for lactose prediction. • Larger illumination-to-detection distances were optimal for fat and protein. • The accuracy of the composition predictions are well within the ICAR requirements. To support in-line quality control of raw milk and ensure the close monitoring of the metabolic and udder health of dairy cows, we propose a fiber-optic spatially-resolved spectroscopy (SRS) setup. This setup allows to vary the interaction of long-wave near-infrared (LW-NIR) light with the milk by submerging a separate optical illumination and detection fiber into the sample and altering their relative distance to optimize measurements for specific milk component(s). We evaluated this approach for predicting milk fat, protein, and lactose content and determined the optimal illumination-to-detection distances for each milk component. The region between 1.1 and 1.8 mm was optimal for lactose, and between 2.2 and 3.8 mm for fat and protein. These distance ranges resulted in a root-mean-square error of prediction (RMSEP) of less than 0.10 % weight/weight (wt/wt) for milk fat, and lower than 0.13 % (wt/wt) for protein and lactose. Integration of these distances into a fiber-optic SRS reflectance probe would allow to simultaneously determine the fat, protein, and lactose content of raw milk in-line with high accuracy. [ABSTRACT FROM AUTHOR]

• An online milk composition sensor is implemented on a farm and tested for 8 weeks. • Milk components are predicted more accurately when models are built post hoc. • Real-time prediction better reflects on-farm challenges for a milk analyzer. • The accuracy of the real-time prediction is well within the ICAR requirements. • A cow-specific bias correction further improved the milk lactose prediction. On-farm monitoring of milk composition can support close control of the udder and metabolic health of individual dairy cows. In previous studies, near-infrared (NIR) spectroscopy applied to milk analysis has proven useful for predicting the main components of raw milk (fat, protein, and lactose). In this contribution, we present and evaluate a precise tool for online milk composition analysis on the farm. For each milking, the online analyzer automatically collects and analyses a representative milk sample. The system acquires the NIR transmission spectra of the milk samples in the wavelength range from 960 to 1690 nm and performs a milk composition prediction afterward. Over a testing period of 8 weeks, the sensor collected 1165 NIR transmittance spectra of raw milk samples originating from 36 cows for which reference values were obtained for fat, protein, and lactose. For the same online sensor system, two calibration scenarios were evaluated: training post-hoc prediction models based on a representative set of calibration samples (n = 319) acquired over the entire testing period, with different cows in the calibration and test set, and training real-time prediction models exclusively on the samples acquired in the first week of the testing period (n = 308). The obtained prediction models were thoroughly tested on all the remaining samples not included in the calibration sets (n respectively 846 and 857). For the post-hoc prediction models, this resulted in an overall prediction error (root-mean-squared error of prediction, RMSEP) smaller than 0.080% (all % are in wt / wt) for milk fat (range 1.5–6.3%), protein (2.6–4.3%) and lactose (4–5.1%), with a coefficient of determination R2 of 0.989, 0.947 and 0.689 for fat, protein, and lactose respectively. For the real-time prediction models, the RMSEP was smaller than 0.092% for milk fat and lactose, and 0.110% for protein, with an R2 of 0.989 (fat), 0.894 (protein) and 0.644 (lactose). The milk lactose predictions could be further improved (RMSEP = 0.088%, R2 = 0.675) by taking into account a cow-specific bias. The presented online sensor system using the real-time prediction approach can thus be used for detailed and autonomous on-farm monitoring of milk composition after each individual milking, as its accuracy is well within the requirements by the International Committee for Animal Recording (ICAR) for on-farm milk analyzers and even meet the standards for laboratory analysis systems for fat and lactose. For this real-time prediction approach, a drift was observed in the predictions, especially for protein. Therefore, further research on the development of online calibration maintenance techniques is required to correct for this model drift and further improve the performance of this sensor system. [ABSTRACT FROM AUTHOR]