Novel prediction models for hyperketonemia using bovine milk Fourier-transform infrared spectroscopy

Metabolic diseases driven by negative energy balance in dairy cattle contribute to reduced milk production, increased disease incidence, culling, and death. Cow side tests for negative energy balance markers are available but are labor-intensive. Milk sample analysis using Fourier transform infrared spectroscopy (FTIR) allows for sampling numerous cows simultaneously. FTIR prediction models have moderate accuracy for hyperketonemia diagnosis (beta-hydroxybutyrate (BHB) ≥ 1.2 mmol/L). Most research using FTIR has focused on homogenous datasets and conventional prediction models, including partial least squares, linear discriminant analysis, and ElasticNet. Our objective was to evaluate more diverse modeling options, such as deep learning, gradient boosting machine models, and model ensembles for hyperketonemia classification. We compiled a sizable, heterogeneous dataset including milk FTIR and concurrent blood samples. Blood samples were tested for blood BHB, and wavenumber data was obtained from milk FTIR analysis. Using this dataset, we trained conventional prediction models and other options listed above. We demonstrate prediction model performance is similar for convolutional neural networks and ensemble models to simpler algorithm options. Results obtained from this study indicate that deep learning and model ensembles are potential algorithm options for predicting hyperketonemia in dairy cattle. Additionally, our results indicate hyperketonemia prediction models can be developed using heterogeneous datasets.

Dynamics of metabolic characteristics in dairy cows and their impact on disease-free survival time

Dairy cows are at a greater risk of disease due to increased energy demand during the transition period. Blood biomarkers including beta-hydroxybutyrate (BHBA)¹ and non-esterified fatty acids (NEFA)² are routinely used to identify animals in a state of negative energy balance (NEB)³. Recent research demonstrates cattle have varied response to NEB, that requires multiple blood biomarkers to characterize. This research identified five subcategories (cowtypes) of metabolic responses in transition dairy cows: Healthy, Athlete, Clever, Hyperketonemia, and Poor Metabolic Adaptation Syndrome (PMAS)⁴. The data set used in this study was collected in Germany by VIT - Vereinigte Informationssysteme Tierhaltung w.V. from 2016 to 2020. Health issues with time of diagnostic were included in the dataset. Using previously reported prediction models for blood BHB and blood NEFA and milk Fourier-transform infrared spectroscopy (FTIR)⁵ data, the cowtypes in our dataset were predicted. The objective of this study is to evaluate the association of the cowtypes with the disease-free survival time in dairy cows during early post calving using an accelerated failure time regression model. Additionally, transition probabilities of the population dynamics between cowtypes are studied by means of a Markov chain model. Using Healthy cowtype as reference level, Athlete, Clever, and PMAS cowtypes were found to be significant for the disease-free survival probability (P < 0.01). Conversely, Hyperketonemia cowtype was not significant (P = 0.182). Compared to the Healthy cowtype, all other cowtypes had a negative effect on the survival probabilities, which was higher for PMAS cows. Furthermore, after computing the estimated population transition probabilities among cowtypes, the stationary distribution of the Markov chain, along with bootstrap confidence intervals were computed. The results showed 0.091 (95% CI:0.089,0.092), 0.077 (95 % CI:0.074,0.078), 0.684 (95 % CI:0.067,0.069), 0.138 (95 % CI:0.136,0.139), and 0.009 (95% CI:0.008,0.010) of probability of being in Healthy, Athlete, Clever, Hyperketonemia, and PMAS cowtype, respectively. These estimates represent the proportion of cows belonging to the different cowtypes in a herd; information which may prove useful for herd management. The application of blood biomarker predictions using milk FTIR allows us to investigate differences between predicted cowtype and movements between these states and the association with time to disease. Further research will improve our understanding of the dynamic nature of the transition period.

Developing a predictive model for beta-hydroxybutyrate and non-esterified fatty acids using milk fourier-transform infrared spectroscopy in dairy cows

Negative energy balance following parturition predisposes dairy cattle to numerous metabolic disorders. Current diagnostics are limited by their labor requirements and inability to measure multiple metabolic markers simultaneously. Fourier-transform Infrared spectroscopy (FTIR) data, measured from milk samples, could improve the detection of metabolic disorders using routine milk samples from dairy farms. The objective of this study was to develop a predictive model for numeric values of blood beta-hydroxybutyrate (BHB) and blood non-esterified fatty acids (NEFA). The study utilized a dataset comprised of 622 observations with known blood BHB and blood NEFA values measured concurrently with the milk FTIR data. Using ElasticNet regression on milk FTIR data and production information, we built regression models for numeric blood BHB and blood NEFA prediction and classification models for blood BHB values greater than 1.2 mmol/L and blood NEFA values greater than 0.7 mmol/L. The R² of the best fitting model was 0.56 and 0.51 for log-transformed BHB and log-transformed NEFA, respectively. The BHB classification model had a 90 % sensitivity and 83 % specificity and the NEFA classification model achieved a sensitivity of 73 % and specificity of 74 %. We applied our numeric prediction models to an external dataset (n = 9660) with known blood metabolites to validate the prediction accuracy of log-transformed blood BHB and log-transformed blood NEFA models. Log-transformed BHB root mean square error (RMSE) was 0.4018 and log-transformed NEFA RMSE was 0.4043. The second objective of this study was to develop a categorization for cows as either metabolically disordered or healthy. Responses to negative energy balance in transition cows are related to blood levels of BHB and NEFA. Cows suffering from metabolic disorders without elevated blood BHB values remain unidentified when detection is focused on blood BHB alone. To account for this differentiated metabolic response, we categorized cows as either 'metabolically healthy' or suffering a 'metabolic disorder' by using a combination of blood BHB, blood NEFA, and milk fat to protein quotient. We obtained a balanced accuracy of 94 % for the prediction of cow metabolic status. Direct prediction of metabolic status can be used to identify hyperketonemic cows in addition to cows exhibiting metabolic response patterns not detected by elevated blood BHB alone.

Full model selection using regression trees for numeric predictions of biomarkers for metabolic challenges in dairy cows

Dairy cows suffer poor metabolic adaptation syndrome (PMAS)¹ during early post-calving periods caused by negative energy balance. Measurement of blood beta-hydroxy butyric acid (BHBA)² and blood non-esterified fatty acids (NEFA)³ allow early and accurate detection of negative energy balance. Machine learning prediction of blood BHBA and blood NEFA using milk testing samples represents an opportunity to identify at-risk animals, using less labor than direct blood testing methods. Routine milk testing on modern dairies and computer record keeping provide an immense amount of data which can then be used in machine learning models. Previous research for predicting blood metabolites using Fourier-transform infrared spectroscopy (FTIR)⁴ milk data has focused mainly on individual models rather than a comparison among the models. Full model selection is the process of comparing different combinations of pre-processing methods, variable selection, and statistical learning algorithms to determine which model results in the lowest prediction error for a given dataset. For this project we used a full model selection approach with regression trees (rtFMS)⁵ . rtFMS uses the cross-validated performance of different model configurations to feed a regression tree for selecting a final model. A total of 384 possible model configurations (algorithms, predictors and data preprocessing options) for each outcome (blood BHBA and blood NEFA) were considered in the rtFMS technique. rtFMS allows direct comparison of multiple modeling approaches reducing bias due to empirical knowledge, modeling habits, or preferences, identifying the model with minimal root mean squared prediction error (RMSE)⁶ . An elastic net regression model was selected as the best performing model for both biomarkers. The input data for blood BHBA predictions were FTIR milk spectra, with a second derivative pre-processing, and a filter with 212 wave numbers, obtaining RMSE = 0.354 (0.328-0.392). The best performing model for blood NEFA had input data of FTIR milk spectra, with a second derivative pre-processing, and a filter with 212 wave numbers filter along with the time of milking, obtaining RMSE = 0.601 (0.564-0.654). The comparison of multiple modeling strategies, conducted by rtFMS, present an option for improved FTIR prediction models of blood BHBA and blood NEFA by reducing error due to human bias. The implementation of rtFMS to design future prediction models can guide model inputs and features. Our prediction models have the potential to increase early detection of metabolic disorders in dairy cows during the transition period.

101 Treatment with gonadotrophin-releasing hormone on Day 7 or 21 does not reduce pregnancy loss in dairy heifers receiving invitro-produced embryos

Previous research has shown that induction of an accessory corpus luteum (CL) by administration of gonadotrophin-releasing hormone (GnRH) on Day 5 increases circulating progesterone and reduces pregnancy loss between Days 33 and 60 in heifers receiving invitro-produced (IVP) embryos. Therefore, the objective of the present study was to determine whether timing of induction of an accessory CL influenced pregnancy loss in IVP recipients. Holstein heifers (n=1,658) were synchronized using a modified 5-day CIDR CO-Synch protocol. Briefly, heifers received an intravaginal progesterone device (CIDR) on Day −8, Day −3: CIDR removed and first prostaglandin F2α (PGF2α) treatment, Day −2: second PGF2α, and Day 0: GnRH (G1, 100µg of gonadorelin acetate). Oestrus expression was evaluated on Day 0 with the use of an oestrus detection device. Transfer of fresh IVP embryos was performed on Day 7±1 after G1. At the time of transfer, heifers were randomly assigned, in a 2×2 factorial design, to receive GnRH (200µg of gonadorelin acetate) or remain as Control (untreated) at two different times: Day 7 (time of embryo transfer (ET)) or Day 21. Thus, the following groups were formed: Control-Control (n=410); Control-GnRH (n=409); GnRH-Control (n=419); and GnRH-GnRH (n=420). All heifers were evaluated by transrectal ultrasonography to determine number, size, and location of CL on Days 5, 14, 21, 28, 32, and 60 and pregnancy status on Days 28, 32, and 60. Data collected from each heifer included embryo stage and quality, oestrus expression, body condition score, number of transfers, and technician. Pregnancies per embryo transfer (P/ET) and pregnancy loss data were analysed by logistic regression (SAS 9.4). Ovulation to Day 7 treatment was greater (P&lt;0.01) in GnRH treated heifers (70.0%; 585 out of 836) than in untreated Controls (2.7%; 22 of 819). Ovulation to Day 21 treatment was determined only in animals pregnant at Day 28 and was greater (P&lt;0.01) in heifers treated with GnRH (37.6%; 153 of 407) than in untreated controls (1.0%; 4 of 390). There was no effect of Day 7 treatment (P&gt;0.68) or Day 21 treatment (P&gt;0.18), nor a Day 7×Day 21 treatment interaction (P&gt;0.48) on P/ET at Day 32 or 60 (Table 1). Treatment with GnRH on Day 7 or 21 did not alter pregnancy loss between Days 32 and 60 (P&gt;0.10). Heifers with an accessory CL present at Day 32 (11.7%; 41 of 350) had similar (P=0.55) pregnancy loss compared with heifers with no accessory CL (14.4%; 54 of 375). The number of CL present on Day 32 did not affect (P=0.23) pregnancy loss; however, heifers with 3 CL (4.9%; 3 of 61) tended (P&lt;0.10) to have reduced pregnancy loss compared with heifers with 1 CL (14.4%; 54 of 375) or 2 CL (13.2%; 38 of 289). In conclusion, treatment with GnRH on Day 7 or 21 induced an accessory CL; however, it did not affect P/ET on Days 32 or 60 and pregnancy loss. Table 1.Pregnancies per embryo transfer (P/ET) and pregnancy loss in embryo recipients based on timing of treatment with gonadotrophin-releasing hormone (GnRH) Treatment P/ET Day 32% (n) P/ET Day 60% (n) Pregnancy loss% (n) Day 7 Day 21 Control Control 42.4 (410) 36.1 (410) 15.0 (174) GnRH 46.0 (409) 40.7 (409) 11.7 (188) GnRH Control 43.7 (419) 36.8 (419) 15.9 (183) P-value GnRH 43.1 (420) 38.8 (420) 10.0 (181) Day 7 treatment 0.68 0.72 0.96 Day 21 treatment 0.51 0.18 0.11 Interaction 0.48 0.66 0.71