Mastitis detection in dairy cows by application of neural networks

Cavero, D.; Tölle, K.-H.; Henze, C.; Buxadé, C.; Krieter, Joachim

doi:10.1016/j.livsci.2007.05.012

Cited by 70 publications

(67 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Generally, the models that increase SENS cause a decrease in PPV, due to the higher number of false positives; high SENS is published in some works, but SPEC is not published (Maatje et al 1992(Maatje et al , 1997, leaving the question of the proportion of false positive cases open. Other published results obtained higher SENS with low SPEC (Zecconi et al 2004;Cavero et al 2008). High SENS and SPEC have been published for clinical and subclinical mastitis in cows using the increases above moving average (Mele et al 2001;Cavero et al 2007), Tracking Signal Method (De Mol et al 1999Mele et al 2001) or Fuzzy Models (Cavero et al 2006) that combine one or more variables (for example, temperature, flow or yield) with EC.…”

Section: Resultsmentioning

confidence: 98%

“…In subclinical mastitis, Nielen et al (1995a) reported 54% SENS and 92% SPEC. Cavero et al (2008) reported 84.2% SENS and 51.1% SPEC for SCC > 100 000 cells/ml detection and 78.6% SENS and 74.9% SPEC for SCC > 400 000 cells/ml detection. The fuzzy logic method uses EC combined with another variable; Kamphuis et al (2008b) employed EC combined with SCC and reported 80% SENS in clinical mastitis with a low true positive value (32%); Cavero et al (2006) reported 83-92% SENS and 75-93% SPEC combining EC, yield, and milk flow variables in the algorithm.…”

Section: Introductionmentioning

confidence: 94%

See 1 more Smart Citation

Goat mastitis detection using daily records of milk conductivity: comparative results of different algorithms

Romero¹,

Reinemann²,

Alejandro³

et al. 2014

CZECH J ANIM SCI

View full text Add to dashboard Cite

Milk electrical conductivity is employed for mastitis detection in cows due to its automation, low cost, and infection detectability at early stage. Nevertheless, the number of publications about its use in dairy goats is scarce. The aim of this study was to check and compare the detectability of goat mastitis (sensitivity and specificity) using different algorithms, constructed with individual daily conductivity data from glands, in order to improve the know how about the potential of this variable for goat mastitis detection. A total of 18 goats (8 primiparous and 10 multiparous) free of mastitis were used, and gland milk conductivity was daily monitored. After 16 days of monitoring, some unfavourable situations for gland health were simulated in order to increase the cases of infection. Once infection was established (9 goats and 12 glands got infected), the experiment continued for further 16 days. A total of 19 different algorithms that employed conductivity data from gland were designed; they were tested using gland milk conductivity (EC) and ratio of EC of collateral glands in the same goat (RAT EC ). The algorithms were tested in all the animals and intramammary infection detection ability characteristics (sensitivity (SENS), specificity (SPEC), positive predictive value (PPV), and negative predictive value (NPV)) were recorded. All clinical cases were detected (n = 2, 100% SENS) with all the algorithms. Best global SENS (clinical and subclinical, 33.3-58.3%) and SPEC (77.8-100%) were similar to results reported in previous studies in cows, and obtained with algorithms ARIMA and Rule 1 (3 standard deviations of data). The best algorithms to use in mastitis detection depend on the prevalence and type of mastitis. EC ARIMA and Rule 1 algorithms detected the most severe cases on-line and quickly, with a low proportion of false positives.

show abstract

Section: Resultsmentioning

confidence: 98%

Section: Introductionmentioning

confidence: 94%

Goat mastitis detection using daily records of milk conductivity: comparative results of different algorithms

Romero¹,

Reinemann²,

Alejandro³

et al. 2014

CZECH J ANIM SCI

View full text Add to dashboard Cite

show abstract

“…Because an Se of at least 70% combined with an Sp of at least 99% is desired , the CM detection performance of this AMS is suboptimal. Although new detection models were developed in several previous studies (e.g., de Mol and Ouweltjes, 2001;Cavero et al, 2008;Kamphuis et al, 2010), these models were unable to improve detection performance to the extent that the detection of CM cases remained satisfactory and the number of false-positive alerts was reduced to a reasonable level. Because of the suboptimal detection performance, and particularly the unsatisfactory Sp, interpretation of the alert lists is difficult.…”

Section: Discussionmentioning

confidence: 99%

“…A general complaint of dairy farmers working with an AMS is the relatively large number of false-positive alerts on the mastitis alert lists. Several detection models were developed with the aim of reducing the number of false-positive alerts (e.g., de Mol and Ouweltjes, 2001;Cavero et al, 2008;Kamphuis et al, 2010). The sensitivity (Se) and specificity (Sp) of these models, however, remain too low to substantially reduce the number of false-positive alerts and at the same time retain a sufficient detection of true cases.…”

Section: Introductionmentioning

confidence: 99%

Discriminating between true-positive and false-positive clinical mastitis alerts from automatic milking systems

Steeneveld

Gaag

Ouweltjes

et al. 2010

Journal of Dairy Science

View full text Add to dashboard Cite

Automatic milking systems (AMS) generate alert lists reporting cows likely to have clinical mastitis (CM). Dutch farmers indicated that they use non-AMS cow information or the detailed alert information from the AMS to decide whether to check an alerted cow for CM. However, it is not yet known to what extent such information can be used to discriminate between true-positive and false-positive alerts. The overall objective was to investigate whether selection of the alerted cows that need further investigation for CM can be made. For this purpose, non-AMS cow information and detailed alert information were used. During a 2-yr study period, 11,156 alerts for CM, including 159 true-positive alerts, were collected at one farm in the Netherlands. Non-AMS cow information on parity, days in milk, season of the year, somatic cell count history, and CM history was added to each alert. In addition, 6 alert information variables were defined. These were the height of electrical conductivity, the alert origin (electrical conductivity, color, or both), whether or not a color alert for mastitic milk was given, whether or not a color alert for abnormal milk was given, deviation from the expected milk yield, and the number of alerts of the cow in the preceding 12 to 96 h. Subsequently, naive Bayesian networks (NBN) were constructed to compute the posterior probability of an alert being truly positive based only on non-AMS cow information, based on only alert information, or based on both types of information. The NBN including both types of information had the highest area under the receiver operating characteristic curve (AUC; 0.78), followed by the NBN including only alert information (AUC = 0.75) and the NBN including only non-AMS cow information (AUC = 0.62). By combining the 2 types of information and by setting a threshold on the computed probabilities, the number of false-positive alerts on a mastitis alert list was reduced by 35%, and 10% of the true-positive alerts would not be identified. To detect CM cases at a farm with an AMS, checking all alerts is still the best option but would result in a high workload. Checking alerts based on a single alert information variable would result in missing too many true-positive cases. Using a combination of alert information variables, however, is the best way to select cows that need further investigation. The effect of adding non-AMS cow information on making a distinction between true-positive and false-positive alerts would be minor.

show abstract

“…milk yield for the fi rst lactation and milk yield for the second lactation) due to the limited sample size. It is diffi cult to compare directly the results of the present work with those of other authors because most research in this fi eld [1,[34][35][36][37][38][39] has been aimed at detecting single mastitis cases on the farm rather than diagnosing cows that are susceptible to this disease. Nevertheless, some comparisons can be made.…”

Section: Discussionmentioning

confidence: 99%

Identification of Cows Susceptible to Mastitis based on Selected Genotypes by Using Decision Trees and A Generalized Linear Model

et al. 2016

View full text Add to dashboard Cite

The aim of the present study was to: 1) check whether it would be possible to detect cows susceptible to mastitis at an early stage of their utilization based on selected genotypes and basic production traits in the fi rst three lactations using ensemble data mining methods (boosted classifi cation tress -BT and random forest -RF), 2) fi nd out whether the inclusion of additional production variables for subsequent lactations will improve detection performance of the models, 3) identify the most signifi cant predictors of susceptibility to mastitis, and 4) compare the results obtained by using BT and RF with those for the more traditional generalized linear model (GLZ). A total of 801 records for Polish Holstein-Friesian Black-and-White cows were analyzed. The maximum sensitivity, specifi city and accuracy of the test set were 72.13%, 39.73%, 55.90% (BT), 86.89%, 17.81%, 59.49% (RF) and 90.16%, 8.22%, 58.97% (GLZ), respectively. Inclusion of additional variables did not have a signifi cant effect on the model performance. The most signifi cant predictors of susceptibility to mastitis were: milk yield, days in milk, sire's rank, percentage of Holstein-Friesian genes, whereas calving season and genotypes (lactoferrin, tumor necrosis factor alpha, lysozyme and defensins) were ranked much lower. The applied models (both data mining ones and GLZ) showed low accuracy in detecting cows susceptible to mastitis and therefore some other more discriminating predictors should be used in future research.

show abstract

Mastitis detection in dairy cows by application of neural networks

Cited by 70 publications

References 16 publications

Goat mastitis detection using daily records of milk conductivity: comparative results of different algorithms

Goat mastitis detection using daily records of milk conductivity: comparative results of different algorithms

Discriminating between true-positive and false-positive clinical mastitis alerts from automatic milking systems

Identification of Cows Susceptible to Mastitis based on Selected Genotypes by Using Decision Trees and A Generalized Linear Model

Contact Info

Product

Resources

About