Implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows
Bovine mastitis is one of the biggest concerns in the dairy industry, where it affects sustainable milk production, farm economy and animal health. Most of the mastitis pathogens are bacterial in origin and accurate diagnosis of them enables understanding the epidemiology, outbreak prevention and ra...
| Main Author: | |
|---|---|
| Format: | Thesis (University of Nottingham only) |
| Language: | English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | https://eprints.nottingham.ac.uk/66056/ |
| _version_ | 1848800292431200256 |
|---|---|
| author | Esener, Necati |
| author_facet | Esener, Necati |
| author_sort | Esener, Necati |
| building | Nottingham Research Data Repository |
| collection | Online Access |
| description | Bovine mastitis is one of the biggest concerns in the dairy industry, where it affects sustainable milk production, farm economy and animal health. Most of the mastitis pathogens are bacterial in origin and accurate diagnosis of them enables understanding the epidemiology, outbreak prevention and rapid cure of the disease. This thesis aimed to provide a diagnostic solution that couples Matrix-Assisted Laser Desorption/Ionization-Time of Flight (MALDI-TOF) mass spectroscopy coupled with machine learning (ML), for detecting bovine mastitis pathogens at the subspecies level based on their phenotypic characters.
In Chapter 3, MALDI-TOF coupled with ML was performed to discriminate bovine mastitis-causing Streptococcus uberis based on transmission routes; contagious and environmental. S. uberis isolates collected from dairy farms across England and Wales were compared within and between farms. The findings of this chapter suggested that the proposed methodology has the potential of successful classification at the farm level.
In Chapter 4, MALDI-TOF coupled with ML was performed to show proteomic differences between bovine mastitis-causing Escherichia coli isolates with different clinical outcomes (clinical and subclinical) and disease phenotype (persistent and non-persistent). The findings of this chapter showed that phenotypic differences can be detected by the proposed methodology even for genotypically identical isolates.
In Chapter 5, MALDI-TOF coupled with ML was performed to differentiate benzylpenicillin signatures of bovine mastitis-causing Staphylococcus aureus isolates. The findings of this chapter presented that the proposed methodology enables fast, affordable and effective diag-nostic solution for targeting resistant bacteria in dairy cows.
Having shown this methodology successfully worked for differentiating benzylpenicillin resistant and susceptible S. aureus isolates in Chapter 5, the same technique was applied to other mastitis agents Enterococcus faecalis and Enterococcus faecium and for profiling other antimicrobials besides benzylpenicillin in Chapter 6. The findings of this chapter demonstrated that MALDI-TOF coupled with ML allows monitoring the disease epidemiology and provides suggestions for adjusting farm management strategies.
Taken together, this thesis highlights that MALDI-TOF coupled with ML is capable of dis-criminating bovine mastitis pathogens at subspecies level based on transmission route, clinical outcome and antimicrobial resistance profile, which could be used as a diagnostic tool for bo-vine mastitis at dairy farms. |
| first_indexed | 2025-11-14T20:49:15Z |
| format | Thesis (University of Nottingham only) |
| id | nottingham-66056 |
| institution | University of Nottingham Malaysia Campus |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-14T20:49:15Z |
| publishDate | 2021 |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | nottingham-660562021-12-31T04:40:28Z https://eprints.nottingham.ac.uk/66056/ Implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows Esener, Necati Bovine mastitis is one of the biggest concerns in the dairy industry, where it affects sustainable milk production, farm economy and animal health. Most of the mastitis pathogens are bacterial in origin and accurate diagnosis of them enables understanding the epidemiology, outbreak prevention and rapid cure of the disease. This thesis aimed to provide a diagnostic solution that couples Matrix-Assisted Laser Desorption/Ionization-Time of Flight (MALDI-TOF) mass spectroscopy coupled with machine learning (ML), for detecting bovine mastitis pathogens at the subspecies level based on their phenotypic characters. In Chapter 3, MALDI-TOF coupled with ML was performed to discriminate bovine mastitis-causing Streptococcus uberis based on transmission routes; contagious and environmental. S. uberis isolates collected from dairy farms across England and Wales were compared within and between farms. The findings of this chapter suggested that the proposed methodology has the potential of successful classification at the farm level. In Chapter 4, MALDI-TOF coupled with ML was performed to show proteomic differences between bovine mastitis-causing Escherichia coli isolates with different clinical outcomes (clinical and subclinical) and disease phenotype (persistent and non-persistent). The findings of this chapter showed that phenotypic differences can be detected by the proposed methodology even for genotypically identical isolates. In Chapter 5, MALDI-TOF coupled with ML was performed to differentiate benzylpenicillin signatures of bovine mastitis-causing Staphylococcus aureus isolates. The findings of this chapter presented that the proposed methodology enables fast, affordable and effective diag-nostic solution for targeting resistant bacteria in dairy cows. Having shown this methodology successfully worked for differentiating benzylpenicillin resistant and susceptible S. aureus isolates in Chapter 5, the same technique was applied to other mastitis agents Enterococcus faecalis and Enterococcus faecium and for profiling other antimicrobials besides benzylpenicillin in Chapter 6. The findings of this chapter demonstrated that MALDI-TOF coupled with ML allows monitoring the disease epidemiology and provides suggestions for adjusting farm management strategies. Taken together, this thesis highlights that MALDI-TOF coupled with ML is capable of dis-criminating bovine mastitis pathogens at subspecies level based on transmission route, clinical outcome and antimicrobial resistance profile, which could be used as a diagnostic tool for bo-vine mastitis at dairy farms. 2021-12-31 Thesis (University of Nottingham only) NonPeerReviewed application/pdf en cc_by https://eprints.nottingham.ac.uk/66056/1/THESIS-06_08_2021_NE_1.pdf Esener, Necati (2021) Implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows. PhD thesis, University of Nottingham. Machine learning MALDI-TOF Bovine mastitis Antimicrobial resistant Diagnostic solutions Bioinformatics Dairy farms Biomarkers |
| spellingShingle | Machine learning MALDI-TOF Bovine mastitis Antimicrobial resistant Diagnostic solutions Bioinformatics Dairy farms Biomarkers Esener, Necati Implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows |
| title | Implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows |
| title_full | Implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows |
| title_fullStr | Implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows |
| title_full_unstemmed | Implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows |
| title_short | Implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows |
| title_sort | implementation of machine learning for the evaluation of mastitis and antimicrobial resistance in dairy cows |
| topic | Machine learning MALDI-TOF Bovine mastitis Antimicrobial resistant Diagnostic solutions Bioinformatics Dairy farms Biomarkers |
| url | https://eprints.nottingham.ac.uk/66056/ |