When Buffalo Mastitis is Suspected, Can Udder Ultrasound Help Pinpoint the Affected Quarter?

When mastitis is suspected in a buffalo, identifying the affected animal is only part of the clinical problem. The next question is more specific: Which part of the udder is affected, and can the lesion be identified early enough to guide management? 

Udder ultrasonography offers a non-invasive way to examine the teat, teat canal, gland cisterns, and mammary parenchyma. It can also reveal structural changes associated with udder disease, making it a potential adjunct to conventional mastitis assessment^[1]. 

What does ultrasound add? 

Mastitis can produce changes within the mammary tissue that are not necessarily obvious from external examination alone. Ultrasound has been used to visualize pathological changes associated with Staphylococcus aureus infection of the udder^[2]. 

The challenge is interpreting these images consistently. Complex ultrasound images, differences in clinical experience, and limited availability of skilled personnel can make image-based assessment difficult. 

Could automated image analysis help? 

What did the buffalo data show?^[2] 

In a study involving 271 buffaloes and 3,054 udder ultrasound images, a deep-learning-based detection system was developed to distinguish healthy and mastitis-affected udders. 

The researchers assessed the system using two SCC thresholds. When SCC ≥ 2 × 10^5 cells/mL was used to define mastitis, the model achieved 70.02% accuracy, 77.93% specificity, 63.11% sensitivity, and a 69.21% F1-score. 

When the threshold was increased to SCC ≥ 4 × 10^5 cells/mL, performance improved. Accuracy reached 75.93%, while specificity was 80.23%, sensitivity 70.35%, and the F1-score 71.79%. 

The corresponding AUC values were 0.77 and 0.83, respectively. Thus, within the conditions evaluated, the higher SCC threshold produced better discrimination between healthy and mastitis-affected udders. 

Why does quarter-level detection matter? 

For the practicing veterinarian, identifying mastitis at the whole-animal level may not be enough. The more useful question can be whether the affected quarter can be localized. 

In this work, the ultrasound-based system was able to identify the area of mastitis at the 1/4 level. The configuration using SCC ≥ 4 × 10^5 cells/mL as the mastitis threshold was considered the more suitable option for on-site buffalo mastitis detection^[2]. 

That localization could provide useful information when subsequent treatment and production-management decisions are being considered. 

The ultrasound findings also have a clinical basis. As mastitis becomes more severe, changes in mammary parenchymal echotexture can become more pronounced. Previous findings cited in the source showed a positive relationship between ultrasound texture characteristics and SCC, while another report found increased pixel values and pixel heterogeneity with increasing mastitis severity^[3],[4]. 

Where could it fit in practice? 

For buffalo farms, particularly those operating with limited diagnostic resources, ultrasound-based assessment may offer an additional diagnostic route. Ultrasound equipment is already commonly used for reproductive work in cattle-farm settings, which may make udder imaging more accessible than technologies requiring specialized equipment or extensive datasets. 

However, ultrasound image analysis should be viewed as diagnostic support rather than a replacement for veterinary assessment. Conventional approaches, including SCC, pathogen identification, milk conductivity testing, and biomarker assessment, have their own diagnostic strengths and practical limitations^[2]. 

The practical value of this approach lies in adding another layer of information: Using udder imaging not only to detect possible mastitis, but potentially to identify the affected quarter and support more informed herd-management decisions.

References:

1. EZ Kotb E, El-Fattah A, Ola AA, Azab AM, Leil AZ. Ultrasonography, histopathological udder alterations and bacteriological investigations for diagnosis of mastitic goats. Journal of Applied Veterinary Sciences. 2020 Apr 25;5(2):77-86. https://javs.journals.ekb.eg/article_85593_e7cd75ad7ae6ccc823cc42bb99075af7.pdf
2. Zhang X, Li Y, Zhang Y, Yao Z, Zou W, Nie P, Yang L. A new method to detect buffalo mastitis using udder ultrasonography based on deep learning network. Animals. 2024 Feb 23;14(5):707. https://www.mdpi.com/2076-2615/14/5/707
3. Zhang X, Ahmad MJ, An Z, Niu K, Wang W, Nie P, Gao S, Yang L. Relationship between somatic cell counts and mammary gland parenchyma ultrasonography in buffaloes. Frontiers in veterinary science. 2022 Mar 21;9:842105. https://www.frontiersin.org/journals/veterinary-science/articles/10.3389/fvets.2022.842105/pdf
4. Abdullah OM, Aslam S, Khan MA, Mushtaq H, Hassan M, Ijaz M. Healthy, sub-clinical, and clinical mastitis in Holstein–Friesian cattle: A comparative echotextural and electrical conductivity study. South African Journal of Animal Science. 2023 Apr 1;53(2):221-30. https://www.scielo.org.za/pdf/sajas/v53n2/07.pdf