Álvaro García
Lameness is a painful gait disorder prevalent in dairy cows characterized by abnormal locomotion. With a global prevalence of nearly 23% over the past three decades, its implications for animal welfare, productivity, and farm economics are highly significant. The condition adversely affects milk production, reproduction, and often necessitates premature culling. Lameness assessment traditionally relies on observing the cow’s gait and scoring their locomotion, a time-intensive process impractical for regular monitoring. Consequently, there’s growing interest in automated lameness detection utilizing camera systems. Various studies have explored automating locomotion scoring and lameness detection using video cameras. These systems employ computer vision techniques to localize body parts and compute locomotion traits to identify lameness. Recent advancements in computerized deep learning enable marker-less pose estimation and object detection, facilitating the computation of locomotion traits crucial for lameness assessment. Most studies, however, focus on individual locomotion traits to detect lameness, with limited exploration of combining multiple traits. Moreover, the reliance on individual observer-reliant locomotion scores as “ground truth” introduces subjectivity and reliability issues.
Recent Research
Locomotion scoring in dairy cows employs a 5-point scale, where scores 1 to 5 represent normal gait to severely lame, respectively. A recent experiment conducted in the Netherlands (Russello et al. 2024) showed specific agreement varied across scores, with score 5 showing the lowest consistency. The researchers noted ambiguity in distinguishing between scores 4 and 5 particularly in less-experienced observers. Most studies on lameness detection, however, utilize 2- or 3-level locomotion scores instead of the 5-level scale due to the rarity of severely lame cows in commercial dairy farms and the subjectivity of visual scoring, which can lead to low agreement and reliability. In this experiment they computed six locomotion traits: back posture measurement, head bobbing, tracking distance, stride length, stance duration, and swing duration.
- Back Posture Measurement (BPM): assesses the curvature of the cow’s back during walking. It measures the degree of arching or straightness in the cow’s back as it moves. Changes in back posture indicate discomfort or pain, often associated with lameness or musculoskeletal issues.
- Head Bobbing: It measures the vertical movement of the cow’s head as it walks. Normally, cows have a smooth, rhythmic movement of the head while walking. However, lameness or discomfort can disrupt it, causing the cow’s head to bob or jerk unnaturally.
- Tracking Distance: Refers to the lateral movement of the cow’s body during walking, typically measured on one side (e.g., left or right). It assesses how far the side of the cow’s body moves horizontally as it walks. Lameness can cause asymmetrical movement patterns, leading to differences in tracking distance between the left and right sides.
- Stride Length: Measures the distance covered by a single step of the cow’s limbs during walking. It indicates how far the cow’s limbs extend forward or backward with each step. Changes in stride length can signify alterations in gait patterns, resulting from lameness or other locomotor issues.
- Stance Duration: It evaluates the amount of time the cow’s limbs spend in contact with the ground during walking. It measures the duration of each phase of the walking cycle when a limb is in contact with the ground (stance phase) versus off the ground (swing phase). Variations in stance duration can reflect changes in weight distribution and movement patterns associated with lameness.
- Swing Duration: It quantifies the duration of the swing phase of the walking cycle, during which a limb is lifted off the ground and moved forward. It measures the time taken for the limb to complete its forward movement before contacting the ground again. Alterations in swing duration can indicate abnormalities in limb movement and coordination, which indicate lameness or gait disturbances.
The three most important traits determined in this experiment were back posture measurement, head bobbing, and left tracking distance, in this order. These findings underscore the importance of combining multiple traits for accurate classification and enhancing lameness detection in cattle management.
Impact on Milk Production
There is conflicting evidence regarding how lameness affects milk yield. Some studies indicate reduced yields after diagnosis, while others observe decreases both before and after treatment, with some reporting no change. Accurately evaluating lameness’s impact on production, particularly for high-yielding cows, necessitates assessing milk loss by examining deviations from lactation curves.
A comprehensive study conducted in England encompassed approximately 8000 test-day milk yields from 900 cows spanning 18 months. Clinical lameness instances were identified, and the most common diagnoses included sole ulcer, white line disease, interdigital necrobacillosis, and digital dermatitis. Certain lameness types, such as sole ulcers and white line lesions, may affect milk production well before visible signs emerge, emphasizing the need for timely intervention.
Lameness compromises the potential of high-yielding cows, with decreased milk yield observed from four months pre-diagnosis to five months post-diagnosis. The total estimated reduction in mean milk yield for a cow lame from the fifth month of lactation onwards was almost 360 kg per 305-day lactation period. At a milk price of $15 per hundredweight (cwt) and above, this equates to a substantial economic loss. These insights are critical for evaluating lameness’s economic and health impacts, emphasizing the urgency of early lameness identification and diagnostic improvement.
Farm cases
Here are two cows in this case farm, number 251 (above) and 605 (below), that the 3D camera system has detected dropped almost 1.5 points in their locomotion score (last column to the right). As we mentioned above, what is important is to compare their current production to the expectations based on the expected lactation curve. Their profits on the herd per day are –9.77 and –7.77, respectively.
Cow 251 experienced a notable decline in locomotion score, coinciding with a decrease in milk production and a drop in body condition score (BCS) around 250 days in milk (DIM). Despite this decline, she showed signs of recovery in both production and BCS thereafter. It’s evident that while lameness affected her milk yield, her overall health and nutritional status have been well maintained, as indicated by her ability to regain weight and condition. This highlights the importance of closely monitoring not only milk production but also locomotion score and BCS to assess the overall health and well-being of the cow.
On the other hand, Cow 605 also saw a decline in locomotion score, albeit less severe than Cow 251, which coincided with a drop in milk production around 110 DIM. Despite maintaining her BCS within expectations for her production level and stage of lactation, the significant change in locomotion score warrants urgent attention. This emphasizes the importance of promptly addressing any changes in locomotion score, even if they are not accompanied by significant changes in BCS, as they can be indicative of underlying health issues that may affect productivity and overall welfare.
In summary, both cows demonstrate the importance of considering locomotion score, milk production, and body condition score at the same time. Monitoring all three parameters allows for a more comprehensive assessment of a cow’s health and productivity, enabling early detection and intervention when necessary to optimize herd management and individual animal welfare.
© 2025 Dellait Knowledge Center. All Rights Reserved.
