TAKE HOME MESSAGES

• A modified ultrasonic instrument was able to measure changes in subdermal fat thickness along lactation in dairy cattle.
• Digitized cross sections of ultrasonic scans, followed by identification and measurement of subdermal fat via an algorithm, was a promising method for development of a sensor for scoring body condition in dairy cows.

INTRODUCTION

Body condition score (BCS) has become recognized as a valuable tool to monitor the energy reserves of the dairy cow. Limitations for widespread use of BCS include:

1. even experienced scorers assign somewhat different scores for the same cow with a specific amount of subdermal fat reserves; and
2. scoring BCS is labor intensive to follow BCS changes in individual cows as lactation progresses.

We have examined ultrasonic measurement of subdermal fat thickness for assigning BCS in dairy cattle, with the goal of being able to computerize the results to the recognized dairy cattle visual scoring system between 1.0 and 5.0 scale with .25 units. Although ultrasonic measurement of subdermal fat has become a widespread tool for production systems in beef cattle and swine, we quickly found that assigning BCS in dairy cattle is much challenging than for routine use in beef or swine because the fat reserves in dairy cattle are more dispersed, and a distinguished subdermal fat layer is very sparse in comparison to beef and swine.

PROCEDURE

We formed a team with scientists from Israel, using an approach that allowed the measured ultrasonic values to be digitized, and, as a long range goal, automatically analyzed by a computer system for rapid interpretation . The specific goals of our studies were to:

1. determine the overall feasibility of the proposed technique; and
2. to determine which body sites were the optimum locations to measure subdermal fat thickness.

Several technical problems were associated with adapting the equipment for the approach that we used. Our desire was to measure and digitize a cross section of the subdermal fat layer. We obtained a small 7.5 MHZ probe, and had the instrument modified to allow the peaks of different tissue densities to be identified automatically by a special algorithm. This approach was to allow many measurements to be analyzed rapidly by a computer. We followed body weight changes, BCS, and subdermal fat thickness at 14-day intervals between week 7 pre-calving to 120 days post-calving in five cows. Ultrasonic measurements were obtained between the 12th and 13th rib, at the loin between the 3rd and 4th transverse process and 5 cm from the vertebrae, and on the center of the rump.

TABLE 1. Body Condition Score, and Subdermal Fat Thickness at Three Body Sites for Fat and Thin Cows Three Weeks Before Calving

RESULTS

We were able to develop an algorithm that successfully separated skin thickness and subdermal fat thickness, Figure 1. This accomplishment was particularly useful since individual cows have their individual skin thickness that is difficult to differential from subdermal fat thickness. We found in dairy cattle, that the body condition is partly a function of fat that is dispersed in interstitial tissue instead of a definite layer of fat as occurs in beef cattle or swine. As a result, all body sites showed slight differences, but we concluded that any of the body sites that we tested were acceptable. The ultrasonic system was able to measure changes in subdermal fat thickness that corresponded to body weight changes and body condition scores along lactation. Patterns of subdermal ultrasonic measurements in fat and thin cows were able to differentiate between the classes of BCS, Table 1.

CONCLUSIONS

We concluded that the use of digitized cross sections of ultrasonic scans, followed by identification and measurement of subdermal via an algorithm for analysis of peak values, was a promising method for development of a sensor for scoring body condition in dairy cattle. The method removes most of the subjectivity of BCS scoring, but substantial development still is to occur to reduce the time required for data collection (for example, a robotic sensor that could scan the cows in the milking parlor) before expecting a commercial ultrasonic sensor for scoring cows for BCS automatically.

Figure 1. Typical video cross section image (A) showing ultrasound density of various tissue (top) and digitized values of peaks (bottom). Computer printouts (B) may be analyzed by algorithm to determine edge of skin, and to separate thickness of skin and thickness of subdermal fat layer.

Figure may be viewed in PDF