Frequently Asked Questions about precision and accuracy of measurements

• How do we know that our system is measuring accurately?
• How do we know that our system is measuring precisely?
• Why is the precision different for different EchoMRI™ systems?
• What is the difference between standard precision, high precision, and extra high precision modes?
• What are the possible sources of error for the data measured by the EchoMRI™ system?
• How many measurements will it take to achieve the desired precision for my experiment?

Q. How do we know that our system is measuring accurately?
A. In order to validate the system performance, a simple experiment can be performed. Take muscle from a rat or a lean chicken breast, measure it, and then combine it with a known amount of canola oil warmed to temperature between 30°C and 40°C, and measure again. Repeat both measurements about 10 times to gather appropriate statistics. The difference should reproduce the weight of the lean and canola oil.

^Top

Q. How do we know that our system is measuring precisely?
A. Our determination of precision is based on repeatability. For example, one can measure 1 animal many times, expressing the occurrence of error in terms of Standard Deviation; alternatively, one can measure many animals once and express the occurrence of error in terms of the Standard Error of the Mean. For our instruments, we take one animal and do repeated measurements under the assumption that animal's properties will not change over a short period of time. We believe that this is the best means for defining precision.

^Top

Q. Why is the precision different for different EchoMRI™ systems?
A. The pricing and precision for EchoMRI™ technology is related to the magnet and antenna sizes that are based on the weight ranges of your animal subjects. Before you order any equipment, we will consult you regarding your required weight ranges, and recommend appropriate solutions.

The key advantage of optimizing your system for different weight ranges is to maintain a higher relative precision. The narrower the weight range is - the higher the relative precision. For example, the precision of measured fat for a 1200 grams rat will be about 3 grams with the system for rats between 500 to 1200 grams. However, if you were to use the same system for a 100 grams rat, the relative measured precision of 3 grams for fat determination would not be acceptable.

In order to serve you better, Echo Medical Systems provides several standard configurations of equipment for mice and rats based on weight ranges 20-100, 100-700, 250-900, and 500-1200 grams. We also offer custom weight ranges as an option to our customers, and we encourage you to discuss your requirements with us.

^Top

Q. What is the difference between standard precision, high precision, and extra high precision modes?
A. The major source of imprecision in NMR methodology is due to subject movement. However, in cases where random error plays a significant role in the analysis, it can be reduced by increasing the number of scans conducted on the subject and then averaging all of the data points, thus reducing the effect of the random error. The main source of random error in NMR systems comes from antenna noise.

^Top

Q. What are the possible sources of error for the data measured by the EchoMRI™ system?
A. These include subject movement, random noise (from the antenna which can be mitigated by multiple subject scans), and environmental noise (which we shield for).

^Top

Q. How many measurements will it take to achieve the desired precision for my experiment?
A. It depends. The more you measure a single sample (repeatability), the more you will reduce random noise. However, there is no reason to measure to a precision of more than 0.1 grams for mice weighing under 100g. Based on the size of the animals, the systems are pre-calibrated to give the researcher the most precise measurement within a reasonable sample size. The systems can be programmed and adjusted based on a researcher's precision needs, and the process can be optimized (size of holder, time of management, etc.).

^Top