SUBCUTANEOUS FAT DEPTH MAGNITUDE INFLUENCES ITS MEASUREMENT ERRORS: A SIMULATION STUDY

• Type: Article in International Conference Proceedings Book
• Date: 2010

Authors

Cadavez, VAP (Author) Other; Amaro, JR (Author) Other; Antonio J M Fonseca (Author) ICBAS

Conference proceedings International

Title: 6TH INTERNATIONAL CONFERENCE ON SIMULATION AND MODELLING IN THE FOOD AND BIO-INDUSTRY (FOODSIM 2010)

Pages: 118-121

6th International Conference on Simulation and Modelling in the Food and Bio-Industry

Braganca, PORTUGAL, JUN 24-26, 2010

Indexing

ISI Web of Knowledge - 0 Citations

Scopus - 0 Citations

Other information

Authenticus ID: P-003-DGG

Abstract (EN): The objectives of this study were to evaluate the impact of proportional and absolute errors on subcutaneous fat depth (SFD) measurements, and the effects on the stability of models to predict the lean meat proportion (LMP) of lamb carcasses. Ninety eight lambs (72 males and 26 females) of Churra Galega Bragancana breed were slaughtered, and carcasses were weighed and hot carcass weight (HCW) was recorded approximately 30 min after exsanguinations. During carcasses quartering a calliper was used to perform SFD measurements, over the maximum depth of longissimus muscle (LM), between the 12th and 13th ribs (C12), and between the 3rd and 4th lumbar vertebrae (C3). A computer program was written in order to simulate measurement errors for C12 and C3 measurements. Two scenarios were simulated, and C12 and C3 were contaminated with: 1) proportional errors of 5, 10, and 15%, and 2) absolute errors of 0.25, 0.50, and 0.75 mm. Simple linear models to predict LMP were developed using as independent variables: 1) the measured (original) SFD measurements, and 2) the biased SFD measurements. The coefficient of determination and the residual standard deviation were computed. Our study demonstrates that measurement errors can have a high impact on the SFD measurements, and on models stability. We conclude that SFD measurements of higher magnitude should be preferred as predictors of LMP since they are less influenced by measurement errors, thus contributing to more stable regression models.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 4