Activity Number:
|
430
- Contributed Poster Presentations: Section on Statistical Consulting
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, July 31, 2018 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Statistical Consulting
|
Abstract #329842
|
|
Title:
|
Development and Comparison of Predictive Models for Woody Breast in Commercial Broilers
|
Author(s):
|
Andy Mauromoustakos* and JUAN P CALDAS-CUEVA and CASEY OWENS-HANNING
|
Companies:
|
Univ. of Arkansas and University of Arkansas and University of Arkansas
|
Keywords:
|
ordinal logistic regression;
generalized regression ;
classification trees;
NN;
model comparison;
validation
|
Abstract:
|
Woody breast (WB) causes a significant economic loss to poultry producers and the lack of an objective tool to identify WB is a contributing factor. The aim here is to develop and compare several different predictive models to predict WB using image analysis of broiler carcasses. Images of male broiler carcasses of high yielding commercial strain were captured prior to evisceration. Whole breast fillets were scored for WB severity based on the tactile evaluation (0 - 1 as normal; 1 or 1.5 - 3 as woody). Broiler carcass images were processed and analyzed using ImageJ software. Ten model potential inputs predictors from the image analysis included M1: breast width in the cranial region; M2: a vertical line from the tip of keel to 1/5th of breast length; M3: breast width at the end of M2; M4: angle formed at the tip of keel and extending to outer points of M3; M5: area of the triangle formed by M3 and lines generated by M4; M6: area of the breast above M3; M7: M6 minus M5. In addition, three ratios [M8 (M3/M1), M9 (M3/M2), and M10 (M7/M5)] were considered. Validation was used to select the "best" models considered on new data for the ability to predict both WB and compression force.
|
Authors who are presenting talks have a * after their name.