Download the poster

Campagnoli A., Paltanin C., Savoini G., Baldi A., and L. Pinotti
Department of Veterinary Science and Technology for Food Safety, Veterinary Faculty
University of Milan - Italy

Aim of this study was to evaluate the potential of  image analysis measurements in combination with the official analytical method for the detection of constituents of animal origin in feedstuffs [i.e. the microscopic examination technique as described in Commission Directive 2003/126/EC of 23 December 2003 (OJ L 339, 24.12.2003, p. 78.)] in distinguishing between land animals (i.e. poultry vs. mammals). For this purpose reference samples containing poultry (AV) or mammalian (MAM) meat and bone meals (Agricultural Research Centre of Gembloux, Belgium, and SAFEED-PAP Project; VSA, University of Milan) were analysed. Sediment fractions of each samples were observed with a compound microscope (Olympus BX41, Germany) at several magnifications. Through a digital camera and an image analysis software (Image-for Plus 4.5.1, Media Cybernetics Inc., Silver Springs, USA), we obtained 823 bone fragment lacunae images at X40. Images have been processed and elaborated in order to obtain for each lacuna a monochrome mask on which several measurements were performed. On each lacuna 26 geometric variables were obtained. Data were analysed by ANOVA (GLM procedure) and by PROC BOXPLOT procedure of SAS statistic software. In order to show the variability of the most discriminant variables we have performed graphic test (box-plot) for mean and median comparisons. Results obtained in the present study indicated that even though most of variables measured were significantly (P <0.001) different between mammalian and poultry in term of mean, only area (area of the object), axis minor (length of minor axis of ellipse with same moments of order 0, 1 and 2 as object), area polygon (area included in the polygon defining the object outline), diameter min (length of shortest line joining two points of object outline and passing through the centroid), feret min (smallest caliper (feret) length), radius min (minimum distance between object centroid and outline), roundness [reports the roundness of the object, as determined by the following formula (perimeter2)/(4*pi*area)], and size width [feret diameter (i.e. caliper length) along minor axis of object], produced discrimination among class (i.e. poultry vs. mammals). In conclusion, data here presented indicate that some of the variables/descriptors provided by image analysis related to lacunae dimensions and features appear promising  for a reliable distinction between animal meal at the level of vertebrate classes, while for further characterisation, at higher taxonomic level, contribution of morphological variables deserve further investigation.

Keywords:
Meat and bone meals, bone lacunae, microscopic method,  image analysis

#“This work is funded by the European Commission, within the framework of the FOOD-CT-2006-036221 Project SAFEED-PAP ("Detection of presence of species-specific processed animal proteins in animal feed"; http://safeedpap.feedsafety.org) under the specific FP6 research and technological development program "Integrating and strengthening the European Research Area".
The information contained in this work reflects the authors’ views; the European Commission is not liable for any use of the information contained therein”.