Final Report Summary - MACSYS (Development of an objective method to perform quality classification of comminuted poultry meat) Executive Summary:Executive SummaryDeboning of meat is an important technology to improve yield and sustainability of meat production. But Mechanically Separated Meat (MSM) is generally considered to be of an inferior quality, and its use is subject to strict regulations. In contrast, manually separated meat is accepted as regular meat, and not subject to the same restrictions. But development of the technology for mechanical processes of meat-bone separation makes it possible to produce a meat raw material that cannot be distinguished from regular minced meat. There are no objective reasons to classify all MSM as being of inferior quality. EU regulations, however, still differ between MSM and meat strictly based on the production method. MACSYS has addressed a major market opportunity for the SME consortium through the development of new, efficient and objective solutions for the quality classification of comminuted poultry meat related directly to degree of degradation of muscle tissue. A prototype of a spectroscopy-based fast-track system has been developed to monitor meat quality in processing lines. Software for an automatic image analysis assessment of meat quality has been developed, based on a commonly accepted immuno-histochemical method differentiating between intact and non-intact muscle fibre membranes. To achieve these outputs of the project five project results were defined at project start: N°1 Preparation, handling, documentation and characterization of standard meat samples, N°2 New antibodies and histochemical staining techniques, N°3 Common EU histochemical method to assess muscle fibre structure, N°4 Automated histological image analysis system, and N°5 Fast-track method and system for wide in-line industrial application. These project results have been obtained through 18 deliverables, which have been submitted during the 27 months project period. The main achievement is the agreement of a common immune-histochemical method to quantify muscle fibre degradation which has further been supported with the method being brought to a Cen Workshop Agreement (CEN-CENLEC/WS 83). This result is very important as a first step for the two other main outcomes of MACSYS – the cloud-based automated histochemical image analysis system, which have been developed to quantify the muscle fibre degradation based on the immuno-histochemical staining, and the developed near-infrared spectroscopy-based (NIRS) prototype for in-line measurement of muscle fibre degradation in comminuted poultry meat, as this method should be calibrated against the immuno-histochemical method. The results can now be explored by the participating companies and in future may benefit the whole poultry industry which however will require a change in legislation.Project Context and Objectives:Deboning of meat is an important technology to improve yield and sustainability of meat production. But Mechanically Separated Meat (MSM) is generally considered to be of an inferior quality, and its use is subject to strict regulations. In contrast, manually separated meat is accepted as regular meat, and not subject to the same restrictions. But development of the technology for mechanical processes of meat-bone separation today makes it possible to produce a meat raw material that cannot be distinguished from regular meat mince. There are no objective reasons to classify all MSM as being of inferior quality. EU regulations, however, still differ between MSM and meat strictly based on the production method as there are currently no accepted method available by which degree of degradation of muscle tissue can be measured which is a prerequisite for quantification of loss or modification of muscle fibre structure. 1.1 ObjectiveThe overall objective of MACSYS is to overcome scientific and technical barriers associated with the development of efficient and objective solutions for the quality classification of comminuted poultry meat. This implies the development of a fast-track system, based on spectra and/or impedance analysis which can be integrated on the production line and which will be able to measure and classify different qualities of comminuted poultry meat. Furthermore, software for analysis of histological slides that enables an automatic measurement of meat parameters will be developed, together with a standard EU histochemical method that identifies muscle structure degradation to be used by laboratories and authorities within the EU.To fulfil this overall objective of MACSYS, seven scientific and technological objectives have been identified and are presented in table 1.To achieve these objectives of the project five project results were defined at project start: • N°1 Preparation, handling, documentation and characterization of standard meat samples, • N°2 New antibodies and histochemical staining techniques, • N°3 Common EU histochemical method to assess muscle fibre structure, • N°4 Automated histological image analysis system, and • N°5 Fast-track method and system for wide in-line industrial application.and they were obtained through research and development in 5 Work packages (WP3-7), 1 exploitation and dissemination work package (WP2) and one management work package (WP1) as shown in figure 1. The work packages have been connected by the flow of samples, results and generated knowledge among WP’s and by involving inputs from RTD, SME and LE partners in the fulfilment and dissemination of the project results. he content of the RTD WP (3-7) are as follows:• WP3 establish a protocol for the preparation, handling and documentation of standard comminuted meat samples for the experimental work in WP4-7, and coordinate the development of a database for sample documentation and to save the results of all analytical work performed throughout the project.• WP4 will focus on establishing a common histochemical method for the assessment of muscle fibre degradation. It will provide the basic data for developing the image analyses system and the online systems and will evaluate the possibility of including biochemical data into the histochemical analysis.• WP5 develops and tests the image analysis algorithms for the development of the automated histological image analysis solution.• WP6 involves the development of a fast-track system, based on spectroscopy and impedance measurements, for the evaluation of meat quality. Statistical models for the correlation between spectral and electrical signals and meat quality parameters will also be developed in this task. • WP7 is dedicated to the integration of the fast-track method in processing machines, and to its testing and subsequent validation at meat production plants.MACSYS has addressed a major market opportunity for the SME consortium through the development of new, efficient and objective solutions for the quality classification of comminuted poultry meat related directly to quality traits. A spectroscopy-based fast-track system has been developed to monitor meat quality in processing lines and software for an automatic image analysis assessment of meat quality has been developed, based on a commonly accepted immuno-histochemical method agreed upon at a CEN workshop (CEN-CENELEC/WS 83). These project results have been obtained through 18 deliverables, which have been submitted during the 27 months project period. Project Results:The scientific and the technological achievements of the project can be divided in 5 main results.2.1 Preparation, handling, documentation and characterization of standard meat samplesFirst of all a description of the samples, the preparation, the handling and the characterization of the standard samples as well as industrial samples used in the project was done. The standard samples were mixtures of emulsified meat and coarsely minced meat in different concentrations, in steps of 10% from completely emulsified meat in one end to pure coarsely minced meat at the other end. The mixtures were produced for both thigh and breast muscle. The industrial samples were obtained from three different slaughterhouses in Denmark which included soft separated and hard separated samples, different machine types and different settings of machines. The collection also included two samples of manually separated meat, which by present legislation is considered as meat. All together 14 industrial samples were described. Both standard and industrial samples are documented in a common database with a comprehensive description of the circumstances for sampling of the individual samples. 2.2 New antibodies and histochemical staining techniquesThe application of new antibodies for visualization of cell outlining’s derived from chicken tissue was tested against the ones previously used. The advantage of using immunohistochemistry is that individual tissues and proteins can be specifically stained whereas with standard histochemical staining methods different tissue are identified basically on differences in staining intensity of the dye. The new myosin antibody developed from chicken myosin clearly was superior to the ones used before developed from pig myosin and they resulted in a more uniform staining. Whereas using an antibody for laminin developed from chicken was only marginally better from the ones previously used.Four different methods to measure degree of degradation were tested: the Danish laminin method, a Toluidine blue (TB) and a haematoxylin eosin (HE) method developed in Germany at the Max Rubner Institut and an English TB method developed at Leatherhead Food Limited. The three latter methods are based on subjective evaluation as opposed to the Danish which is fully objective. Both the English and the German methods were developed to give an overall assessment of the quality of the products and to separate them into different classes. As the legislation refer to alterations or degradation of muscle structure the technicians who performed the analysis were asked to evaluate the relative muscle structure degradation in the samples and this is the value used in the calculations. It should be emphasised that the laminin method is exclusively designed to measure degree of degradation. A basic test of the ability of the methods to measure degree of degradation of muscle tissue was to calculate the correlation of the individual methods to a hypothetical straight line. This line was based on the expected values of the standard mixtures of emulsified and coarsely minced meat made from breast and thigh chicken meat. The results from the standard samples showed that the laminin method had a higher correlation (r value) for both muscles which implies that the laminin is the most reliable of the methods. A further advantage was that the method was less challenging for development of the automated image analysis system. In table 2 the individual r and R2 values from the methods and muscles are displayed. These results led to the laminin method being the one, considered by the consortium, to be the best candidate for a future standard method for measuring degree of degradation of muscle tissue. The next step was to evaluate the potential of the methods used on the industrial samples. Much to our surprise the results showed that the range of degradation was much less than expected for the Danish and German methods, the range was only 30% from the most to the least degraded. This of course limits the number of qualities that could be separated. At present 2 qualities can be clearly differentiated but a possibility of defining 3 quality groups may be realistic with further validation of the method. The present concept that the meat separated by highest pressure is of very low quality because of a very high level of degradation of muscle structure cannot be confirmed. There are effects of carcass parts, equipment etc. The high pressure separated meat may well be of poorer quality but this is for other reasons than muscle structure degradation as referred to in the legislation. Other quality traits can and should be identified in the sample and in combination with the degradation level (to comply with legislation) this would give an overall assessment of the quality, which would be more accurate and more informative for end users as well as consumers. The comparison of the methods has been described in a manuscript submitted to Food Control (Raudsepp et al. 2016).2.3 Common EU histochemical method to assess muscle fibre structureThe consensus of the immuno-histochemical method as the most promising method to measure muscle fibre degradation was obtained in the consortium, and the method refined and then proposed for a CWA on “Comminuted and fragmented poultry meat – quantification of muscle fibre structure degradation”. The CWA was proposed by the MACSYS consortium with the secretariat assistance from NEN and announced at the CEN WEB page 27th of January, 2016. The kick off meeting held the 1st of March, and the final Workshop with an agreement of the CWA was held the 26th of May. The method is described in CWA 83.2.4 Automated histological image analysis systemThe original laminin methods analyses were performed by an image analysis system (Coolscope by NIKKON) sufficed with software developed for this particular purpose. As, however Coolscope is no longer commercially available and the company who developed the software no longer exists, there was a need for development of new and updated systems. Softcrits was responsible for this development in collaboration with AU. Coolscope is a combination of a microscope, a digitalized stage board and a camera and some standard software. The first analyses were carried out directly on the sections. The Coolscope system cannot handle a whole image thus the analysis was carried out on random selected areas within the section. These were features we would like to have included in the new system. This was initially in the MACSYS project achieved by taking pictures of random areas within each section and then send them for processing and analysis. Another alternative was to use a high resolution scanner and work on whole sections rather than selected areas. This was tested and with a positive outcome, however the price of these scanners is too high to become a standard system in most food labs. Late in the project, however, we were introduced to a scanner at a price compatible with that of a microscope, digitalized stage board and a camera. This scanner has proven its value and is definitely a solution we will support for work with the further validation of the method and its application in other animal species.The process of development was an iterative process where each individual version was examined. At the beginning of the project all staining methods were examined for feasibility for image analysis, implying that software to identify degraded muscle tissue was developed for all methods. The challenges with the standard staining methods (German and English) were much higher than with the immunological method (Danish). Thus efforts were put into the development of the analysis of this staining method. The focus on the development of the image analysis system was initially on standard samples and a correlation with results from the newly developed system to the reference samples (thigh: r= 0.85 breast: r=0.92) were reached almost similar to the one of the older system. Concerning measurements of the industrial samples more challenges was expected. So a further iterative process was initiated. We have continuously throughout the rest of the project period improved the algorithm and have raised the correlation to the older method substantially. There is still room for minor improvements, but we are confident that these can be done in the upcoming validation procedure. We have alongside development of the algorithm developed a cloud based handling procedure of the system which implies that the analyses can be carried out at the individual labs performing the sectioning and staining at a commercial base. The result of the image analyses is a percentage of structured muscle out of total muscle tissue in the sample. A business plan has been described for the cloud-based image analysis software.2.5 Fast-track method and system for wide in-line industrial application The development of a fast track prototype to measure muscle structure degradation in comminuted poultry meat began with a screening off a small set of samples with impedance as well as the spectroscopic techniques NIR, SRS-NIR, NIT and LIBS in order to find the best method to use for analysis of the MACSYS samples. Among those methods two spectroscopic techniques were shown to have the potential to be applied for either at-line or online monitoring a range of quality parameters in MSM. The first method, near-infrared spectroscopy (NIRS), is demonstrated to enable prediction of fat, protein and moisture and in addition, NIRS could be used to separate MSM samples with high degree of muscle degradation from samples with low degree of muscle degradation. NIRS is a well-known technique for online monitoring and therefore a good candidate for prototype development. However, not all quality traits were possible to cover with NIRS. Cholesterol and calcium content could not be predicted, probably due to low concentration of cholesterol and the fact that minerals cannot be measured directly with NIRS. For calcium prediction, the other technique, laser induced breakdown spectroscopy (LIBS), was applied. LIBS is an upcoming technique, which can be used for direct measurement of minerals in a sample and a advantage is that no sample preparation is required.The results for LIBS when applied to MSM show that this technique can be used to separate samples with very low calcium levels (regular minced meat and MSM samples with the lowest calcium levels) from other MSM samples. However, there is room for optimization of this method that may lead to better quantification of the calcium content. For this, a better understanding of how many measurements are optimal to achieve representative sampling and how much the detection of calcium is affected by sample composition would be interesting to study further. Even though LIBS has the potential for use directly on a product stream, an at-line solution is probably more feasible for this method. The LIBS method has been published in Schmidt Andersen et al. (2016).LIBS is novel, and can only give an estimate of a few inorganic compounds, the consortium decided that the NIR probe should be tested for further development. A description of mechanical design, optical system, software and data handling was described for the NIRS-based fast track prototype , including a protocol on how to use the prototype. The prototype was then tested by integration of the prototype into deboning machinery to be used for outgoing quality control of MSM raw material produced by machines supplied by Marel and Lima and also tested for the capability of the fast-track prototype to serve as an incoming quality control tool for users (e.g. RBD) of MSM raw material. The conclusions were clear: NIR is a method capable of estimating protein, moisture, fat and muscle degradation in chicken meat for both outgoing and incoming quality control of MSM raw material, and integration of the prototype into the deboning machinery seems reliable and stable. A business plan has been described for further exploration of the NIRS-based prototype.Potential Impact:The MACSYS project was funded as a research project for the benefit for SME’s. The outcome of the project for the SME’s is very much dependent of the future legislation on comminuted meat, however one step forward has been taken with MACSYS as an objective immune-histochemical method which can measure muscle structure degradation has been developed, and the consensus obtained for this method at the CEN workshop (CEN-CENLEC/WS 83). From here the envisaged socioeconomic impact for the MACSYS SMEs is: • The automated image analysis software is expected to be used by at least one laboratory per country/region in the 28 EU countries and an expansion of laboratories and number of images to be analysed over the first 5 years of commercialisation leading to a positive output in this time period. If the change of the EU regulation does not occur, the system will not be profitable and the project will not have commercial viability. It is therefore compulsory for the success of the product that the MSM regulations change and it is strongly necessary to lobby in the EU commission for this change to take place.• The installations of the fast-track device across Europe are expected to grow over a 10 year period which could make it an attractive investment however it is estimated that 3-4 years is required to fully mature the prototype. As for the image analysis software this scenario requires the change in the EU regulation of comminuted meat.3.1 Socio-economic impactThe vast majority of chicken meat is sold as cuts of breast, thigh and wings, but when these parts have been removed by machine, there is still meat left on the carcass. In fact, machines are often very inefficient butchers, often leaving some 50% (and sometimes much more) of the meat on the bone. Present legislation stipulates that if this meat is recovered by machine, the use is highly restricted - producers are not allowed to call it meat and if used it should be included in the list of ingredients as MSM. On the other hand if separation is performed manually it can be called meat and the product is not subjected to any restriction in use. The legislation anticipates that using a machine for the recovery process will cause the muscle fibre structure to be destroyed to an extent where it is no longer acceptable as meat. There is however, no method available by which the extent of degradation can be measured therefore all mechanically separated meat is declassified exclusively on the basis of the production method. The development of methods for separation and the option of changing settings of the machines, using lower pressure with different sizes of sieve holes has made it possible to make products which in an overall quality assessment is at the same level or even better than standard minced products and definitely better than hand separated meat. This has implied that large amounts of potentially high quality mechanically separated meat are downgraded simply because of legislation. This leads to high economical losses for the producers and food waste as large amounts of these products are processed into pet food. The MACSYS project has now made it possible to objectively determine the muscle structure degradation, and it is documented what is already known empirically that muscle tissue degradation depends on the type of muscle and machinery used and on the pressure used to separate meat from bones. The automated image analysis software and the NIRS-based fast-track device for objective quantification of the level of muscle structure degradation enables product differentiation and quality classification of mechanically deboned meat. This will have a positive influence on several poultry meat related markets (quality laboratories and regulatory authorities, deboning equipment manufacturers and producers and users of MSM). However, the positive impact will only occur if legislation changes and an economic incentive to perform such differentiation are created. With the establishment of the CWA the food quality laboratories and regulatory authorities is able to use a common standardised method for assessment of the level of muscle structure degradation. Following the CWA next step is to work for a change in the EU legislation and therefore it is crucial that policy makers actively promote the MACSYS results following the completion of the project as the socio-economic impact of MACSYS highly depend on the legislation.If legislation changes, it will enable producers of MSM to upgrade raw materials obtained using low pressure treatment with a low degree of muscle structure degradation to meat. For a small chicken slaughterhouse producing 1000 tons low pressure MSM per year it would increase income with approximately € 805.000 per year. For large producers of low pressure MSM it is obvious that income would increase even more. At present, the MSM raw material sales prize can get so low that the producers decide to sell to the pet food industry instead of human consumption. This constitutes a very important ethic problem as many people in the world are lacking protein and, therefore, it seems unreasonable and unethical to downgrade these products to pet food. Some of the MSM raw materials produced under high pressure, giving rise to high levels of muscle structure degradation, may be shifted to the low pressure group which would also result in increased income generation. It is hypothesised that a small poultry slaughterhouse producing 6000 tons of high pressure MSM per year would be able to transfer 2000 tons to low pressure treatment. This would result in an expected added value of around € 270.000 per year. An increased income through added value of certain products is also foreseen for users of mechanically deboned raw material. If legislation changes and mechanically deboned chicken meat using low pressure is recognised as meat, product development will become a new area of focus within this industry. New high quality products such as sausages and charcuterie chicken products e.g. chicken pepperoni, mortadella-like products and Fois Poulette (a poultry variety of the highly deemed Fois gras) are already requested by some consumer segments but not yet produced as the incentive to do so is lacking. If product development becomes an increased focus area new jobs would emerge and an estimated number of 2000 additional employees in Europe within the processing industry would be required. If legislation changes, it is likely that the number of producers of MSM would also increase and therefore equipment manufacturers would increase sales of deboning machines. Furthermore the results of the MACSYS project will generally enhance the sustainability of poultry meat production first and foremost within the EU, by substantially reducing waste of potentially acceptable poultry meat. In predictions of the market developments it is expected that EU will gradually loose its net export status. Because of the results of the MACSYS project the industry may now be provided with tools that will allow for an extreme focus on quality which eventually will enhance competitiveness of the industry, thereby reducing the demand for import of lower quality products from countries where in addition neither food safety nor animal welfare is of any importance. 3.2 Dissemination and exploitation activitiesA wide range of dissemination activities has been carried out during the time span of the project including the creation of a public MACSYS webpage (www.macsysproject.eu) with regular updating of new knowledge and results, production of a project video launched on YouTube, writing of scientific contributions accepted in peer-reviewed journals and popular articles published in trade-oriented magazines. In addition, seminars and workshops have been organised aiming at creating general awareness of MACSYS and disseminating the obtained results. The audience of these dissemination activities covers both higher education, industry and policy makers and span a wide range of stakeholders world-wide.Specifically the MACSYS project has generated new knowledge which has resulted in 2 peer-reviewed articles and 1 manuscript recently submitted. Numerous public dissemination activities have been carried out in the form of workshops, seminars and conference presentations. A CEN workshop was arranged with the purpose of making a CWA on the technical requirements and methods to classify objectively the level of degradation of muscle structure in MSM meat. Stakeholders and policymakers from 10 countries participated in the meeting. Among others AVEC (Association of Poultry Processors and Poultry Trade in the EU countries), NEPLUVI (Association of Dutch Poultry Processing Industries), Hungarian Poultry Product Board, the British Poultry Council and the Danish Agriculture and Food Council were represented at the kick-off meeting and subsequent workshop. Finally, the CWA was agreed upon on the 26th of May, 2016.List of Websites:http://macsysproject.euSME Partners:• Robert Damkjaer Ltd, Denmark (RBD)• Carometec A/S, Denmark (CMT)• LIMA, France• SoftCrits, Spain (SCS)RTD Partners:• Aarhus University, Denmark (AU)• University of Copenhagen (KU)• Leatherhead Research Limited, UK (LFR)• Max Rubner Institut, Germany (MRI)LE Partner/end-user:• Marel, Iceland.
