ارزیابی تازگی گوشت مرغ با استفاده از روش‌های ماشین بویایی و شبکه‌های عصبی مصنوعی

امروزه توجه ویژه بشر به کیفیت مواد غذایی موجب شده تا روش‌های سریع، آسان و غیرمخرب نظیر ماشین بویایی برای ارزیابی ویژگی‌های کیفی این مواد به کار گرفته شود. گوشت یکی از مهمترین مواد غذایی است و تازگی مهمترین ویژگی‌ کیفی آن به شمار می‌رود؛ بنابر این بررسی کیفیت آن برای مصرف کننده از ارزش بسزایی برخوردار است. هدف اصلی مطالعه حاضر بررسی امکان استفاده از روش‌های ماشین بویایی و شبکه‌های عصبی مصنوعی برای تشخیص تازگی گوشت مرغ در طول دوره نگهداری در یخچال بود. برای رسیدن به این هدف، قسمت‎های ران مرغ به عنوان نمونه‎های مورد مطالعه انتخاب و در دمای ˚c4 یخچال نگهداری شدند. در زمان‌های تعیین شده‎ای، نمونه‌ها پس از قرارگیری در محفظه‎های بسته‎ای از یخچال خارج و داده‌های بویایی آن‌ها اکتساب گردید. پس از پیش پردازش داده‌ها، از طبقه‌بند شبکه‌های عصبی مصنوعی با ساختار بهینه 3-6-10 برای طبقه-بندی و تشخیص تازگی نمونه ها استفاده شد. شاخص‌های آماری به کار رفته به منظور ارزیابی طبقه‌بند جهت تخمین تازگی گوشت مرغ شامل دقت، صحت، حساسیت، اختصاصی بودن و سطح زیر منحنی بودند. مقادیر این شاخص‌ها برای طبقه‌بندی با ویژگی‌های منتخب به ترتیب برابر 95.77، 94.7، 92.18، 95.95 و 94.1 درصد محاسبه گردیدند. نتایج قابل قبول به دست آمده از بررسی حاضر به وضوح نشان داد که سامانه پیشنهادی بکار رفته به عنوان یک روش هوشمند و قابل اعتماد توانایی طبقه‌بندی بلادرنگ تازگی گوشت مرغ به صورت سریع، آسان، اقتصادی، غیر مخرب و با دقت مناسب را دارد.

evaluation of chicken meat freshness using olfaction machine and artificial neural networks

introduction: today, with the increasing awareness and concern for food quality, there has been a significant development in the use of fast, easy, and non-destructive methods to assess the quality of foodstuffs. in particular, the use of olfaction machine and artificial intelligence has gained prominence in this field. among the various food items, meat holds a special importance as a primary source of nutrition for humans. the freshness of meat is considered to be the most crucial qualitative feature, and it plays a vital role in ensuring the health and well-being of individuals. in this regard, chicken meat has emerged as a popular choice due to its nutritional value and widespread consumption worldwide. therefore, it becomes imperative to devise effective methods for checking the quality and freshness of chicken meat to ensure its suitability for consumption.materials and methods: the main objective of the present study is to explore the feasibility of employing an electronic nose and artificial neural network methods for detecting the freshness of chicken meat during its storage in a refrigerator at a temperature of 4 ºc. the electronic nose, a novel technology in the area of food quality assessment, utilizes an array of sensors to mimic the olfactory system of humans and detect various volatile compounds released by the meat. these sensors generate a unique pattern of responses, which can be analyzed to determine the freshness of the meat. the artificial neural network, on the other hand, is a computational model inspired by the functioning of the human brain. it has the capability to learn and make predictions based on the patterns it identifies in the input data. in the neural network system employed in this study, the input layer consists of 10 neurons, corresponding to the number of sensors in the electronic nose, while the output layer comprises 3 neurons representing different freshness classes of chicken meat. to optimize the performance of the network, various classifier networks were designed and evaluated. after careful examination of different network structures, it was determined that the best structure consisted of a hidden layer with 6 neurons. consequently, an optimal network with a general structure of 10-6-3 was created to detect the freshness of chicken meat during different days of storage.results and discussion: to assess the performance of the classifier in evaluating the freshness of chicken meat, several statistical indices were employed. these included precision, accuracy, sensitivity, specificity, and area under the curve factors. the obtained values of these indices for the classification using selected characteristics were found to be 95.77, 94.7, 92.18, 95.95, and 94.1, respectively. these results indicate that the developed intelligent diagnosis system based on the electronic nose and artificial neural networks is capable of online classification of chicken meat with high accuracy and reliability. the successful implementation of the proposed system has significant implications for the food industry. it offers a fast, easy, economical, and non-destructive method for assessing the freshness of chicken meat. this can help in reducing food waste by ensuring that only fresh and safe meat is made available to consumers. additionally, it can aid in improving the overall quality control processes in the food supply chain, thereby enhancing consumer satisfaction and trust.conclusions: in conclusion, the present study demonstrates the potential of using an electronic nose system and artificial neural networks for detecting the freshness of chicken meat during storage. the obtained results highlight the effectiveness of this approach in providing accurate and reliable classification of chicken meat based on its freshness. this research contributes to the advancement of food quality assessment methods and offers valuable insights for the food industry.