Harnessing the Electronic Nose for Early Detection of Diseases and Quality Control in Fresh Postharvest Produce: An Exhaustive Analysis

by Jorge Luis Alonso with ChatGPT The journal Comprehensive Reviews has published a research paper that aims to review various research methodologies used in electronic nose technology, with a specific focus on early disease detection and quality monitoring of fresh produce postharvest. This is a summary of the paper. 1. Introduction The UN Sustainable Development Goals aim to reduce global food waste by 50% by 2030. However, total elimination may not be feasible due to postharvest losses caused by pathogen infection and other factors. Electronic noses (EN) are increasingly being used to detect changes in the aroma of fruits and vegetables. This is important because it can act as a biomarker for the safety and quality of the produce. Early detection of postharvest diseases using EN can minimize food and economic losses. EN has also been applied in various other sectors to meet consumer expectations for safety and quality, such as healthcare, pharmaceuticals, forestry, environmental protection, dairy and food industries. Several review articles have been published on the mechanism and application of EN sensors. They have been used for fruit identification, maturity and quality assessment, food safety evaluation, and early disease detection in crops. However, publications specifically reviewing studies on early disease detection in harvested produce are rare. EN has also been investigated for monitoring produce quality throughout the supply chain.? The purpose of this paper is to review the different research methodologies of EN, including volatile metabolite profiling, biomarker and quality assessment, limitations and future directions, applied to early disease detection and postharvest produce quality monitoring.? The review also recognizes the need to reduce postharvest losses and provides guidance to researchers investigating this application in unstudied produce types or other food products. 2. Fundamental principles of?EN Sensory array materials The EN system uses a sensory array setup that mimics the mammalian nose to detect changes in aroma from volatile metabolites. This setup typically includes multiple sensors, each sensitive to specific compounds, to detect different volatile metabolites. Sealed vials hold the samples, allowing the volatiles to interact with the sensors and change their physical and chemical properties.? EN has been explored for detecting changes in volatile metabolites at various stages of crop maturity, disease identification, quality assessment and shelf-life evaluation, ultimately creating a real-time monitoring system. Compared to other detection tools, EN offers a simpler, more portable, cost-effective and efficient field implementation.? Several sensor types are available, including? Metal-Oxide-Semiconductor (MOS) Metal Oxide Semiconductor Field Effect Transistor (MOSFET) Piezoelectric crystal (PC) Quartz crystal microbalance (QCM) Surface Acoustic Wave (SAW) The choice of sensor has a significant impact on the analysis result. Developing these technologies is critical to improving crop production and ensuring safety.? The three main steps in characterizing volatile metabolite profiles include odor sampling, disease detection through sensory array analysis, and data acquisition and pattern recognition. Odor sampling In odor analysis, volatile metabolites are attracted to a sensory array using appropriate sampling methods. Due to the nature of fresh produce and consumer preferences, rapid sampling with minimal mechanical damage is preferred. Common methods include static headspace, preconcentrator, and solid phase microextraction (SPME).? The static headspace method uses sealed vials that allow samples to reach equilibrium prior to detection. Parameters such as temperature, sample size, and equilibration time must be optimized.? The preconcentrator, also known as the ?purge and trap? method, uses an inert gas stream to direct volatile metabolites to the detector. However, this technique can result in volatile metabolite depletion during purging.? SPME has gained popularity due to its ease of use and superior detection rates. It captures, equilibrates, and thermally desorbs volatile compounds on a specific fiber for detection. This effective extraction method is highly recommended for fresh produce odorant sampling. Disease detection and quality characterization Plant diseases are divided into non-infectious and infectious types. Non-infectious diseases are caused by genetic factors, environmental stress, or nutrient deficiencies and don?t spread to other plants. Conversely, infectious diseases are caused by pathogens such as fungi, bacteria or viruses. Early detection of infectious diseases is critical to controlling or preventing their spread.? Volatile metabolites produced by plants and fresh produce provide bioinformation for rapid, non-invasive disease diagnosis. In particular, several volatile metabolites produced during pathogenesis can reveal crop quality attributes. Fungal pathogens exhibit species-specific metabolite characteristics that can be identified using techniques such as GC-MS and FTIR. For example, Aspergillus candidus produces a unique volatile compound (monoterpene) that differs from other Aspergillus species. Data acquisition and pattern recognition In EN-based disease detection and quality characterization, post-processing steps analyze signals to produce final results. These signals require pattern recognition analysis, which can be supervised or unsupervised, and parametric or non-parametric.? Statistical models such as Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Partial Least Square Analysis (PLS), and Artificial Neural Network (ANN) help process EN system data for improved signal representation and accurate results.? Different pattern recognition components have different applications depending on the study. For example, PCA works well with unknown samples or hidden correlations, while LDA, a supervised method, maintains class-specific data. Similarly, PLS manages collinear data and minimizes the number of calibration samples required, and ANN is suitable for analyzing large volatile concentrations and provides flexibility. SVM is a supervised machine learning technique applicable to classification or regression problems.? Preprocessing standardizes odor data, removes unnecessary information, and prepares data for further analysis. 3. EN for early detection of postharvest diseases Postharvest diseases resulting from latent and senescent infections by pathogens negatively affect the quality of fresh produce and lead to economic losses. Although molecular techniques such as PCR and serological methods have proven effective, they have limitations such as invasiveness, complexity and lack of rapid monitoring.? EN is a non-destructive and cost-effective alternative that can detect volatile metabolites released by fresh produce. EN can differentiate between healthy and infected samples and can be used to identify specific biomarkers for different fungal species.? Research on strawberries and peaches has successfully identified unique volatile metabolites associated with decay due to fungal infection. This distinctive volatiles could serve as biomarkers for the presence of the fungi being tested. EN can be used to distinguish and monitor the growth of common fungi that infect peaches and to investigate the detection of other postharvest diseases. Numerous studies have investigated the use of EN to detect fungal contamination and disease in a variety of food products. These studies have reported high predictive accuracy for identifying contamination levels in peanuts and apples, and for classifying healthy and infected apples and garlic. However, certain sensors were ineffective at detecting volatile changes, requiring complementary analysis for more detailed determinations. Several pattern recognition systems were used to streamline the analysis and improve accuracy. 4. EN for postharvest quality monitoring Traditional methods for evaluating the quality of fresh produce, such as titration and penetrometer, are inaccurate and can destroy samples. Although aroma and flavor are important sensory cues for customers, evaluating them with human evaluators can be costly and time-consuming. However, modern standards have made it possible to perform rapid and reliable identification for postharvest food quality monitoring using EN. This approach can detect changes in volatile profiles to analyze and predict the various quality indices and maturity stages of produce. Freshness quality Several studies have used EN technology to monitor the freshness of various fruits and vegetables. This technology is based on MOS sensors that can detect changes in volatile compounds that may indicate the quality of spoilage. Statistical analyses such as PCA, PLS, and CDA have been used to distinguish the differences in volatile compounds over different storage periods. In addition, some studies compared EN results with traditional quality characteristics or human sensory evaluation. The results of these studies demonstrate that EN technology can provide portable and cost-effective quality monitoring of fruits and vegetables during storage. Ripening quality Studies have investigated the use of electronic devices to monitor fruit quality based on its ripening stage. In one study, a portable electronic detector using gas chromatography was used to analyze the volatile compounds produced by mangoes at various stages of ripeness. The system predicted unripe and ripe mango stages with highly concentrated compounds, but could not access data from fully ripe mangoes due to the crop profile.? Another study used Gas chromatography?mass spectrometry (GC-MS) and an EN with 10 MOS sensors to classify mulberries into five maturity levels. The authors classified the different maturity classes of the berries using PCA, LDA, and ANN algorithms, with ANN providing the best classification results.? Finally, an EN equipped with 13 gas sensors was developed to discriminate between unripe, ripe, and rotten stages of banana ripeness with 100% accuracy using the BPNN approach. The EN setup also displayed the final ripeness stage determination on a website. Sensory quality A research study was conducted to differentiate between hulled and unhulled Triticum sp. wheatgrass species based on flavor quality and volatile metabolite characteristics. To accomplish this, the study established a protocol using a portable EN instrument with MOS sensors and Headspace-solid phase microextraction (HS-SPME) analysis. The results of the study showed that EN sensors could differentiate common volatile compounds such as alcohols, ketones, and aldehydes, regardless of genotype. The study found that samples with higher levels of certain volatiles had higher levels of TSS. While the volatiles comparison was useful in distinguishing between peeled and unpeeled samples, it was not effective in identifying genotypic differences. PCA analysis confirmed that the genotypic differences had a low total variance of 61.8%. Other quality parameters A study was conducted on the use of QCM biosensors coupled with HS-SPME GC-MS analysis to profile carrot aroma. The study monitored changes in aroma compounds for up to 26 days at different storage temperatures and identified 18 compounds. The EN successfully tracked changes in carrot aroma, but there was no correlation between data obtained from the EN via QCM and GC-MS.? In another study, a homemade low-cost EN consisting of 8 MOS sensors and near-infrared spectroscopy (NIR) spectroscopy was used to study quality changes in pitaya. The study found that EN sensors sensitive to alcohol showed the largest peaks, and the developed EN technology had a high degree of accuracy in classifying samples on different storage days. However, the sensors were unable to capture enough volatiles from the thick peel of pitaya in the measurements performed on a non-broken fruit.? Future studies should focus on developing low-cost EN technologies that are more likely to be adopted by varieties and useful in the field. 5. Future trend, limitations, and conclusion Recent reports suggest that EN technology can serve as a powerful tool for early detection of postharvest diseases and real-time, non-destructive quality assessment of produce. However, to realize its full potential, researchers must take steps to improve and expand its use. One critical action that researchers can take is to develop and incorporate sensor materials that can detect multiple responses for a more efficient early detection process. In addition, the EN mechanism can be integrated with automated agricultural equipment for real-time disease detection. Another important step researchers can take is to enhance the EN structure by integrating sensor materials with data processing tools, such as phase-based reconstruction (PSR) and radial basis function (RBF) models. This step can create a more intelligent olfactory system while reducing noise and disturbances in the raw data matrix to improve the stability and applicability of the system. In addition, EN should be integrated with sensory analysis to provide a comprehensive analysis of product quality. To avoid subjective and biased results from individual panelists, predefined quality classes need to be established. In addition, researchers should work to develop EN sensors capable of acquiring data based on the concentration and severity of each disease characterized, coupled with other analyses such as GC-MS or molecular techniques. To make EN more practical and economically viable for on-farm use, researchers should focus on mobile and in-field detection, particularly for horticultural crops. This technology has enormous potential to contribute to achieving the United Nations Sustainable Development Goals (UN SDGs) of zero hunger (SDG 1), zero poverty (SDG 2), and sustainable production and consumption (SDG 12). Author:?Jorge Luis?Alonso G. (with ChatGPT)?Maximizing Agribusiness Profits with Expert Postharvest Storage Strategies | Horticultural Writing Specialist.?This article was written exclusively for the business platform Postharvest. Image by Couleur from Pixabay