IVD technology丨"G-Sense" VOC detection products
If there is a POCT product that can use the volatile compounds emitted by our body as indicators of specific diseases, then IVD testing will usher in a big change. In this way, many serious and degenerative diseases can be detected long before current standard methods. The author proposes a concept product called "G-Sense" for detecting volatile organic compounds in exhaled breath.
An IVD product that utilizes volatile compounds emitted by our body as indicators of specific diseases is an exhaled volatile organic compound (VOC)-based detection method for rapid, real-time identification of various pathogens. VOCs are derived from many endogenous biochemical processes in the human body, including lipid oxidation, and carbohydrate and fatty acid metabolism. Since cellular metabolism is altered by disease, changes in VOCs can serve as biomarkers for specific pathophysiological conditions. At present, the detection methods of exhaled VOC mainly include mass spectrometry, chromatography and other large-scale analytical instruments and gas sensors. Compared with large-scale instruments, gas sensors are small in size, low in cost, and easy to operate, and have good application prospects in the detection of VOC in the exhaled breath of large-scale people. The types of gas sensors include semiconductor type, electrochemical type, mass type, optical type, gas chromatography type, contact combustion type, etc.
01
Product name: G-Sense (graphene sensing)
Product introduction: G-Sense is a graphene electronic olfactory sensor based on machine learning, which can detect and identify VOCs in your breath for various applications such as disease diagnosis, air quality assessment, and food quality inspection. It has the characteristics of high sensitivity, high selectivity, low power consumption, and low cost, while providing a non-invasive, convenient, and comfortable detection experience. G-Sense makes your breath not only a sign of life, but also an indicator of health, making your breath more valuable
Product purpose: Detect and identify different VOC odors, used in disease diagnosis, environmental monitoring, public safety, food quality and other fields
Product principle: Use graphene nanoribbon arrays and gold nanoparticles modified electronic olfactory sensors to respond to VOC odors, and use machine learning algorithms to extract response features and classify and identify them
Product composition: sensor chip, signal acquisition module, data processing module, display output module
Product specifications: The sensor chip size is 10mm×10mm, the signal acquisition module is a voltage conversion circuit, the data processing module is an embedded microcontroller, and the display output module is a liquid crystal display
Product performance: The detection threshold of the sensor chip for various VOC odors is between 0.1 ppm and 1 ppm, the accuracy of odor discrimination reaches 100%, and the accuracy of odor identification reaches 95%; the response characteristics of the binary odor mixture are similar to those of a single odor Correlation, which can reflect the ratio of the mixture; the sensor chip has good stability and repeatability, the response time is within 10 seconds, and the recovery time is within 30 seconds
Product advantages: Compared with other types of IVD testing products, this product has the following advantages:
High response sensitivity, able to detect low concentration of VOC odor
High response specificity, capable of distinguishing and identifying different VOC odors
Fast response, capable of real-time monitoring of VOC odor changes
Data processing intelligence, able to use machine learning algorithms to improve classification and recognition performance
The equipment is small in size, easy to carry and use
Product Market: According to market research, this product is mainly aimed at customers in the following fields:
In the medical field, it can be used for rapid, non-invasive, and convenient disease screening and diagnosis, such as diabetes, lung cancer, and new coronary pneumonia;
In the environmental field, this product can be used to monitor the VOC odor in environmental factors such as air quality, water quality, soil, etc., to discover pollution sources and harmful substances in time, and to protect the environment and human health. Such as formaldehyde, benzene, carbon dioxide, etc.;
In the field of food, this product can be used to detect VOC odor in food, beverage, agricultural products and other food industries, evaluate the freshness, composition, nutritional value of food, etc., to ensure the quality and safety of food. Such as fruit, meat, dairy products, etc.;
Public safety: This product can be used to detect the VOC smell of dangerous items such as explosives, drugs, and toxic gases, and improve the safety of public places and vehicles. Such as explosives, drugs, gunpowder, etc.
Product competition: According to market analysis, the main competitors of this product are as follows:
AlphaSense: This is an American company that focuses on the development of electronic olfactory sensors based on metal oxide nanowire arrays. Its products can detect a variety of VOC odors and have high sensitivity and stability.
Aryballe: This is a French company that focuses on the development of electronic olfactory sensors based on biomolecular recognition technology. Its products can detect a variety of VOC odors with high specificity and reliability.
Koniku: This is an American company that focuses on the development of electronic olfactory sensors based on synthetic biology technology. Its products can detect a variety of VOC odors and have high intelligence and scalability.
Product advantages: Compared with competitors, this product has the following advantages:
The response features are rich, and machine learning algorithms can be used to extract multiple dynamic features as labeling information to improve classification and recognition performance.
The response mechanism is clear, and molecular dynamics simulation and density functional theory calculations can be used to reveal the competitive adsorption mechanism between odor molecules and sensing materials, improving the interpretation of the response.
The response simulates the human sense of smell, which can imitate the Sniffin' Sticks test to evaluate the sensor's olfactory performance, and reflects the masking effect in the binary odor mixture, improving the realism of the response.
02
source of concept
The concept of G-Sense comes from an artificial olfactory research paper titled "Machine learning-enabled graphene-based electronic olfaction sensors and their olfactory performance assessment", the authors of which are Shirong Huang, Alexander Croy, Antonie Louise Bierling, Vyacheslav Khavrus, Luis Antonio Panes-Ruiz, Arezoo Dianat, Bergoi Ibarlucea and Gianaurelio Cuniberti, published in Applied Physics Reviews on May 15, 2023
DOI: https://doi.org/10.1063/5.0132177. This article mainly introduces the development of graphene electronic olfactory sensors based on machine learning and the method for evaluating the olfactory performance of volatile organic compounds (VOCs).
picture
Image from this paper
The background of the article points out that olfaction is an evolutionarily ancient sensory system that can provide complex information about the chemical environment. Detection and identification of VOCs produced by natural and artificial processes can serve as signature fingerprints to help determine their origin. Inspired by the biological sense of smell, artificial olfaction aims to achieve similar performance, thereby digitizing the sense of smell. The effectiveness of artificial olfaction depends on the sensor hardware on the one hand and signal processing on the other. The article aims to demonstrate the development of a machine learning-based graphene electronic olfactory sensor and propose a method to evaluate its olfactory performance against VOCs. The author used an electronic olfactory sensor based on a graphene nanoribbon array and gold nanoparticles modification, extracted 11 dynamic features as fingerprint information using a machine learning algorithm, and imitated the Sniffin' Sticks test to evaluate the sensor's response to four VOC odors (clove , eucalyptus, lemon and rose aromas), detection threshold, odor discrimination and recognition performance. At the same time, the authors also performed experiments on binary odor mixtures and revealed the competitive adsorption mechanism between odor molecules and sensing materials using molecular dynamics simulations and density functional theory calculations. The results show that the electronic olfactory sensor has high olfactory performance for four VOC odorants, with a detection threshold between 0.1 ppm and 1 ppm, an odor discrimination accuracy of 100%, and an odor identification accuracy of 95%. In experiments with binary odor mixtures, the sensor's response characteristics were more similar to a single odor and related to its ratio, similar to the masking effect in human olfactory perception. The simulation and calculation results showed that different odor molecules have different adsorption sites and adsorption capacities on the graphene surface, which may affect the sensor's response to the mixture. In this paper, we propose a machine learning-based graphene electronic olfactory sensor and propose a method to evaluate its olfactory performance on VOCs. This method can simulate the human olfactory system, and reveal the sensor's response rules and mechanisms to single and mixed odors. This may promote the application of electronic olfactory sensors in some emerging fields, such as environmental monitoring or public safety.
The experimental method in the article mainly includes the use of electronic olfactory sensors based on graphene nanoribbon arrays and gold nanoparticles modification, using machine learning algorithms to extract 11 dynamic features as fingerprint information; imitating the Sniffin' Sticks test to evaluate the sensor for four VOCs Detection Threshold, Odor Discrimination and Recognition Performance of Odors (Clove, Eucalyptus, Lemon, and Rose Aromas); Experiments on Binary Odor Mixtures, Odor Molecules and Sensing Materials Revealed Using Molecular Dynamics Simulations and Density Functional Theory Calculations competitive adsorption mechanism.
03
Discussion analysis
The author summarizes the key steps:
●The first step is to prepare the graphene sensing array. The graphene nanotubes can be grown on the metal catalyst by chemical vapor deposition and transferred to the flexible substrate, and then the graphene nanotubes can be processed with different functional agents Surface modification to enhance its response to different VOCs;
The second step is to build a signal acquisition and processing module, which can use a microcontroller unit (MCU) as the core chip, connect with the graphene sensor array through a resistance measurement circuit, and communicate with external devices through a Bluetooth or Wi-Fi module, At the same time, corresponding software programs are written to realize the reading, processing and transmission of signals;
The third step is to train the machine learning model. Support vector machine (SVM) can be used as a classification algorithm to construct training and test sets by collecting data of different VOCs and respiratory gases of lung cancer patients, and extracting dynamic features as label information , and carry out cross-validation and parameter optimization to improve the accuracy and generalization ability of the model;
●The fourth step is to design the display and interaction module. Liquid crystal display (LCD) can be used as the display device, and a touch screen or buttons can be used as the interaction mode. At the same time, a simple, friendly and beautiful user interface is designed to display the test results and suggestions. And provide relevant operation guidance and feedback;
●The fifth step is to conduct product testing and evaluation. The Sniffin' Sticks test can be used as a reference standard to test and evaluate the olfactory performance of G-Sense, including odor detection threshold, odor discrimination and recognition performance, and compare with other traditional lung cancer Test methods are compared and analyzed to verify the effectiveness and advantages of the product.
Possible risks or challenges:
There may be some technical difficulties in the preparation process of graphene sensing arrays, such as the uniformity, stability and repeatability of graphene nanotubes, and the selection and control of functionalizing agents;
●The design of the signal acquisition and processing module may have some performance problems, such as circuit noise, interference and anti-interference ability, as well as software compatibility, security and updateability;
●There may be some data problems in the training of machine learning models, such as the quality, quantity and representativeness of data, as well as the labeling, cleaning and analysis of data;
●The design of display and interaction modules may have some user problems, such as user needs, preferences and habits, as well as user satisfaction, trust and loyalty;
●Product testing and evaluation may have some standard issues, such as testing methods, conditions and indicators, as well as evaluation criteria, basis and results.
Risk and challenge response strategies:
●For the preparation process of graphene sensor array, you can refer to existing literature and patents, learn from successful experience and methods, and carry out some experimental optimization and parameter adjustment at the same time to improve the quality and functional effect of graphene nanotubes;
●For the design of signal acquisition and processing modules, mature circuit schemes and software platforms can be used to conduct some simulation tests and debugging to reduce circuit noise and interference and improve software compatibility and security;
●For the training of machine learning models, high-quality data can be obtained from public databases or cooperative hospitals, and some data preprocessing and feature engineering can be performed to increase the validity and representativeness of data;
●For the design of display and interaction modules, some user research and feedback can be conducted to understand user needs and preferences, and some user experience testing and improvement can be carried out to improve user satisfaction and trust;
●For product testing and evaluation, you can refer to relevant international or domestic standards and norms, select appropriate testing methods and indicators, and conduct some controlled experiments and statistical analysis to ensure the validity and objectivity of testing.
