Periodic Reporting for period 4 - SCENT (SCENT: Hybrid Gels for Rapid Microbial Detection)
Berichtszeitraum: 2020-06-01 bis 2022-05-31
The possibility to mimic nature using artificial intelligent systems to perform complex tasks is increasingly relevant. Artificial olfaction, for example, is being implemented in key societal areas namely for early and non-invasive disease diagnostics.
These sensing systems use electronic nose devices (e-noses). E-noses include an array of gas sensors associated with signal-processing tools. Chemical sensors are the most common and usually present limited stability and selectivity, requiring aggressive conditions during processing and operation. Bioinspired sensors employing biological olfactory receptors are an alternative. Unfortunately, their complexity and low stability are a limitation. The SCENT team focused on an innovative class of stimulus-responsive gels, which tackle these key challenges. The gels are customisable and have a low environmental footprint associated. The SCENT project explored the potential of these new materials to advance the field of odour detection, while providing new tools for the scientific community. To accomplish this, the SCENT team: 1) built libraries of gels with semi-selective and selective properties, 2) assembled tailor-made opto-electrical e-noses for analysis of volatile compounds, 3) developed customized algorithms for signal analysis, processing and classification, 4) assessed e-nose applications namely in identifying pathogenic bacteria, including those with acquired antimicrobial-resistances.
These sensing systems use electronic nose devices (e-noses). E-noses include an array of gas sensors associated with signal-processing tools. Chemical sensors are the most common and usually present limited stability and selectivity, requiring aggressive conditions during processing and operation. Bioinspired sensors employing biological olfactory receptors are an alternative. Unfortunately, their complexity and low stability are a limitation. The SCENT team focused on an innovative class of stimulus-responsive gels, which tackle these key challenges. The gels are customisable and have a low environmental footprint associated. The SCENT project explored the potential of these new materials to advance the field of odour detection, while providing new tools for the scientific community. To accomplish this, the SCENT team: 1) built libraries of gels with semi-selective and selective properties, 2) assembled tailor-made opto-electrical e-noses for analysis of volatile compounds, 3) developed customized algorithms for signal analysis, processing and classification, 4) assessed e-nose applications namely in identifying pathogenic bacteria, including those with acquired antimicrobial-resistances.
Gels that smell?
Can you imagine a gelatine that can distinguish different smells? SCENT team made this vision come true.
We developed gas-sensitive materials with a unique combination of biological and chemical components, which self-assemble to form gels. The new gels mimic the biological olfactory system but are much simpler in composition and robust in their design. They change their optical and electrical properties upon contact with volatile organic compounds, and these changes can be converted into signals using tailor-made electronic noses (e-noses) developed during the project. The generated signals are further processed and analyzed with artificial intelligence tools. With this approach, it is possible to accurately predict the nature of a smell previously learnt by the e-nose system, showing the potential for discrimination of distinct odors.
There are several potential applications for the SCENT technology, namely for the quantification of ethanol in automotive fuel, and monitoring of fish deterioration due to microbial action. In the clinical area, the team explored applications in the non-invasive diagnostics field, namely, to distinguish between antimicrobial resistant and sensitive bacteria, or for cancer monitoring.
The SCENT project was a unique opportunity to train over 23 young researchers, between Master, PhD students and young postdoctoral fellows. The team was also very active in dissemination, communication and exploitation activities, having received 13 awards/recognitions. We published 15 scientific papers; presented our work as 6 plenary/keynote lectures, 46 oral communications and 23 posters at international conferences. The IP technology developed during the SCENT project was protected in 2 European patent applications (1 granted) and the team participated in tech transfer courses and contest, having won a 1st prize. Considering the importance to divulge our work to the general public, we had several participations in Outreach events, media appearances and one dedicated video (https://www.youtube.com/watch?v=7m6uqpB9vzU ).
Can you imagine a gelatine that can distinguish different smells? SCENT team made this vision come true.
We developed gas-sensitive materials with a unique combination of biological and chemical components, which self-assemble to form gels. The new gels mimic the biological olfactory system but are much simpler in composition and robust in their design. They change their optical and electrical properties upon contact with volatile organic compounds, and these changes can be converted into signals using tailor-made electronic noses (e-noses) developed during the project. The generated signals are further processed and analyzed with artificial intelligence tools. With this approach, it is possible to accurately predict the nature of a smell previously learnt by the e-nose system, showing the potential for discrimination of distinct odors.
There are several potential applications for the SCENT technology, namely for the quantification of ethanol in automotive fuel, and monitoring of fish deterioration due to microbial action. In the clinical area, the team explored applications in the non-invasive diagnostics field, namely, to distinguish between antimicrobial resistant and sensitive bacteria, or for cancer monitoring.
The SCENT project was a unique opportunity to train over 23 young researchers, between Master, PhD students and young postdoctoral fellows. The team was also very active in dissemination, communication and exploitation activities, having received 13 awards/recognitions. We published 15 scientific papers; presented our work as 6 plenary/keynote lectures, 46 oral communications and 23 posters at international conferences. The IP technology developed during the SCENT project was protected in 2 European patent applications (1 granted) and the team participated in tech transfer courses and contest, having won a 1st prize. Considering the importance to divulge our work to the general public, we had several participations in Outreach events, media appearances and one dedicated video (https://www.youtube.com/watch?v=7m6uqpB9vzU ).
The innovative concept of hybrid gels in artificial olfaction was introduced in the SCENT project. It has been patented (EP3256545A1, Granted) and then published in high-impact peer-reviewed journals in the field of materials science (Adv.Funct.Mat. 2017, 27, 1700803; Materials Today Bio. 2019, 1, 100002). It was also spotted by the media, resulting in one press release and 1-full page in a daily national newspaper. These first publications placed our research and SCENT in the spotlight of society, academia and industry.
The work published in Adv. Mater. 2022, 34, 2107205 was an important milestone of the project. It was also spotted by the media, resulting in one press release and 2-full page in a daily national newspaper. The relevance of this work arises from our solution to the problem of humidity interference in gas sensing. We obtained either humidity sensors or humidity-tolerant VOC sensors that do not require sample pre-conditioning or further signal processing. These results were extremely important as they formed the basis to apply SCENT technology in humid environments as those envisioned for clinical diagnostics.
Another important progress was the use, for the first time, of deep convolutional neural networks (CNN) as pattern recognition systems to analyse dynamic optical textures in LC droplets exposed to a set of different VOCs. With our classification models, we showed that a single individual droplet can recognise 11 VOCs with small structural and functional differences. The optical texture variation pattern of a droplet also reflects VOC concentration changes (Sensors 2021, 21(8), 2854).
The work published in Adv. Mater. 2022, 34, 2107205 was an important milestone of the project. It was also spotted by the media, resulting in one press release and 2-full page in a daily national newspaper. The relevance of this work arises from our solution to the problem of humidity interference in gas sensing. We obtained either humidity sensors or humidity-tolerant VOC sensors that do not require sample pre-conditioning or further signal processing. These results were extremely important as they formed the basis to apply SCENT technology in humid environments as those envisioned for clinical diagnostics.
Another important progress was the use, for the first time, of deep convolutional neural networks (CNN) as pattern recognition systems to analyse dynamic optical textures in LC droplets exposed to a set of different VOCs. With our classification models, we showed that a single individual droplet can recognise 11 VOCs with small structural and functional differences. The optical texture variation pattern of a droplet also reflects VOC concentration changes (Sensors 2021, 21(8), 2854).