The annual global production of chickens exceeds 25 billion birds, which are often housed in very large groups, numbering thousands. Distress calling triggered by various sources of stress has been suggested as an 'iceberg indicator' of chicken welfare. However, to date, the identification of distress calls largely relies on manual annotation, which is very labour-intensive and time-consuming. Thus, a novel convolutional neural network-based model, light-VGG11, was developed to automatically identify chicken distress calls using recordings (3363 distress calls and 1973 natural barn sounds) collected on an intensive farm. The light-VGG11 was modified from VGG11 with significantly fewer parameters (9.3 million versus 128 million) and 55.88% faster detection speed while displaying comparable performance, i.e. precision (94.58%), recall (94.89%), F1-score (94.73%) and accuracy (95.07%), therefore more useful for model deployment in practice. To additionally improve light-VGG11's performance, we investigated the impacts of different data augmentation techniques (i.e. time masking, frequency masking, mixed spectrograms of the same class and Gaussian noise) and found that they could improve distress calls detection by up to 1.52%. Our distress call detection demonstration on continuous audio recordings, shows the potential for developing technologies to monitor the output of this call type in large, commercial chicken flocks.

Rapid identification of infectious disease pathogens such as foot-and-mouth disease virus (FMDV) during new outbreaks of disease is of fundamental importance in disease control. SpectroSensTM optical microchip sensors demonstrating rapid, label-free detection of FMDV are presented; these contain multiple high-precision planar Bragg gratings and function as low-cost, robust refractive-index sensors. Sensor selectivity to FMDV is imparted by functionalising the top-surface of specific sensing channels with anti-FMDV monoclonal antibodies (mAbs). Selective binding of cognate antigens within the test sample to surface-immobilised FMDV mAbs results in localised changes in refractive index within specific sensing channels; these antibody-antigen interactions manifest as increases in wavelength of light reflected from the multi-channel sensor chip (light is coupled into and out of the chip via optical fibres). Selective identification of FMDV within minutes of sample introduction has been demonstrated by referenced measurement of changes in sensor reflected wavelength from anti-FMDV channels against sensor controls; simplified ‘snap-shot’ assay data are displayed in the form of a simple yes/no readout using a robust, hand-portable device, with further semi-quantitative information available to the ‘super-user’. The characteristics of the SpectroSensTM multiplexed detection platform highlight its potential for in-field detection of foot-and-mouth disease and prospective expansion into diagnoses of other infectious veterinary diseases.

An optical waveguide array biosensor suitable for rapid detection of multiple bio-hazardous agents is presented. SpectroSens™ optical microchip sensors contain multiple spatially-separated waveguide channels with integral high-precision Bragg gratings sensitive to changes in refractive-index; selective surface-functionalisation of discrete sensing channels with different antibodies as bio-recognition elements enables selective multi-analyte biological detection. Interactions between target antigens in the test sample and respective surface-immobilised antibodies result in localised changes in refractive-index; the biosensor response manifests as increases in wavelength of light reflected from specific sensing channels. Multiplexed, label-free detection of 8 different biological agents, encompassing bacterial spores, vegetative cells, viruses and proteinaceous toxins has been demonstrated in real-time. Selective detection of Bacillus atrophaeus (BG) spores, Escherichia coli cells, MS2 viruses and ovalbumin (OVA) protein (simulant bio-hazardous agents) was first demonstrated as proof-of-concept; subsequently, detection of Bacillus anthracis (BA) spores (UM23CL2 strain), Franciscella tularensis (FT) cells (live vaccine strain), Vaccinia viruses (heat-killed) and ricin toxin (bio-hazardous agents) was proven. Two optical microchip sensors, each comprising 8 sensing channels were packaged into a single disposable cartridge allowing simultaneous 16-channel data acquisition. The specific antibody deposition sequence used in this study enabled detection of either 4 simulants or 4 bio-hazardous agents using a single consumable. The final device, a culmination of the multidisciplinary convergence of the fields of biology, chemistry, optoelectronics and microfluidics, is man-portable and inherently robust. The performance characteristics of the SpectroSens™ technology platform highlight its potential for exploitation as a 'detect to warn/treat' biodetector in security and defence operations.

A multi-channel optical microchip sensor system suitable for real-time, label-free detection of a wide range of biological agents is presented. SpectroSens (TM) chips containing multiple high-precision planar Bragg gratings are exploited as low-cost, robust refractive index sensors. Sensitivity to biological agents is conferred by functionalising individual sensing regions with different antibodies selected against numerous targets of interest. Antigen binding to the surface-immobilised antibodies results in localised changes in refractive index; upon laser-induced interrogation of the sensing region via optical fibres, these antibody-antigen interactions manifest as increases in wavelength of light reflected from the sensor chip. Real-time detection of multiple biological agents including bacterial cells/spores, viruses and toxins has been demonstrated. Further improvements to sensor performance including physical and chemical methods are also investigated. This multi-analyte capability highlights the potential use of this sensing technology in applications ranging from bio-hazard detection for defence purposes to point-of-care clinical diagnostics.

A rapid, label-free optical biosensor system for sensitive monitoring of bio-molecular interactions in real-time is presented. SpectroSens sensor chips are based on integrated planar Bragg gratings sensitive to localised changes in refractive index. Bio-molecule recognition is imparted by functionalisation of the sensing surface with antibodies against targets of interest. In this study, antibodies against selected proteins were oriented with recombinant Protein A/G, which was covalently immobilised to the sensor chip via amine coupling to a glutaraldehyde-activated silane layer. Immunoassays for the detection of rabbit IgG and ovalbumin proteins as model antibody-antigen interaction systems were performed. Binding of complementary antigens to respective antibody-functionalised sensors manifested as changes in wavelength of light reflected from the optical sensors. Quantitative binding kinetics with detection sensitivities in the mid ng/ml range were obtained for both antigens using this planar, two-dimensional surface coating. Data presented demonstrate the suitability of SpectroSens sensors as a valuable tool in life science research and development for monitoring bio-specific interactions, protein concentration determination and antibody selection; the optical integration and analytical characteristics of these sensors suggest that they may find numerous applications in bio-pharmaceutical development and clinical diagnostics.

A universal optical microchip sensing platform demonstrating real-time, label-free detection of a wide range of biological agents is presented. SpectroSens™ chips containing high-precision planar Bragg gratings are exploited as low-cost, robust refractive index sensors. Sensitivity to biological agents is conferred by functionalising the sensing surface with antibodies selected against targets of interest. Several methods for immobilisation of bio-molecules on the metal oxide-coated sensing surface have been investigated. In this study, surfaces were modified with an amino-terminated silane monolayer and activated by glutaraldehyde cross-linking for covalent attachment of recombinant Protein A/G, to which agent-specific antibodies were immobilised. Binding of target antigens (introduced under flow) to the surface-immobilised antibodies results in localised changes in refractive index; upon laser-induced interrogation of the sensing region via optical fibres, these antibody–antigen interactions manifest as increases in wavelength of light reflected from the Bragg grating. Detection of biological targets including proteins (ovalbumin<10nm), viruses (MS2<100nm), bacterial cells (Escherichia coli>1μm) and spores (Bacillus atrophaeus>1μm) in real-time has been demonstrated. The large size range of detection targets is attributed to a large penetration depth of the sensing light of >1μm into the sample liquid using these sensors. This multi-analyte capability enables utilisation of this sensing technology in applications ranging from bio-threat detection for defence and homeland security to point-of-care clinical diagnostics.

An enzyme-linked holographic sensor for the detection of spore-specific calcium dipicolinate (Ca-DPA) is described. Small, acid-soluble spore proteins (SASPs) extracted from dormant spores are exploited as holographic matrices. Reflection holograms are incorporated into SASP films using a diffusion method for the distribution of ultra-fine silver bromide grains into pre-formed polymers, followed by holographic recording using a frequency doubled Nd:YAG (532nm) laser. Changes in holographic diffraction intensity are used to characterise the structural integrity of SASP-based matrices in response to proteolytic attack by recombinant germination protease (GPRS), following activation with Ca-DPA; the holographic response manifests as an irreversible reduction in diffraction intensity due to progressive fringe removal caused by SASP degradation. This work demonstrates proof-of principle of a Ca-DPA-activated enzyme-linked holographic sensor, highlighting the potential of a system whereby Ca-DPA released from germinating spores may activate GPR in the vicinity of an SASP-based holographic matrix; such a sensor system presents a novel platform for the selective detection of bacterial spores with intrinsic signal amplification.

A novel optical sensor system for rapid, sensitive and robust biological detection is presented. Sensor elements based on integrated optical circuits confine all optical signals into a planar format, resulting in a small, low-cost and mechanically stable refractive index sensor, without any external bulk optics. Consequently, the sensor elements are able to operate in real-world environments, resilient to vibration and temperature changes, whilst maintaining refractive index resolution of 10-6. Oxide surfaces on the sensor are ideal for protein attachment and have a long lifetime in buffer solutions (>100hrs). Real-time, label-free detection of biological agents has been demonstrated using antibodies attached to the sensor surface. The sensor design results in a large penetration depth of the sensing light, up to 1μm into the sample liquid, conferring the ability to detect various classes of biological targets, spanning toxins, viruses and bacteria. Each sensing element utilizes parallel multiple wavelength data to provide additional information at the point of measurement, resulting in on-chip temperature and strain referencing, focused towards increased accuracy and reduction of false alarms. The large size range of biological detection, coupled with the long lifetime of the sensors makes the system ideally suited to applications ranging from medical diagnostics to confirmatory detectors for homeland security

Holographic sensors for the detection of Bacillus species spore germination and vegetative growth are described. Reflection holograms were fabricated using a diffusion method for the distribution of ultra-fine silver bromide grains into pre-formed polymer films, followed by holographic recording using a frequency doubled Nd:YAG (532 nm) laser. Changes in holographic replay wavelength or diffraction intensity were used to characterise the swelling behaviour or structural integrity of a range of holographic matrices in response to various extracellular products of bacterial spore germination and vegetative metabolism. Divalent metal ion-sensitive holograms containing a methacrylated analogue of nitrilotriacetic acid (NTA) as the chelating monomer were successfully used to monitor Ca2+ ions released during B. subtilis spore germination in real-time, which was within minutes of sample addition; the holographic response manifested as a 16 nm blue-shift in diffraction wavelength over the progress of germination. Similarly, pH-sensitive holograms comprising methacrylic acid (MAA) as the ionisable monomer were responsive to changes in pH associated with early vegetative metabolism following germination of B. megaterium spores; a visually perceptible blue-shift in holographic replay wavelength of 75 nm was observed. Casein and starch-based holographic matrices, prepared by co-polymerisation of the appropriate substrate with acrylamide, were used to detect exo-enzymes released during later stages of B. megaterium and B. subtilis vegetative cell growth; holographic responses of both matrices were visible as a reduction in diffraction intensity due to progressive fringe disruption caused by enzymatic cleavage. The combined monitoring of various germination and growth events using the range of aforementioned holographic sensors provides a novel, comprehensive means for the detection of viable bacterial spores.