WO2013068341A1

WO2013068341A1 - Contactless detection of specific molecules from oral exhaust of ruminantia, measurement method and related molecules

Info

Publication number: WO2013068341A1
Application number: PCT/EP2012/071918
Authority: WO
Inventors: Christian Jonasson; Johan STIGWALL; Karin LERSTEN; Marie ÖBERG; Bengt-Ove RUSTAS
Original assignee: Imego Ab
Priority date: 2011-11-07
Filing date: 2012-11-06
Publication date: 2013-05-16
Also published as: EP2775912A1

Abstract

The present invention relates to molecular detection and specifically to detection of specific molecules from oral exhaust of ruminantia, without physical contact with the animal. The invention provides away to collect animal exhalation and eructation samples on a routine basis without interfering with or disturbing the animal. Furthermore, the invention provides specific marker molecules, measurement methods and suitable uses for these marker molecules.

Description

CONTACTLESS DETECTION OF SPECIFIC MOLECULES FROM ORAL

EXHAUST OF RUMINANTIA, MEASUREMENT METHOD AND RELATED MOLECULES Technical Field

The present invention relates to molecular detection and specifically to detection of specific molecules from oral exhaust of ruminantia, specific molecules, measurement method and use. Background of the Invention

Modern farms tend to have larger and larger numbers of cattle, such as cows, which makes the level of attention for every cow lower. At the same time, in order for a farm to be profitable, it is important that each cow delivers maximal amounts of milk, and that the milk has high quality. Taken together, this increases the demand for automatic equipment and systems for controlling the cattle, such as checking health status and fodder regimens. An efficient use of fodder is a crucial in order to create a competitive production, and individually optimized feeding may create a substantial advantage.

The fodder comprises fiber and energy in a delicate balance. However, when the composition is not optimal, the cow may suffer from disturbed rumen. This may have a negative impact on the quality of milk (Plaizer et al. 2009). Today, software exists, which helps a farmer plan the feeding regimen. However, there is no system providing sufficient possibilities to obtain feedback on a specific feeding regimen. It is possible to draw certain conclusion based on physical signals from cow, but due to genetic variations, it may be hard to get sufficient reliable results (Enemark et al. 2002).

Various methods and devices for extracting information from cattle exists. For example, US 2011/0192213 disclose a method and system for reducing methane emissions by ruminants. Also, it is possible to obtain samples from the rumen of a cow, but since this is an invasive procedure, this is cumbersome and associated with risks for the animal, such as increased risk of infection and increased stress levels. Thus, the existing methods are only suitable for small studies and not possible to scale up to track the status of all animals on a farm, which would be desirable.

Thus, to meet the increasing demands for high productivity, there is a need for an apparatus and a method which provides further information from ruminants, which may help obtain large scale data sets, for optimizing production, lower costs etc.

Summary

Accordingly, the present invention preferably seeks to mitigate, alleviate or eliminate one or more of the above-identified deficiencies in the art and disadvantages singly or in any combination and solves at least the above mentioned problems by providing contactless means to routinely sample substances without interfering with or disturbing the animal.

According to a first aspect of the invention, a sensor apparatus for detecting specific molecular substances from oral exhaust of ruminantia is provided. The apparatus comprises a mixing chamber for receiving the oral exhaust of an animal and a measurement chamber for detecting specific molecules, wherein a fan is mounted in relation to the measurement chamber so that, in use, the fan creates a gas flow from the mixing chamber to the measurement chamber, and wherein the sensor apparatus comprises a carbon dioxide sensor, a methane sensor and, at least one substance sensor connected to a processing unit, wherein the mixing chamber comprises an inlet for the oral exhaust and the measurement chamber comprises inlets for the carbon dioxide sensor, the methane sensor and the substance sensor.

This is advantageous, since it enables contactless sampling of specific substances, which does not interfere with the normal daily life of the animal.

In an embodiment the mixing chamber and the measurement chamber are separated by a gas permeable separator, such as a filter.

An advantage with this is that the gas permeable separator may collect excess moisture and/or particles, which makes the readings from the sensors more accurate.

In an embodiment the carbon dioxide sensor, the methane sensor and/or the substance sensor are connected to a sensor pump. This is advantageous, since the sensors may then operate in a more stable fashion, since the sensor pump enables a continuous gas flow. The sensor pump may also increase the flow rate through the sensors, which lowers the sensor response time.

The sensors may be selected from are selected from the group consisting of non-dispersive infrared (NDIR) sensor, solid-state gas sensor, solid electrolyte sensor, tunable diode laser absorption spectroscopy (TDLAS) sensor, Fourier transform infrared spectroscopy (FTIR) sensor, gas chromatography (GC-MS) sensor, gas chromatography/flame ionization detection (GC-FID) sensor, proton transfer reactor mass spectroscopy sensor, mass spectroscopy sensor, ion mobility spectrometry/mass spectroscopy (IMS-MS) sensor, cavity ring-down spectroscopy sensor, or combinations thereof.

In an embodiment, the apparatus further comprises a temperature sensor, for measuring the temperature of the oral exhaust.

This is advantageous, because it can measure the temperature of the flow which reaches the other sensors. Since the normal body temperature of a cow is about 38°C and 39°C, if the ambient temperature is known it is possible to estimate the dilution of the exhaust from the animal.

In an embodiment, the apparatus further comprises animal positioning means, such as based on ultrasound, radio -frequency identification (RFID), IR or camera technology for detecting the presence an animal in proximity to the apparatus.

An advantage with this is that it is possible to measure distance between animal and apparatus, which improves the accuracy of the dilution measurement.

In an embodiment, the sensor apparatus further comprises animal identification means, for detection of a specific animal in proximity to the apparatus.

This is advantageous, since it enables monitoring of individual behavior and tracking of specific animals.

In an embodiment, the sensor apparatus comprises more than one substance sensor.

This is advantageous, since several substances may then suitably be measured. The substance sensor/s may be a detector for detecting a molecular marker substance, such as selected from the group consisting of benzene, ethylhexanol, butanol, acetone, isoprene, toluene, xylene, limonene, hexane, methylcyclopentane, dimethyl sulfide, 2-butanone, heptane, ethyl acetate, ethanol and ammonia, and combinations thereof.

According to a second aspect of the invention, a system is provided, for detecting specific molecules from oral exhaust of ruminantia, comprising a plurality of sensor apparatus units according to the first aspect, connected to a central processing unit.

According to a third aspect, a milking machine is provided, comprising the sensor apparatus according the first aspect, or the system according to the second aspect.

According to a fourth aspect, a fodder station is provided, comprising the sensor apparatus according the first aspect, or the system according to the second aspect.

According to a fifth aspect, a method for detecting specific molecular substances from oral exhaust of ruminantia is provided. Said method comprises the steps of obtaining continuous carbon dioxide, methane and at least one specific molecule concentration data from oral exhaust of ruminantia; utilizing the carbon dioxide and methane data to calculate the dilution of the oral exhaust; and calculating the actual concentration of specific molecular substances in the exhaust based on the calculated dilution and the specific molecular concentration data.

The specific molecule may be selected from benzene, ethylhexanol, butanol, acetone, isoprene, toluene, xylene, limonene, hexane, methylcyclopentane, dimethyl sulfide, 2-butanone, heptane, ethyl acetate, ethanol and ammonia, and combinations thereof.

In an embodiment, the step of utilizing the carbon dioxide and methane data comprises the steps sorting eructation sample data and lung sample data into two segments, by discriminating between eructation samples and lung samples; and compensating for dilution in both eructation samples and lung samples. Optionally, the method may further comprise a second step of compensating for inter-source cross-talk between the lung sample data and eructation sample data. In an embodiment, the step of sorting eructation sample data and lung sample data may comprise the steps of, optionally, performing sample-wise normalization of methane data against the carbon dioxide data; applying band-pass filtering to emphasize sudden increments in the methane; finding relevant peaks by selecting local maxima in the band passed methane signal; applying a threshold value to the peaks; and

determining the time of the eructations based on the peaks fulfilling the threshold value.

According to a sixth aspect, the invention comprises a method for detecting a symptom in ruminantia, comprising the steps of detecting abnormal levels of specific molecular substances from oral exhaust of ruminantia; correlating the abnormal levels to reference data, indicative of the symptom; and thereby detecting the symptom.

In an embodiment, the molecular substance is benzene, ethylhexanol, butanol, acetone, isoprene, toluene, xylene, limonene, hexane, methylcyclopentane, dimethyl sulfide, 2-butanone, heptane, ethyl acetate, ethanol or ammonia, or combinations thereof.

In an embodiment, where the symptom is related to high protein content, the specific molecular substances are selected from the group consisting of: hexane, methylcyclopentane, dimethyl sulfide, 2-butanone, and combinations thereof.

In an embodiment, where the symptom is related to high starch content, the specific molecular substances are selected from the group consisting of dimethyl sulfide, 2-butanone, ethylhexanol, butanol, acetone, toluene, limonene, heptane, ethyl acetate, ethanol, ammonia, or combinations thereof.

The present invention has the advantage over the prior art that provides a way to collect animal exhalation and eructation samples on a routine basis without interfering with or disturbing the animal. The invention also enables collecting and analyzing exhaled air and eructation for marker compounds, closely connected to both levels of said compounds in blood and rumen. This is advantageous, since it enables diagnosis of rumen dysfunction in individual ruminant animals of a herd or livestock. The invention provides for reliable quantification of said marker compounds regardless of air dilution of the sample at the sampling moment. Brief Description of the Drawings

These and other aspects, features and advantages of which the invention is capable of will be apparent and elucidated from the following description of

embodiments of the present invention, reference being made to the accompanying drawings, in which

Fig. 1 is a schematic illustration of a sensor apparatus according to an embodiment;

Fig. 2 is a schematic illustration of a system according to an embodiment;

Figs 3 to 7 are graphs illustrating different results according to an embodiment;

Fig. 8 are different views of the apparatus according to an embodiment, Fig. 8A is a perspective view and Fig. 8B is a side view;

Figs 9 and 10 are schematic overviews of methods according to embodiments of the invention;

Fig. 11 A is a graph of pH level in rumen of animals and Fig. 1 IB is a graph of corresponding level of butanol; and

Figs 12 to 15 are graphs showing correlations between substances and protein content in fodder.

Description of Embodiments

Several embodiments of the present invention will be described in more detail below with reference to the accompanying drawings in order for those skilled in the art to be able to carry out the invention. The invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. The embodiments do not limit the invention, but the invention is only limited by the appended claims. Furthermore, the terminology used in the detailed description of the particular embodiments illustrated in the accompanying drawings is not intended to be limiting of the invention.

In an embodiment according to Fig. 1, a contactless sensor apparatus 10 for detecting specific molecules from oral exhaust of ruminantia is provided. The oral exhaust is partly a gas flow originating from the lungs of the animal and partly an eructation exhaust originating from the digestion system of the animal (rumen).

Ruminantia, typically cattle (Bos taurus), raised as livestock for meat, or as dairy animals, even though the invention is not limited to any specific animal.

An important feature of the contactless sensor apparatus 10 is that no part of the apparatus is intended to be in physical contact with the animal. This is

advantageous, since the apparatus may then be easier to install and/or use, which provides easier and cheaper measurements.

A perspective view of the contactless sensor apparatus 10 is provided in Fig. 8A and a side view is provided in Fig. 8B. The arrows in these figures are examples of how the air may flow through the apparatus 10.

The apparatus 10 has a mixing chamber 100 and a measurement chamber 110, separated by a gas permeable separator 150. The mixing chamber 100 is, in use, placed close to the muzzle of an animal, so that the oral exhaust of the animal is directed into the mixing chamber 100. In the mixing chamber 100, the exhaust is allowed to diffuse and mix so that it becomes more homogenous. This is advantageous, since

measurements may then be more accurate. The mixing chamber 100 is placed under the measurement chamber 110, which is advantageous since it collects liquids and/or larger particles due to gravity. In an embodiment (not shown), the mixing chamber 100 may also have a drainage device, such as a hole. However, as will be appreciated by a person skilled in the art, the mixing chamber may also be positioned above or along the side of the measurement chamber 110. The exhaust is directed from the mixing chamber 100 into the measurement chamber 110 by a fan 120. The fan 120 is positioned along the outer wall of the measurement chamber 110 in fluid communication with the chamber, such as by a hole, so that it sucks air into the mixing chamber 100, through the gas permeable separator 150 and the measurement chamber 110. This is advantageous because it enables sufficient circulation in the measurement chamber 110. Another advantage is that the exhaust is more effectively sampled, since it is actively drawn into the mixing chamber 100. The fan 120 may also increase the active range of the apparatus, i.e. increase the maximum distance between the muzzle of the animal and the mixing chamber 100, needed for correct operation. Typically, this distance may be between 0 cm and 20 centimeters, and with the fan 120 between 0 cm and 40 cm.

Another advantage with the fan 120 is that it improves the response rate of

measurement instruments in the measurement chamber 110. The mixing chamber 100 also serves to collect and allow the exhaust from the animal to properly diffuse, so that the gas is homogenous when it reaches the measurement chamber 110. In the measurement chamber 110, a carbon dioxide sensor 130 and a methane sensor 140 are positioned. Having a mixing chamber 100 is advantageous since the sensors in the measurement chamber will then receive gas samples with identical properties.

Having both a carbon dioxide sensor 130 and a methane sensor 140 is advantageous, since it makes it possible to distinguish between gas flow originating from the lungs of the animal and partly eructation exhaust originating from the digestion system of the animal.

The inventors have noted that carbon dioxide level in exhaled air directly from the lungs (the alveolar concentration) of cow is about 5-6%. Even though the exact figure may vary, it is however important to note that the carbon dioxide level is constant, to enable detection of deviating concentration levels, such as abnormalities, in the concentration/s of other molecules. The carbon dioxide level in rumen of cow is about 60-80%. The inventors have also noted that it is possible to approximate the total gas content of the gas flow from rumen (eructation exhaust) to consist only of carbon dioxide and methane. Thus, it is possible to measure dilution from the exhaust, which is important to provide a dilution reference.

Thus, by approximating that the exhaust from the cow is either lung exhaust or eructation exhaust, it is possible to know the dilution rate. If it is eructation exhaust, the total amount of carbon dioxide and methane corresponds to 100% in the rumen, and if the exhaust is lung exhaust, the measured carbon dioxide level corresponds to 5-6%> in the lungs. Thus, this gives a proper dilution reference both in the case of lung exhaust and eructation exhaust.

The carbon dioxide sensor 130 may be any kind of sensor suitable to measure carbon dioxide levels, such as a sensor selected from the group consisting of infrared (IR) sensor, non-dispersive infrared (NDIR) sensor, solid-state sensor, solid electrolyte sensor, tunable diode laser absorption spectroscopy (TDLAS) sensor, Fourier transform infrared spectroscopy (FTIR) sensor, gas chromatography (GC-MS) sensor, gas chromatography/flame ionization detection (GC-FID) sensor, proton transfer reactor mass spectroscopy sensor, mass spectroscopy sensor, ion mobility spectrometry/mass spectroscopy (IMS-MS) sensor, cavity ring-down spectroscopy sensor, or combinations thereof. However, in this particular embodiment, the carbon dioxide sensor may be a NDIR sensor (such as the C0₂ Engine® K30 FR from SenseAir AB, Sweden).

The methane sensor 140 may be any kind of sensor suitable to measure methane levels, such as a sensor selected from the group consisting of infrared (IR) sensor, non-dispersive infrared (NDIR) sensor, solid-state sensor, solid electrolyte sensor, tunable diode laser absorption spectroscopy (TDLAS) sensor, Fourier transform infrared spectroscopy (FTIR) sensor, gas chromatography (GC-MS) sensor, gas chromatography/flame ionization detection (GC-FID) sensor, proton transfer reactor mass spectroscopy sensor, mass spectroscopy sensor, ion mobility spectrometry/mass spectroscopy (IMS-MS) sensor, cavity ring-down spectroscopy sensor, or combinations thereof. However, in this particular embodiment, the methane sensor is a NDIR prototype sensor based on same principles as the carbon dioxide sensor 130 above.

In this particular embodiment, the apparatus 10 is a plastic box with suitable perforations for allowing the gas to flow between the compartments, and subsequently exit the box. The exact design of the box will be appreciated by the person skilled in the art. However, in one embodiment (not shown) the mixing chamber is simply a space on one side of perforated surface, which space is in proximity to the muzzle of the animal. The measurement chamber is positioned on the other side of the perforated surface and the fan is directing the gas flow into it, for subsequent measurement.

This is advantageous, since it is easier to install and more durable.

In an embodiment, the gas permeable separator 150 is a filter and is positioned between the mixing chamber 100 and the measurement chamber 110. This may be any kind of gas permeable filter, which may be a standard exhaust filter for fans, such as a filter of model PFA from Pfannenberg.

The filter is advantageous, since it collects excess moisture and/or particles, which makes the readings from the sensors 130, 140 more accurate. The apparatus 10 also has a substance sensor 170, which is intended to measure substances other than methane and carbon dioxide. The carbon dioxide sensor 130 and the methane sensor 140 constitute the sensory reference system, and the substance sensor 170 constitutes the sensory substance measurement system.

This is advantageous, since the substance sensor 170 may then be adapted according to the specific substance which is to be analyzed. Thus, the exact

specification of the substance sensor 170 may vary.

In an embodiment, more than one substance sensor 170 is used.

The substance sensor 170 may be any kind of sensor suitable to measure the specific substance or substances of interest, such as a sensor selected from the group consisting of infrared (IR) sensor, non-dispersive infrared (NDIR) sensor, solid-state sensor, solid electrolyte sensor, tunable diode laser absorption spectroscopy (TDLAS) sensor, Fourier transform infrared spectroscopy (FTIR) sensor, gas chromatography (GC-MS) sensor, gas chromatography/flame ionization detection (GC-FID) sensor, proton transfer reactor mass spectroscopy sensor, mass spectroscopy sensor, ion mobility spectrometry/mass spectroscopy (IMS-MS) sensor, cavity ring-down spectroscopy sensor, or combinations thereof. If a full-spectrum sensor technology such as FTIR is used, the same sensor can be used for measuring both the reference gas concentrations (carbon dioxide and methane) and the specific substance or substances of interest. Thus, in one embodiment, the carbon dioxide sensor 130 and the methane sensor 140 is the same sensor. In another embodiment, the carbon dioxide sensor 130, the methane sensor 140 and the substance sensor 170 is the same sensor.

This is advantageous, because it is less expensive, or may require less space in the apparatus 10.

If an NDIR sensor similar to the ones used for the reference gases is used as the substance sensor 170, the sensor may be modified by replacing the detection filter with a custom designed one to target suitable wavelengths, specific for the substance to be analyzed.

As will be appreciated by a person skilled in the art, the respective specific substance or substance of interest may vary. Examples of substances detectable in the oral exhaust of ruminantia may however be selected from the group consisting of benzene, ethylhexanol, butanol, acetone, isoprene, toluene, xylene, limonene, hexane, methylcyclopentane, dimethyl sulfide, 2-butanone, heptane, ethyl acetate, ethanol and ammonia.

In an embodiment, the sensors 130, 140, 170 are connected to a sensor pump 180, through a tube connected to the sensor outlet. Said sensor pump 180 is diverting a portion of the gas flow from the fan 120 into the sensors 130, 140, 170. An advantage with this is that the sensors 130, 140, 170 may operate in a more stable fashion, since the sensor pump 180 enables a continuous gas flow. The sensor pump 180 also increases the flow rate through the sensors, which lowers the sensor response time.

In an embodiment (not shown), each sensor 130, 140, 170 has a separate sensor pump 180. This is advantageous, because the flow through in each sensor may then be individually regulated, which may provide optimized sensor operation.

As will be appreciated by a person skilled in the art, the pump 180 may also be a fan, i.e. a sensor fan, positioned in proximity to the sensor inlet/s or outlet/s and thus configured to "push" or "pull" a stream of gas into, or out from, the sensor/s.

An advantage with this is that the sensor may operate a more stable fashion, since the sensor fan enables a continuous gas flow. The sensor fan also increases the flow rate through the sensors, which lowers the sensor response time.

The sensors may be any kind of sensors suitable to detect the desired molecules, such as one or several sensors selected from the group consisting of non- dispersive infrared (NDIR) sensor, solid-state gas sensor, solid electrolyte sensor, tunable diode laser absorption spectroscopy (TDLAS) sensor, Fourier transform infrared spectroscopy (FTIR) sensor, gas chromatography (GC-MS) sensor, gas chromatography/flame ionization detection (GC-FID) sensor, proton transfer reactor mass spectroscopy sensor, mass spectroscopy sensor, ion mobility spectrometry/mass spectroscopy (IMS-MS) sensor, cavity ring-down spectroscopy sensor, or combinations thereof.

The sensors operate at a measurement rate of 2 Hz. However, the measurement rate may be adapted. If the measurement rage is more frequent, it is possible to follow the singular breaths of the animal. This is advantageous, since knowledge of the breath rate may enable estimating the total gas volumes produced, and also serves as an indicator of animal health status.

The sensors 130, 140, 170 are connected to a processing unit 190, and send signals to said unit. The processing unit 190 processes the sensor signals in a way which will be explained further below.

The processing unit 190 may be a single unit or several units, normally used for performing the involved tasks, e.g. a hardware, such as a digital processor with a memory, running suitable software. The processor may be any of variety of processors, such as Intel or AMD processors, CPUs, microprocessors, Programmable Intelligent Computer (PIC) microcontrollers, Digital Signal Processors (DSP), etc. However, the scope of the invention is not limited to these specific processors. The memory may be any memory capable of storing information, such as Random Access Memories (RAM) such as, Double Density RAM (DDR, DDR2), Single Density RAM (SDRAM), Static RAM (SRAM), Dynamic RAM (DRAM), Video RAM (VRAM), etc. The memory may also be a FLASH memory such as a USB, Compact Flash, SmartMedia, MMC memory, MemoryStick, SD Card, MiniSD, MicroSD, xD Card, TransFlash, and MicroDrive memory etc. However, the scope of the invention is not limited to these specific memories.

In an embodiment (not shown), the processing unit 190 is integrated with the sensors 130, 140, 170.

As will be appreciated by a person skilled in the art, the electricity supply and/or data transfer capabilities of the sensors 130, 140, 170, the fan 120, and the pump 180 may be any kind of wiring suitable to enable operation of the sensors 130, 140, 170, the fan 120, and the pump 180.

In an embodiment (not shown) a temperature sensor is comprised in the apparatus.

This is advantageous, because it can measure the temperature of the flow which reaches the other sensors. Since the normal body temperature of a cow is about 38°C and 39°C, if the ambient temperature is known it is possible to estimate the dilution of the exhaust from the animal. In an embodiment (not shown), the apparatus is further equipped with animal positioning means. The positioning means may be suitable for detecting the presence of an animal in proximity to the apparatus (10). This is advantageous, because it is possible to measure distance between animal and apparatus, which improves the accuracy of the dilution measurement. In an embodiment, such positioning means are based on radio-frequency identification (RFID), IR, ultrasound, or camera technology, such as CCD or CMOS.

The positioning means may also be suitable for detecting the presence of a specific individual animal in proximity to the apparatus (10) and thus identify the individual. This is advantageous, since it enables monitoring of individual behavior and tracking of specific animals.

In an embodiment, such positioning means are based on RFID technology. In an embodiment, the processing unit 190 is equipped with wired or wireless communication means, for sending and transmitting data.

This is advantageous, because several sensor apparatus 10 units may then be interconnected and/or connected to a central processing unit.

Thus, in an aspect of the invention according to Fig. 2, a system 20 for detecting specific molecules from oral exhaust of ruminantia is provided. A plurality of sensor apparatus units 10 are connected to a central processing unit 200.

The central processing unit 200 may be a single unit or several units, normally used for performing the involved tasks, e.g. a hardware, such as a processor with a memory, running suitable software. The processor may be any of variety of processors, such as Intel or AMD processors, CPUs, microprocessors, Programmable Intelligent Computer (PIC) microcontrollers, Digital Signal Processors (DSP), etc. However, the scope of the invention is not limited to these specific processors. The memory may be any memory capable of storing information, such as Random Access Memories (RAM) such as, Double Density RAM (DDR, DDR2), Single Density RAM (SDRAM), Static RAM (SRAM), Dynamic RAM (DRAM), Video RAM (VRAM), etc. The memory may also be a FLASH memory such as a USB, Compact Flash, SmartMedia, MMC memory, MemoryStick, SD Card, MiniSD, MicroSD, xD Card, TransFlash, and MicroDrive memory etc. However, the scope of the invention is not limited to these specific memories.

In an embodiment (not shown), the sensor apparatus 10 is mounted in a milking machine. Thus, according to an aspect, a milking machine is provided, comprising the sensor apparatus 10.

In an embodiment (not shown), the sensor apparatus 10 is mounted in a fodder station, for feeding animals such as cow. Thus, according to an aspect, a fodder station is provided, comprising the sensor apparatus 10. Signal processing

In order to create a reliable baseline, the processing unit 190 of the sensor apparatus 10 is configured to process the signals from the carbon dioxide sensor 130 and the methane sensor 140. In an embodiment, the process is performed according to the following algorithm.

Inputs to the algorithm are two signals: the carbon dioxide concentration, as measured by the carbon dioxide sensor 130, and the methane concentration, as measured by the methane sensor 140. Normally, these are given in PPM (parts per million). An example of obtained concentrations is shown in Fig. 3, where the top curve is carbon dioxide (C0₂) and the bottom curve is methane (CH₄).

First, the two signals are synchronized in time to compensate for differences in delay. This is done by calculating the cross-correlation between the two and selecting the delay that gives the highest correlation. The signals are then shifted accordingly. For this example data, the shift is given above in the data output of the previous section.

For a specific reference collection system (and settings), the delay between carbon dioxide and methane signals should be constant, reducing the need to perform per-sample time synchronization. Instead, the instrument can be calibrated once and for all after which the same delay is used to correct the timing in all measurements.

Then, to compensate for variations in mixing with surrounding air, the methane signal may be normalized against the carbon dioxide signal by element- wise division. However, if the carbon dioxide varies above a specific threshold value, the methane signal can also be used without compensation. The compensated methane value in this example is shown in Fig. 4.

To find the time of an eructation, or burp, the compensated (or noncompensated) signal is filtered using a type of edge detection filter. This filter is in effect a band-pass filter that passes typical eructation frequencies and rejects both very low frequencies (offset levels) as well as high frequency noise. Another interpretation is that it is a temporal differentiation, which acts as a high pass filter, followed by low pass filtering. It has been designed in the time domain to give a single soft local maxima at each eructation instant. In effect, it relates one time range of the signal to another, separated 30 seconds apart. At the eructation instant there will be an unusually large difference between the two 30-second blocks. The soft edges reduce the impact of noise compared to if a square kernel was used. The filter kernel used is a 60-seconds long Hann window multiplied by a linear ramp, where the Hann-part creates soft tails to the sides and the linear ramp (-1 : 1) inverts the sign around zero (also in a soft manner). It is designed so that a step up in methane concentration results in a positive local maxima after convolution.

The length of the filter implies that it will work optimally on eructation periods greater than -60 seconds (or slightly less). To detect much faster eructation, the length may be reduced at the cost of higher noise sensitivity.

Next, a convolution function is applied, i.e. the sum of sample wise products between a signal and a kernel time series (the latter mirrored in time) for all possible time-offsets between the two.

A kernel function is shown in Fig. 5 and the filtered, compensated methane concentration is shown in Fig. 6.

After edge filtering, the eructation times are retrieved by looking for points that are local maxima and have filtered values above zero. Local maxima are found by looking for zero crossings in the first differentials coinciding with negative second differentials.

The output from the algorithm is the timing of each individual eructation, which may be used to switch e.g. a sample collection system so that the collected gas consists either of gas with a larger amount of gas from the rumen (eructations) or from the lungs (non-eructations). Due to the length of the filter kernel, the output will be delayed by 30 seconds. This is no problem if the marker analysis system is also running continuously, since its output data can then be delayed the same amount in software.

From the timing information it is trivial to calculate higher level statistics such as e.g. average eructation frequency and variability in the eructation-to-eructation duration.

The compensated methane concentration, the edge filtered signal and positions where an eructation is detected are shown in Fig. 7. The dotted lines are detected eructations, the signal with large fluctuations is the edge filtered signal and the smaller signal in the middle is the compensated methane concentration.

Thus, in an embodiment according to Fig. 9, a method 90 for determining the timing of eructations in ruminants from a continuously recorded measurement of carbon dioxide and methane concentrations is provided. The method comprises a step of obtaining continuous carbon dioxide and methane concentration data. Optionally, the method may comprise a first step of performing 900 sample-wise normalization of methane data against the carbon dioxide data. Next, the method comprises a step of applying 910 band-pass filtering to emphasize sudden increments in the methane and a step of finding 920 local maxima, in the band passed methane signal. Next, the method comprises a step of applying 930 a threshold value to the peaks, thus enabling selection of the peaks above a certain threshold, which enables determining 940 the time of the eructations based on the peaks fulfilling the threshold value.

The band-pass filter applied 910 may be any kind of algorithm that gives a strong response to the steep upward slope in methane concentration at the eructation instant.

In an embodiment, the band-pass filter has a pass band starting at -0.008 Hz (125s period) and ending at -0.025 Hz (40s period).

In an embodiment, the band-pass filter is a FIR filter with a kernel that is a Hanning window multiplied by a linear ramp passing through the origin:

kernel = ha * t, wherein

0 < n < N, t = n * dt

The band-pass filter may also be a matched filter, which is optimized to a prerecorded recording of the methane level during an eructation event.

The peak finding step 920 may be selected to finds all points with zero first derivative and negative second derivative.

In an embodiment, applying 930 a threshold may be done by accepting all peaks with a filtered concentration value above a first threshold and a second derivative below a second threshold.

The pattern of the carbon dioxide signal at an eructation event may also be used to further enhance the accuracy of the method 90 for determining the timing of eructations. In particular, one may choose to not use the methane level alone, but instead use the relative level compared to the carbon dioxide level. This is

advantageous, since it may reduce sensitivity to fluctuations of the dilution.

Once the eructations have been detected, e.g. with the abovementioned method 90, it provides a basis for separately calculating the relative concentration of specific gaseous molecules, so called markers, from oral exhaust of ruminantia.

The data from about 30s before each eructation (with some margin to reduce contamination from the upcoming eructation) is used to primarily characterize the gases from lungs. The data from about 10s after each eructation is used to primarily characterize gases from the rumen.

However, if two eructations are too close in time, such as less than about 60 seconds, there may still be rumen-gas in sample segment a, therefore that segment is ignored.

Next, a compensation for dilution may be performed, e.g. by normalizing against:

Carbon dioxide level (which is assumed to be -5% in the lungs) =>

'-marker ^— '-measured * (5% -C02)

C0₂ + CH₄ is assumed to be -100% in the rumen. Although C0₂ and CH₄may vary individually, the sum of the two does not, which provides for a a good dilution estimate.

cmarker ^{= c}measured/ (^CC02 + ^CCH₄) Another possibility for compensating dilution is by relating the temperature of the analyzed air to that of the surrounding air. This will have to be done separately for lung/rumen if the temperature of the gas flow from the two sources differ.

'-marker ^— '-measured * (Tcow ^~ T_envi_r0nment)/(Tmeasured ^_ ^environment)

In an embodiment, a combination of gas levels and temperature is used, for improved accuracy.

To further reduce cross-talk between the two sources (lungs and rumen), it is assumed that the measured concentrations in a time segment a (lung selection) and a time segment b (rumen selection) sum up contributions both from the primary source (lungs for segment a, rumen for segment b), but also from the secondary one, according to the formulas:

¾ ^—

d2t> rumen

wherein the left hand side are the measured concentrations and d are the dilution factors. To obtain source concentrations corrected for the non-complete separation, the equation system should be solved for c_lung and c_rumen in a way which is obvious to a person skilled in the art.

As is obvious to a person skilled in the art, the abovementioned calculation provides compensation for inter-source cross-talk between the lung sample data and eructation sample data, i.e. the data in segments a and b, respectively.

When dilution compensation is performed, it is possible to accurately determine the levels of substance molecules in the oral exhaust sample.

Thus, in an embodiment according to Fig. 10, a method 100 for detecting specific molecules from oral exhaust of ruminantia is provided.

The method 100 comprises a step of obtaining 1000 continuous carbon dioxide, methane and at least one specific molecule concentration data from oral exhaust of ruminatia.

Next, the method 100 comprises a step of utilizing 1100 the carbon dioxide and methane data to calculate the dilution of the oral exhaust, such as with the method 90 for determining the timing of eructations. Once the timing of eructations have been determined, it is possible to discriminate between eructation samples and lung samples by sorting 1200 eructation sample data and lung sample data.

The method 100 further comprises a first step of compensating 1300 for dilution in both eructation samples and lung samples.

The method 100 comprises a step of calculating the level of specific molecules in the eructation samples and lung samples, respectively.

The method 100 comprises an optional second step of compensating 1400 the finite mixing of lung- and rumen gas into time segments to obtain the concentration data with less cross-talk between the two sources of interest, such as lung sample data and eructation sample data.

Finally, the method comprises a step of calculating 1500 the actual

concentration of specific molecular substances in the exhaust based on the calculated dilution and the specific molecular concentration data

Molecular markers

Rumen is essentially a fermentation chamber in which the microbial population digests the diet into nutrients. The efficiency of the bovine digestive system is dependent upon a functional rumen. When the rumen becomes dysfunctional, feed digestion is impaired and the cow becomes susceptible to different diseases as well as a decrease in milk production, for example. When a change in nutrient supply takes place (new substrate by changing feed or delivery of substrate by altering feeding time) inefficiency is usually created since the microbes are forced to shift their metabolic activity as they adapt to the new nutrient regime. For healthy and high producing cows, it is critical that the rumen microbial population is fed with a composition of fodder that is optimal. The composition of the fodder affects the ruminal environment and if not optimal for the microbes, the content in the fodder is not utilized in an optimal way. This leads to decreased milk production due to the genetic potential of the cow not being fully utilized, and may also lead to health problems.

Thus, in the rumen of ruminantia, gaseous substances are formed in a fermentation process. These substances are present in the breath and eructation exhaust of the animal. Normally, there is a specific balance between such substances in an animal. This balance may be altered, such as if the fodder is not optimal or the animal is sick or in any other way disturbed or afflicted by various symptoms. The inventors have now surprisingly found a number of molecular markers that may be correlated to the status of the rumen and thereby also provide a possibility of detecting various kinds of imbalances and/or abnormalities, such as diseases, suboptimal feeding patterns etc.

The substances were found in three studies. In one study (starch), increased levels of starch (to disturb the natural balance in the rumen) were given to cow. In one study (protein), varied levels of protein were given to cow. In one study (starch/silage), two different feeding regiments were used according to a cross-over design. The feeding regimens were designed to give different effects on the micro flora of rumen, especially regarding starch content, but also quality of silage.

In all studies, the content of the oral exhaust (both lung and eructation) was studied and compared to the respective feeding regimen.

The starch study was performed by giving fodder with increased starch content, while simultaneously monitoring the animal. The pH levels were measured three or five times daily for each cow, according to standard methods. The rumen is buffered over a range of pH 5 to 8 mainly by phosphate and bicarbonate from saliva and bicarbonate from rumen fermentation. The buffering capacity goes down below pH 6. When the cow is fed with high starch content, more volatile fatty acids (VFA) are produced and the rumen pH drops. This can lead to diseases, for example but not limited to, sub-acute rumen acidosis, laminitis and acute clinical ruminal acidosis. Low pH is thus an indicator of suboptimal rumen status. In Fig. 11 A, a graph of the pH level is shown. The levels of volatile fatty acids (VFA) may also be used to detect suboptimal rumen status. These values were also measured according to standardized methods (data not shown). Specific substances were found to be linked to suboptimal rumen conditions indicated by pH level. In Fig. 1 IB, a graph of the occurrence of butanol (ng/1 sample) is plotted a number of days in the study. It can be seen that the level of butanol increases with maximum on day 16, which corresponds to the pH decrease shown in Fig. 11A. Similar results were found for the substances benzene, ethylhexanol, acetone, isoprene, toluene, xylene, and limonene (data not shown). The protein study was performed by feeding cow with fodder with low "L", medium "M" and high "H" raw protein content. The protein content in milk and the nitrogen and creatinine levels in urine samples were analyzed to study how cows responded to different protein content in the fodder. In Fig. 12, correlation between protein levels of the fodder and the protein levels of milk is shown. The three treatments had varying protein levels of the fodder: 12 (low "L"), 16 (medium "M") and 20 % (high "H") raw protein content. There was significant a significant difference, with a p value of 0.004 (data not shown).

Urine samples were taken from the cows. The urine samples were analyzed for nitrogen and creatinine levels according to standard methods.

In Fig. 13, the correlation between nitrogen in urine and dietary protein is shown for L, M and H; and in Fig. 14, the correlation between the quota

nitrogen/creatinine in urine and dietary protein is shown for L, M and H.

In Fig. 15, the correlation between dimethyl sulfide in oral exhaust (respire air) and dietary protein levels is shown. It can be seen that dimethyl sulfide levels correlate to the markers provided in Figs 12-14. Similar results were obtained for the markers hexane, methylcyclopentane, and 2-butanone, as can be seen in Table 1 below.

Table 1. Results from the protein study.

Substance Protein level Average cone. Min cone. Max cone.

Hexane L 14.1 -59.6 87.0

M 139.2 57.7 218.8

H 280.7 185.0 375.5

Methylcyclopentane L 0.927 -9.6 11.5

M 18.7 7.1 30.5

H 39.0 25.2 53.0

Dimethyl sulfide L 333.9 117.5 554.1

M 546.5 304.6 788.9

H 778.0 505.6 1049.0

2-butanone L 137.6 29.0 243.3

M 168.9 48.6 287.9

H 208.7 76.8 343.2

The starch/silage study also gave pH values and VFA values, like the starch study (data not shown). The abovementioned measurables serve as comparable reference values to be correlated with oral exhaust samples to identify suitable molecular markers in said oral exhaust.

The oral exhaust was subject to sampling and analysis according to the following.

Samples from oral exhaust from the cows were taken from the cows in their normal positions. Background samples from ambient air were also taken on every sample occasion.

Even though an apparatus according to an embodiment of the invention, as described above, could suitably be used for the sampling, a person skilled in the art will appreciate that also other apparatuses may be used.

The following substances were found to correlate with the reference data from the comparative studies in one starch study:

benzene, ethylhexanol, butanol, acetone, isoprene, toluene, xylene, and limonene.

The following were found to correlate with the reference data from the comparative studies in the protein study:

hexane, methylcyclopentane, dimethyl sulfide, and 2-butanone.

Furthermore, in the starch/silage study, the following substances were found to correlate with suitable reference data:

dimethyl sulfide, 2-butanone, ethylhexanol, butanol, acetone, toluene, limonene, heptane, ethyl acetate, ethanol and ammonia.

Thus, according to an aspect of the invention, there is provided marker substances detectable in the oral exhaust of ruminantia, preferably Bos taurus (cattle), and more preferably cow.

In an embodiment, the molecular markers are selected from the group consisting of: benzene, ethylhexanol, butanol, acetone, isoprene, toluene, xylene, limonene, hexane, methylcyclopentane, dimethyl sulfide, 2-butanone, heptane, ethyl acetate, ethanol and ammonia.

In an embodiment, the molecular markers are selected from the group consisting of: benzene, ethylhexanol, butanol, acetone, isoprene, toluene, xylene, and limonene.

These markers have been found to correlate with starch content of fodder, and thereby be useful to monitor and provide optimal fodder starch content. Starch is important, since it provides necessary energy to the cow, but only to a certain level. If the cow receives too much starch, the rumen is disturbed, which leads to problems regarding productivity. Thus, the abovementioned markers are advantageous, since they provide a possibility to prevent over feeding the cow.

In an embodiment, the molecular selected from the group consisting of:

hexane, methylcyclopentane, dimethyl sulfide, and 2-butanone.

These markers have been found to be correlated to the protein content of the fodder, and thereby be useful to monitor and provide optimal fodder protein content. This is advantageous, because it prohibits feeding the cow with higher protein levels than it can absorb. If the cow is fed excessive amounts of protein it leads to increased nitrogen discharge .

Furthermore, two substances were also found to correlate with the quality of silage. These are hexanal and acetic acid.

Although the present invention has been described above with reference to specific embodiments, it is not intended to be limited to the specific form set forth herein. Rather, the invention is limited only by the accompanying claims and, other embodiments than the specific above are equally possible within the scope of these appended claims.

In the claims, the term "comprises/comprising" does not exclude the presence of other elements or steps. Furthermore, although individually listed, a plurality of means, elements or method steps may be implemented by e.g. a single unit.

Additionally, although individual features may be included in different claims, these may possibly advantageously be combined, and the inclusion in different claims does not imply that a combination of features is not feasible and/or advantageous. In addition, singular references do not exclude a plurality. The terms "a", "an", "first", "second" etc do not preclude a plurality. Reference signs in the claims are provided merely as a clarifying example and shall not be construed as limiting the scope of the claims in any way.

Claims

1. A sensor apparatus (10) for detecting specific molecular substances from oral exhaust of ruminantia, comprising a mixing chamber (100) for receiving the oral exhaust of an animal and a measurement chamber (110) for detecting specific molecules, wherein a fan (120) is mounted in relation to the measurement chamber (110) so that, in use, the fan (120) creates a gas flow from the mixing chamber (100) to the measurement chamber (110), and wherein the sensor apparatus (10) comprises a carbon dioxide sensor (130), a methane sensor (140) and at least one substance sensor (170) connected to a processing unit (190), wherein the mixing chamber (100) comprises an inlet for the oral exhaust and the measurement chamber (110) comprises inlets for the carbon dioxide sensor (130), the methane sensor (140) and the substance sensor (170).

2. The sensor apparatus (10) according to claim 1, wherein the mixing chamber (100) and the measurement chamber (110) are separated by a gas permeable separator

3. The sensor apparatus (10) according to claim any of claims 1 or 2, wherein the carbon dioxide sensor (130), the methane sensor (140) and/or the substance sensor (170) are connected to a sensor pump (180).

4. The sensor apparatus (10) according to any of the preceding claims, wherein the sensors (130, 140, 170) are selected from the group consisting of non-dispersive infrared (NDIR) sensor, solid-state gas sensor, solid electrolyte sensor, tunable diode laser absorption spectroscopy (TDLAS) sensor, Fourier transform infrared spectroscopy (FTIR) sensor, gas chromatography (GC-MS) sensor, gas chromatography/flame ionization detection (GC-FID) sensor, proton transfer reactor mass spectroscopy sensor, mass spectroscopy sensor, ion mobility spectrometry/mass spectroscopy (IMS-MS) sensor, cavity ring-down spectroscopy sensor, or combinations thereof.

5. The sensor apparatus (10) according to any of the preceding claims, wherein the gas permeable separator (150) is a filter.

6. The sensor apparatus (10) according to any of the preceding claims, further comprising a temperature sensor, for measuring the temperature of the oral exhaust.

7. The sensor apparatus (10) according to any of the preceding claims, further comprising animal positioning means, for detecting the presence an animal in proximity to the apparatus (10).

8. The sensor apparatus (10) according to any of the preceding claims, further comprising animal identification means, for detection of a specific animal in proximity to the apparatus (10).

9. The sensor apparatus (10) according to claim 7 or 8, wherein the animal positioning means are based on ultrasound, radio-frequency identification (RFID), IR or camera technology.

10. The sensor apparatus (10) according to any of the preceding claims, comprising more than one substance sensor (170).

11. The sensor apparatus (10) according to any of the preceding claims, wherein said substance sensor (170), is a detector for detecting a molecular marker substance.

12. The sensor apparatus (10) according to claim 11, wherein the molecular marker substance is selected from from the group consisting of:

benzene, ethylhexanol, butanol, acetone, isoprene, toluene, xylene, limonene, hexane, methylcyclopentane, dimethyl sulfide, 2-butanone, heptane, ethyl acetate, ethanol and ammonia, and combinations thereof.

13. A system (20) for detecting specific molecules from oral exhaust of ruminantia, comprising a plurality of sensor apparatus units (10) according to any of claims 1 to 12, connected to a central processing unit (200).

14. A milking machine comprising the sensor apparatus according to any of claims 1 to 12, or the system according to claim 13.

15. A fodder station comprising the sensor apparatus according to any of claims 1 to 12, or the system according to claim 13.

16. A method (100) for detecting specific molecular substances from oral exhaust of ruminantia is provided, comprising the steps of:

obtaining (1000) continuous carbon dioxide, methane and at least one specific molecule concentration data from oral exhaust of ruminantia;

utilizing (1100) the carbon dioxide and methane data to calculate the dilution of the oral exhaust; and

calculating (1500) the actual concentration of specific molecular substances in the exhaust based on the calculated dilution and the specific molecular concentration data.

17. The method (100) according to claim 16, wherein specific molecule is selected from the group consisting of:

18. The method according to claim 16 or 17, wherein the step of utilizing (1100) the carbon dioxide and methane data comprises the steps of:

sorting (1200) eructation sample data and lung sample data into two segments, by discriminating between eructation samples and lung samples; and compensating (1300) for dilution in both eructation samples and lung samples; and

optionally, a second step of compensating (1400) for inter-source cross-talk between the lung sample data and eructation sample data.

19. The method (100) according to claim 18, wherein the step of sorting (1200) eructation sample data and lung sample data comprises the steps of:

optionally, performing (900) sample-wise normalization of methane data against the carbon dioxide data;

applying (910) band-pass filtering to emphasize sudden increments in the methane;

finding (920) relevant peaks by selecting local maxima in the band passed methane signal;

applying (930) a threshold value to the peaks; and

determining (940) the time of the eructations based on the peaks fulfilling the threshold value.

20. A method for detecting a symptom in ruminantia, comprising the steps of detecting abnormal levels of specific molecular substances from oral exhaust of ruminantia;

correlating the abnormal levels to reference data, indicative of the symptom; and thereby

detecting the symptom.

21. The method according to claim 20, wherein the molecular substance is benzene, ethylhexanol, butanol, acetone, isoprene, toluene, xylene, limonene, hexane, methylcyclopentane, dimethyl sulfide, 2-butanone, heptane, ethyl acetate, ethanol or ammonia, and combinations thereof.

22. The method according to claim 21, wherein the symptom is related to high protein content and the specific molecular substances are selected from the group consisting of:

hexane, methylcyclopentane, dimethyl sulfide, 2-butanone, and combinations thereof.

23. The method according to claim 21, wherein the symptom is related to high starch content and the specific molecular substances are selected from the group consisting of:

dimethyl sulfide, 2-butanone, ethylhexanol, butanol, acetone, toluene, limonene, heptane, ethyl acetate, ethanol, ammonia, and combinations thereof.