Method for analyzing samples using cavity ring-down spectroscopy and method for generating predictive models

Abstract

To disclose a method and a system for analyzing a sample using cavity ring-down spectroscopy and generating a prediction model.SOLUTION: At least a portion of a sample is input into a ring-down cavity. A laser beam is generated and directed into the ring-down cavity for each wavelength in the set of wavelengths. The laser beam entering the ring-down cavity is dissipated. The light intensity attenuation data of the light emitted from the ring-down cavity is recorded with a light intensity sensor system. From the light intensity decay data for a set of wavelengths, a probability that an analyte from which the sample was withdrawn has a given physiological state or degree of physiological state is determined using at least indirectly a data set of light intensity decay data from a pre-analyzed sample from which the presence or absence of that physiological state or degree of physiological state has been identified.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] This application claims priority to U.S. Provisional Application No. 62 / 828,750, filed April 3, 2019, the entire contents of which are incorporated herein by reference.

[0002] This application relates generally to spectroscopy, and more particularly to methods for analyzing samples and generating predictive models using cavity ring-down spectroscopy. [Background technology]

[0003] Diagnosing certain physiological conditions can be challenging. One such condition is lung cancer, which is one of the most prevalent and deadly forms of cancer. While lung cancer can have an improved prognosis if detected at an early, localized stage, it is most often diagnosed after the cancer has already begun to spread. Survival rates for lung cancer decrease significantly as it moves from localized cases to other areas. Therefore, early detection of lung cancer is crucial.

[0004] Currently, low-dose computed tomography (LDCT) is the recommended method for screening for lung cancer. However, this technique is associated with a high false-positive rate, resulting in many patients undergoing unnecessary follow-up evaluations. Because LDCT's low specificity can lead to a high false-positive rate, along with the high cost and patient radiation exposure associated with the technique, its poor sensitivity makes it an unattractive solution.

[0005] Numerous recent studies on breath biomarkers for lung cancer detection show the promise of this alternative screening method. Thousands of volatile organic compounds (VOCs) have been identified in human breath, at concentrations ranging from parts per million by volume (ppmv) to parts per trillion by volume (pptv). While there are reports that cancer cells influence the body's production of a certain set of VOCs that travel through the bloodstream to the lungs and leave the body in exhaled breath, there is no strong consensus on the identity of that specific set of VOCs.

[0006] The most widely used method for detecting VOCs in breath is mass spectroscopy (MS), commonly coupled with gas chromatography (GC) ("GC-MS"). GC-MS is popular due to its high selectivity and sensitivity in compound detection and high accuracy in compound identification. Two compounds are unlikely to behave similarly in both mass spectrometry and gas chromatography, making GC-MS a reliable option when certainty is important. The first use of GC-MS to analyze lung cancer breath biomarkers was in 1985. Several studies followed, and dozens of potential biomarkers have been identified en masse. Despite this advancement, GC-MS remains impractical for widespread clinical use and is limited to research applications due to its expensive, time-consuming, and complex sample collection procedures.

[0007] Another emerging technology for breath analysis is the electronic nose (E-nose), which uses an array of gas-sensing sensors to detect a diverse set of VOCs. This VOC detection method has several advantages over GC-MS, including relatively low cost, small size, speed, and the potential for use in "online" applications. Online breath analysis eliminates the need for sample storage and potentially provides more rapid feedback to clinicians. However, E-nose technology is far from ideal. E-nose technology requires frequent calibration, is sensitive to changes in humidity and temperature, and suffers from drift and memory effects. E-nose technology also tends to lack the characteristics that make GC-MS attractive, such as high sensitivity, high selectivity, and the ability to identify individual compounds.

[0008] Cavity ring-down spectroscopy (CRDS) is a commonly used technique for analyzing a single analyte in a gaseous sample using its absorption spectrum. A typical CRDS system uses a laser to generate a beam that is directed into the cavity of a chamber containing two highly reflective mirrors. The beam is tuned to a single wavelength, usually in the visible or near-infrared (NIR) spectrum. The beam is then repeatedly reflected between the mirrors, causing a portion of the light to exit the ring-down cavity.

[0009] To "fill" the ring-down cavity, its length must be tuned to the laser wavelength. This is typically done by adjusting the position of one of the two mirrors. When the laser is resonant with a cavity mode, constructive interference builds up the intensity in the cavity. As the light entering the cavity disappears and empties, the intensity of the light in the ring-down cavity decays by a certain percentage. A small amount of light is not reflected by the mirrors and exits the ring-down cavity. The intensity of the exiting light is measured by a sensor to determine the decay rate.

[0010] When a sample is placed in the ring-down cavity, analytes (e.g., volatile organic compounds) present in the sample absorb a portion of the light, thereby accelerating the decay of the light intensity within the ring-down cavity. An absorption spectrum is obtained by measuring the decay time of light at a particular wavelength in the presence of a sample relative to the decay time of light at that wavelength in the absence of a sample. Identification and quantification of individual analytes in a sample can be achieved in several ways, for example, by linear regression of the measured absorption spectrum of the gaseous sample with known absorption spectra of various analytes.

[0011] Although individual analytes are easily identifiable, the specific analytes that can be used to correctly detect the presence or absence of a particular physiological condition, such as lung cancer, can be unclear. Breath samples can be collected noninvasively from patients and analyzed by CRDS, but there is no universally accepted agreement on which analytes in breath are indicative of the presence or absence of lung cancer, and breath samples contain multiple confounding components that can make the detection of a single analyte difficult, so alternative approaches are needed. Summary of the Invention [Means for solving the problem]

[0012] In one aspect, a method for analyzing a sample using cavity ring-down spectroscopy is provided, comprising: introducing at least a portion of the sample into a ring-down cavity; for each wavelength of a set of wavelengths, generating a laser beam of that wavelength using at least one laser that is directed into the ring-down cavity; extinguishing the laser beam entering the ring-down cavity; and recording light intensity attenuation data of light exiting the ring-down cavity using a light intensity sensor system; and determining, using at least one processor, from the light intensity attenuation data for the set of wavelengths, a probability that a subject from whom the sample was drawn has a certain physiological condition or a degree of that physiological condition, using at least indirectly a dataset of light intensity attenuation data of pre-analyzed samples in which the presence or absence of the physiological condition or the degree of the physiological condition has been identified.

[0013] The determining may include using a predictive model based at least in part on the dataset of light intensity attenuation data.

[0014] The physiological condition can be lung cancer.

[0015] The sample and the pre-analysis sample may be a breath sample.

[0016] The set of wavelengths may be a first set of wavelengths, the sample may be introduced from a thermal desorption tube, and at least a portion of the sample may be a first portion of the sample desorbed from the thermal desorption tube heated to a first desorption temperature for introduction into the ring-down cavity, and the method may further comprise introducing a second portion of the sample desorbed from the thermal desorption tube heated to a second desorption temperature; for each wavelength in the second set of wavelengths, generating a laser beam of that wavelength using at least one laser that is directed into the ring-down cavity; extinguishing the laser beam entering the ring-down cavity; and recording light intensity attenuation data of the light exiting the ring-down cavity using a light intensity sensor system.

[0017] The second set of wavelengths may be equal to the first set of wavelengths.

[0018] The method may further include combining the light intensity attenuation data of the second portion of the sample with the light intensity attenuation data of the first portion of the sample for each wavelength in the first set of wavelengths.

[0019] The determining may be performed using a predictive model trained at least in part using a dataset of light intensity attenuation data of the pre-analysis samples.

[0020] The generating, the turning off, and the recording may occur until the control module determines that a desired level of light intensity attenuation data has been collected.

[0021] Another aspect provides a system for analyzing a sample using cavity ring-down spectroscopy, comprising: a ring-down cavity; at least one laser operative to generate a laser beam at each wavelength of a set of wavelengths and direct the laser beam into the ring-down cavity; a sample input system for inputting at least a portion of a sample into the ring-down cavity for analysis and removing at least a portion of the sample from the ring-down cavity; a light intensity sensor system positioned to record light intensity attenuation data of light exiting the ring-down cavity; at least one processor operatively coupled to the sample input system, the at least one laser, and the light intensity sensor system; and storage having computer readable instructions stored thereon, the computer readable instructions being executable by the at least one processor. and determining, from the light intensity attenuation data for the set of wavelengths, a probability that the subject from whom the sample was drawn has a certain physiological condition or a degree of the physiological condition, using at least indirectly a dataset of light intensity attenuation data of pre-analysis samples in which the presence or absence of the physiological condition or the degree of the physiological condition has been identified.

[0022] The at least one processor may determine the probability using a predictive model based at least in part on the dataset of light intensity attenuation data.

[0023] The physiological condition can be lung cancer.

[0024] The sample and the pre-analysis sample may be a breath sample.

[0025] The set of wavelengths may be a first set of wavelengths, the sample may be placed in a thermal desorption tube, and at least a portion of the sample may be a first portion of the sample; the at least one processor may control the sample loading system to desorb a first portion of the sample by heating the thermal desorption tube to a first desorption temperature and load the first portion of the sample into the ring-down cavity; the at least one processor may control the sample loading system to desorb a second portion of the sample by heating the thermal desorption tube to a second desorption temperature and load the second portion of the sample into the ring-down cavity after removing the first portion of the sample from the ring-down cavity; and the at least one processor may, for each wavelength in the second set of wavelengths, operate the at least one laser to generate a laser beam directed into the ring-down cavity, extinguish the laser beam entering the ring-down cavity, and record light intensity attenuation data of light exiting the ring-down cavity using the light intensity sensor system.

[0026] The second set of wavelengths may be equal to the first set of wavelengths.

[0027] The at least one processor may combine light intensity attenuation data for the second portion of the sample with light intensity attenuation data for the first portion of the sample for each wavelength in the first set of wavelengths.

[0028] The at least one processor may determine probabilities from the light intensity attenuation data using a predictive model trained at least in part using a dataset of light intensity attenuation data of the pre-analysis samples.

[0029] The at least one processor may repeat the activating, the deactivating, and the recording until the at least one processor determines that a desired level of light intensity attenuation data has been collected.

[0030] In a further aspect, a method for analyzing a sample using cavity ring-down spectroscopy is provided, comprising: heating a thermal desorption tube to a first desorption temperature to desorb a first portion of a sample contained in the thermal desorption tube; introducing the first portion of the sample into a ring-down cavity; for each wavelength in a first set of wavelengths, generating a laser beam of that wavelength using at least one laser directed into the ring-down cavity; extinguishing the laser beam incident on the ring-down cavity; and recording light intensity attenuation data of light exiting the ring-down cavity using a light intensity sensor system; and removing the first portion of the sample from the ring-down cavity. and removing the sample from the thermal desorption tube; heating the thermal desorption tube to a second desorption temperature to desorb a second portion of the sample contained in the thermal desorption tube; introducing the second portion of the sample into the ring-down cavity; for each wavelength of a second set of wavelengths, using the at least one laser to generate a laser beam of that wavelength directed into the ring-down cavity; extinguishing the laser beam entering the ring-down cavity; and recording light intensity attenuation data of light exiting the ring-down cavity using the light intensity sensor system; and analyzing the light intensity attenuation data for desorption at the first desorption temperature and the light intensity attenuation data for desorption at the second desorption temperature.

[0031] The first set of wavelengths may be equal to the second set of wavelengths.

[0032] The method may further include combining the light intensity attenuation data of the second portion of the sample with the light intensity attenuation data of the first portion of the sample for each wavelength in the first set of wavelengths.

[0033] The second desorption temperature can be higher than the first desorption temperature.

[0034] The first desorption temperature can be equal to the second desorption temperature.

[0035] The analyzing may include determining, from the light intensity attenuation data of the first portion of the sample and the light intensity attenuation data of the second portion of the sample, a probability that the subject from whom the sample was drawn has a physiological condition or a degree of physiological condition, at least indirectly using a dataset of light intensity attenuation data of a pre-analyzed sample in which the presence or absence of the physiological condition or the degree of the physiological condition has been identified.

[0036] The determining may be performed using a predictive model trained at least in part using a dataset of light intensity attenuation data of the pre-analysis samples.

[0037] In yet another aspect, a system for analyzing a sample using cavity ring-down spectroscopy is provided, comprising: a ring-down cavity; at least one laser operative to generate a laser beam directed into the ring-down cavity; a sample introduction system for introducing at least a portion of a sample into the ring-down cavity and removing at least a portion of the sample from the ring-down cavity for analysis, the sample introduction system including a heater configured to heat a thermal desorption tube containing the sample; a light intensity sensor system positioned to record light intensity attenuation data of light exiting the ring-down cavity; at least one processor operatively coupled to the sample introduction system, the at least one laser, and the light intensity sensor system; and storage having computer readable instructions stored thereon, which when executed by the at least one processor, are configured to heat the thermal desorption tube to a first desorption temperature to desorb a first portion of the sample contained in the thermal desorption tube and controlling the sample introduction system to introduce a first portion of the sample into the ring-down cavity; for each wavelength in a first set of wavelengths, operating the at least one laser to generate a laser beam of that wavelength directed into the ring-down cavity, extinguishing the laser beam incident on the ring-down cavity, and recording light intensity attenuation data of light exiting the ring-down cavity with the light intensity sensor system; removing the first portion of the sample from the ring-down cavity; heating the thermal desorption tube to a second desorption temperature to desorb a second portion of the sample contained in the thermal desorption tube; and controlling the sample introduction system to introduce the second portion of the sample into the ring-down cavity; for each wavelength in a second set of wavelengths, operating the at least one processor to generate a laser beam of that wavelength directed into the ring-down cavity, extinguishing the laser beam incident on the ring-down cavity;recording light intensity attenuation data of light exiting the ring-down cavity using the light intensity sensor system, and analyzing the light intensity attenuation data for desorption at the first desorption temperature and the light intensity attenuation data for desorption at the second desorption temperature.

[0038] The first set of wavelengths may be equal to the second set of wavelengths.

[0039] The computer-executable instructions, when executed by the at least one processor, may cause the at least one processor to combine light intensity attenuation data for the second portion of the sample with light intensity attenuation data for the first portion of the sample for each wavelength in the first set of wavelengths.

[0040] The second desorption temperature can be higher than the first desorption temperature.

[0041] The first desorption temperature can be equal to the second desorption temperature.

[0042] The computer-executable instructions, when executed by the at least one processor, may cause the at least one processor to determine, from the light intensity attenuation data of a first portion of the sample and the light intensity attenuation data of a second portion of the sample, a probability that the subject from whom the sample was drawn has a physiological condition or a degree of a physiological condition, at least indirectly using a dataset of light intensity attenuation data of a pre-analyzed sample in which the presence or absence of the physiological condition or the degree of the physiological condition has been identified.

[0043] The determining may be performed using a predictive model trained at least in part using a dataset of light intensity attenuation data of the pre-analysis samples.

[0044] In yet another aspect, a method for generating a predictive model for cavity ring-down spectroscopy analysis is provided, comprising the steps of: for each sample among a plurality of samples identified as having a physiological state or a degree of the physiological state and stored in at least two thermal desorption tubes, selecting, for each desorption temperature sequence among at least two mutually unique desorption temperature sequences, a previously unselected thermal desorption tube among at least two thermal desorption tubes containing the sample; heating, in turn, each desorption temperature in the desorption temperature sequence to the previously unselected thermal desorption tube among the at least two thermal desorption tubes to desorb a portion of the sample contained in the thermal desorption tube; and the method comprises: introducing a portion of the sample into a ring-down cavity; for each wavelength of a set of wavelengths, generating, using at least one laser, a laser beam of that wavelength that is directed into the ring-down cavity; extinguishing the laser beam incident on the ring-down cavity; and recording light intensity attenuation data at the wavelength for the portion of the sample; removing the portion of the sample from the ring-down cavity; and identifying which of the at least two mutually unique desorption temperature sequences for which the light intensity attenuation data has a greater correlation with the presence or absence of the physiological condition or the degree of the physiological condition identified in the plurality of samples.

[0045] Each desorption temperature in at least one of the at least two mutually unique desorption temperature sequences may be higher than the previous desorption temperature in the at least one of the at least two mutually unique desorption temperature sequences.

[0046] In another aspect, a system for generating a predictive model for cavity ring-down spectroscopy analysis is provided, comprising: a ring-down cavity; at least one laser operative to generate a laser beam directed into the ring-down cavity; a sample introduction system for introducing at least a portion of a sample into the ring-down cavity and removing at least a portion of the sample from the ring-down cavity for analysis, the sample introduction system including a heater configured to heat a thermal desorption tube containing the sample; a light intensity sensor system positioned to record light intensity attenuation data of light exiting the ring-down cavity; at least one processor operatively coupled to the sample introduction system, the at least one laser, and the light intensity sensor system; and storage having computer readable instructions stored thereon, which, when executed by the at least one processor, generate a predictive model for at least two thermal desorption tubes identified as having a physiological condition or a degree of the physiological condition. For each sample among a plurality of samples each stored in a desorption tube, for each desorption temperature sequence among at least two mutually unique desorption temperature sequences, selecting a previously unselected thermal desorption tube among at least two thermal desorption tubes containing the sample, heating the previously unselected thermal desorption tube among the at least two thermal desorption tubes to the desorption temperature in turn for each desorption temperature in the desorption temperature sequence to desorb a portion of the sample contained in the thermal desorption tube, introducing the portion of the sample into the ring-down cavity, for each wavelength among a set of wavelengths, generating a laser beam of the wavelength using at least one laser that is directed into the ring-down cavity, extinguishing the laser beam incident on the ring-down cavity, and recording light intensity attenuation data at the wavelength of the portion of the sample, removing the portion of the sample from the ring-down cavity, and determining whether the light intensity attenuation data for each desorption temperature sequence among the at least two mutually unique desorption temperature sequences is:and determining whether the plurality of samples has a greater correlation with the presence or absence of the identified physiological condition or the degree of the identified physiological condition.

[0047] Each desorption temperature in at least one of the at least two mutually unique desorption temperature sequences may be higher than the previous desorption temperature in the at least one of the at least two mutually unique desorption temperature sequences.

[0048] In a further aspect, a method for analyzing a sample using cavity ring-down spectroscopy is provided, comprising: heating a thermal desorption tube to a first desorption temperature to desorb a first portion of a sample contained in the thermal desorption tube; releasing the first portion of the sample without performing cavity ring-down spectroscopy on the first portion of the sample; heating the thermal desorption tube to a second desorption temperature different from the first desorption temperature to desorb a second portion of the sample contained in the thermal desorption tube; introducing the second portion of the sample into a ring-down cavity; for each wavelength in a second set of wavelengths, using at least one laser to generate a laser beam of that wavelength directed into the ring-down cavity; extinguishing the laser beam entering the ring-down cavity; and recording light intensity decay data of light exiting the ring-down cavity using a light intensity sensor system; and analyzing the light intensity decay data.

[0049] The analyzing may include determining from the light intensity attenuation data of the second portion of the sample a probability that the subject from whom the sample was drawn has a certain physiological condition or a degree of that physiological condition, at least indirectly using a dataset of light intensity attenuation data of a pre-analyzed sample in which the presence or absence of the physiological condition or the degree of the physiological condition has been identified.

[0050] In a further aspect, a system for analyzing a sample using cavity ring-down spectroscopy is provided, comprising: a ring-down cavity; at least one laser operative to generate a laser beam directed into the ring-down cavity; a sample introduction system for introducing at least a portion of a sample into the ring-down cavity for analysis and removing at least a portion of the sample from the ring-down cavity, the sample introduction system including a heater configured to heat a thermal desorption tube containing the sample; a light intensity sensor system positioned to record light intensity attenuation data of light exiting the ring-down cavity; at least one processor operatively coupled to the sample introduction system, the at least one laser, and the light intensity sensor system; and storage having computer readable instructions stored thereon, the computer readable instructions being configured to detect and record at least a portion of the at least one processor. When executed by the processor, the method causes the at least one processor to: heat a thermal desorption tube to a first desorption temperature to desorb a first portion of a sample contained in the thermal desorption tube; release the first portion of the sample without performing cavity ring-down spectroscopy on the first portion of the sample; heat the thermal desorption tube to a second desorption temperature different from the first desorption temperature to desorb a second portion of the sample contained in the thermal desorption tube; introduce the second portion of the sample into the ring-down cavity; for each wavelength in a second set of wavelengths, generate a laser beam of that wavelength using at least one laser that is directed into the ring-down cavity; extinguish the laser beam entering the ring-down cavity; and record light intensity decay data of light exiting the ring-down cavity using a light intensity sensor system; and analyze the light intensity decay data.

[0051] The computer readable instructions, when executed by the at least one processor, may cause the at least one processor to determine from the light intensity attenuation data of the second portion of the sample a probability that the subject from whom the sample was drawn has a certain physiological condition or a degree of that physiological condition, at least indirectly using a dataset of light intensity attenuation data of a pre-analyzed sample in which the presence or absence of the physiological condition or the degree of the physiological condition has been identified.

[0052] Other technical advantages may become apparent to one skilled in the art upon reading the following figures and description.

[0053] For a better understanding of embodiments of the present disclosure, and to show more clearly how the same may be carried into effect, reference will now be made, by way of example only, to the accompanying drawings, in which: [Brief explanation of the drawings]

[0054] [Figure 1] FIG. 1 is a schematic diagram of various optical and pneumatic components of a CRDS system according to one embodiment. [Figure 2] FIG. 2 is a cross-sectional view of a thermal desorption tube used to introduce a breath sample into the ring-down cavity of the CRDS of FIG. 1. [Figure 3] FIG. 2 is a schematic diagram of an electrical control system for controlling the various optical and pneumatic components of the CRDS system shown in FIG. 1. [Figure 4] 2 illustrates a general method for analyzing a sample using the CRDS system of FIG. 1 according to one embodiment. [Figure 5] 2 illustrates a method for generating a predictive model using the CRDS system of FIG. 1 according to one embodiment. [Figure 6] 10 illustrates a method for analyzing light intensity decay data of a sample using the CRDS system of FIG. 1 according to another embodiment. [Figure 7] Figure 7A illustrates a computing device in communication with a CRDS system in a CRDS analysis system according to another embodiment, and Figure 7B is a schematic diagram of the computing device of Figure 7A. DETAILED DESCRIPTION OF THE INVENTION

[0055] Unless otherwise indicated, objects shown in the drawings are not necessarily drawn to scale.

[0056] For simplicity and clarity of illustration, reference numerals may be repeated throughout the figures, where appropriate, to designate corresponding or analogous elements. Additionally, numerous specific details are provided to provide a thorough understanding of embodiments of the present disclosure. However, it will be understood by those skilled in the art that embodiments of the present disclosure may be practiced without such specific details. In some instances, well-known methods, procedures, and components have not been described in detail so as not to obscure embodiments of the present disclosure. While exemplary embodiments are initially shown in the drawings and described below, it should be understood that the principles of the present disclosure can be implemented in numerous ways, both presently known and unknown. The present disclosure is not limited to the exemplary embodiments and ways shown in the drawings and described below.

[0057] Various terms used throughout this disclosure should be read and understood as follows, unless otherwise indicated: "Or" as used throughout this application is inclusive as if described as "and / or." "A" and "an" as used throughout this application include the plural and vice versa. Similarly, references to a gender include the opposite gender, and the use, implementation, or performance of the description herein should not be understood as being limited to a single gender. "Illustrative" should be understood as "illustrative" or "example," and not necessarily "preferred" over other embodiments. Further definitions of terms may be provided herein, and as will be understood by reading this disclosure, such terms may apply to both their preceding and following cases.

[0058] Modifications, additions, and omissions may be made to the systems, devices, and methods of the present disclosure without departing from the scope of the present disclosure. For example, system or device components may be integrated or separated. Furthermore, the operations of the systems and devices of the present disclosure may be achieved by more, fewer, or other components, and the methods of the present disclosure may include more, fewer, or different steps. Furthermore, steps may be performed in any suitable order. In this disclosure, "each" refers to each member of a set or each member of a subset of a set.

[0059] Any module, unit, server, computer, terminal, engine, or device exemplified in this disclosure as executing instructions may include or have access to a computer-readable medium, such as a storage medium, computer storage medium, or data storage device (e.g., magnetic disk, optical disk, or tape). Computer storage media includes volatile, nonvolatile, removable, and non-removable media implemented in any method or technology for storage of information, such as computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include RAM, ROM, EEPROM, flash memory, other memory technologies, CD-ROM, DVD (Digital Versatile Disk), other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage, other magnetic storage devices, or any other medium usable to store desired information and accessible by an application, module, or both. Such computer storage media may be part of, accessible to, or connectable to a device. Additionally, unless otherwise specified, a processor or controller in this disclosure may be implemented as a single processor or as multiple processors. The multiple processors may be arrayed or distributed, and any processing function referred to in this disclosure may be performed by one of the multiple processors, but may also be implemented as a single processor. Any method, application, or module of this disclosure may be implemented using computer-readable / executable instructions that may be stored or carried on such computer-readable media and that may be executed by one or more processors.

[0060] The following disclosure provides a novel approach to performing analysis using CRDS. Unlike traditional approaches that compare light intensity decay data with that of a single known analyte, light intensity decay data recorded for a sample is analyzed at least indirectly using a dataset of light intensity decay data for a set of previously analyzed samples that have been identified as having a certain physiological state or degree of that physiological state. In a preferred embodiment, this is done with a predictive model generated using machine learning on a set of training data. Additionally, the process of gradually desorbing the sample from the thermal desorption tube can be used to provide additional information about the sample. Furthermore, the predictive model can be improved by using machine learning to determine which combinations of desorption temperature and / or wavelength provide light intensity decay data that have a greater correlation with the physiological state or degree of that physiological state for which multiple samples have been identified.

[0061] A physiological state is any type of condition that a subject may be in. Physiological states include disease, infection, the functional state of one or more organs or body tissues, etc. The degree of a physiological state can include severity, stage (e.g., cancer stage), growth size, gland secretion, etc. The degree of a physiological state can sometimes be binary, representing the presence or absence of the physiological state in a subject. The degree can be discrete or continuous.

[0062] The various components of a CRDS system 20 for analyzing a sample according to certain embodiments are shown in FIG. 1. A CO laser 24 and a carbon-13 ( 13 C) An O2 laser 28 is provided. The CO2 laser 24 and the Carbon-13O2 laser 28 are gas tube lasers that emit light at a series of quasi-equally spaced known frequencies that can be rapidly selected using an adjustable diffraction grating device. Gas tube laser technology has a long history and is a stable and robust method for generating infrared light at precisely known frequencies. Both the CO2 laser 24 and the Carbon-13O2 laser 28 emit light in the mid-infrared spectrum.

[0063] Each of the CO2 laser 24 and the Carbon-13O2 laser 28 has an actuator to change the angle of the grating after the cavity, along with an output coupler to allow adjustment of the laser cavity length, thereby changing its pitch to adjust which wavelengths are reflected. By both changing the laser cavity length and changing the grating angle, the laser can be tuned very precisely to a specific wavelength and desired mode quality.

[0064] The CO2 laser 24 generates the first laser beam 32, and the carbon-13O2 laser 28 generates the second laser beam 36. Depending on the desired optical frequency, either the CO2 laser 24 is tuned to generate the first laser beam 32 while the carbon-13O2 laser 28 is detuned, or the carbon-13O2 laser 28 is tuned to generate the second laser beam 36 while the CO2 laser 24 is detuned. In this way, at most, only one of the CO2 laser 24 and the carbon-13O2 laser 28 outputs a beam at any given time, preventing simultaneous combination of the first beam 32 and the second beam 36. Mid-infrared, and particularly long-wavelength infrared, was chosen as the type of light because most volatile organic compounds absorb light within this range. As a result, multiple volatile organic compounds can be measured with a single system. The CO2 laser operates in this range and has sufficient power and narrow linewidth for ring-down spectroscopy. Using two lasers adds to the range and number of available wavelengths that the CRDS system 20 can use to analyze samples.

[0065] First laser beam 32 is redirected via mirror 40 on an optical mount to beam splitter 44. Beam splitter 44 is partially reflective and partially transmissive and splits each of first laser beam 32 and second laser beam 36 into two beams: a sampling beam 48 and a working beam 52, which may have the same characteristics and be of a similar intensity as sampling beam 48.

[0066] The sampling beam 48 is received by a high-speed infrared detector 56, which measures the amplitude and beat frequency of the sampling beam 48 using an oscilloscope. The beat frequency can indicate the presence of higher-order modes due to less than optimal tuning of the CO laser 24 and the Carbon-13O laser 28. In response to detecting an undesired beat frequency, the corresponding laser 24 or 28 is tuned until the beat frequency amplitude is minimized or eliminated while the intensity is maximized. If the beat frequency amplitude cannot be reduced below an acceptable level, the laser can be tuned to another wavelength.

[0067] The working beam 52 travels to a first optical modulator 60, which deflects the working beam 52 onto a mirror 64 on an optical mount. The mirror 64 redirects the light to a second optical modulator 68, which in turn deflects the working beam 52 onto a focusing lens 72. Optical modulators are used to control the intensity of the light beam generated by the laser. In this embodiment, the first optical modulator 60 and the second optical modulator 68 are acousto-optic modulators (AOMs), also known as Bragg cells. An AOM is a type of optical modulator that uses a piezoelectric transducer coupled to a material such as germanium or glass. In the disclosed embodiment, the material is germanium. When an oscillating electrical signal is applied to the piezoelectric transducer, it vibrates, generating acoustic waves in the material. The acoustic waves stretch the material, causing periodic variations in the refractive index, enabling Bragg diffraction. Light incident on an AOM at the first Bragg angle relative to a plane perpendicular to the axis of propagation of the sound wave is deflected by an amount equal to twice the Bragg angle at maximum efficiency. With the electrical signal removed, the Bragg diffractive properties of the material disappear, allowing the light to pass through undeviated, effectively attenuating the light along the deflected path. A side effect of an AOM is that the frequency of the light being deflected is shifted.

[0068] In other embodiments, the optical modulator can instead be an electro-optical modulator. An electro-optical modulator is another type of optical modulator that applies a DC (direct current) or low-frequency electric field to a material, perturbing the position, orientation, and / or shape of the material's molecules. As a result, the refractive index is altered and the phase of the output beam changes as a function of the applied field. By passing the beam through a polarizer, the phase modulation is converted to intensity modulation. Alternatively, a phase modulator placed in one branch of the interferometer can act as an intensity modulator.

[0069] Also, although the CRDS system 20 is described as having two optical modulators, in other embodiments the CRDS system may have fewer or more optical modulators.

[0070] The first and second optical modulators 60 and 68 act as attenuators, adjusting the intensity of the working beam 52 and extinguishing it at the beginning of a ring-down event. A ring-down event can involve extinguishing the working beam 52 from illuminating the ring-down cavity, detuning the laser relative to the ring-down chamber, and collecting optical intensity data from the ring-down chamber. In an AOM, the first and second optical modulators 60 and 68 use the acousto-optic effect to diffract light with acoustic waves (usually radio-frequency acoustic waves). In each of the first and second optical modulators, a piezoelectric transducer is coupled to a material, such as germanium or glass, and an oscillating electrical signal is used to vibrate the piezoelectric transducer. The vibrating piezoelectric transducer generates acoustic waves in the material, which expand and contract the material, causing periodic variations in the refractive index and enabling Bragg diffraction. Light incident on the AOM at a Bragg angle relative to a plane perpendicular to the propagation axis of the acoustic waves is deflected by an amount equal to twice the Bragg angle for maximum efficiency. The loss of the electrical signal eliminates the Bragg diffractive properties of the material, allowing the light to pass through undeviatedly, effectively eliminating the light along the deflected path. Thus, the intensity of sound can be used to modulate the intensity of light in a deflected beam.

[0071] The intensity of the light deflected by each of the first and second optical modulators 60, 68 can be between approximately 85% of the input light intensity (representing the maximum deflection efficiency of the optical modulators 60, 68) and the attenuation limit of each of the first and second optical modulators 60, 68, of approximately 0.1%. When the acoustic wave applied to the germanium is turned off, the deflected beam loses approximately 30 dB, or 99.9%, of its previous intensity. The attenuation limit represents the upper limit of how much the input light intensity can be attenuated by the optical modulator.

[0072] The optical modulator is asymmetric, meaning that, as a side effect, it Doppler shifts the frequency of the light in a first mode when it receives input light at its first end, and Doppler shifts the frequency of the light in a second mode opposite to the first mode when it receives input light at its second end, but with the same attenuation power. The Doppler shift in the frequency of the light is along the same direction regardless of whether the light is incident on the first end or the second end.

[0073] Conventional CRDS systems use a single optical modulator, resulting in a frequency-shifted working beam. While this frequency shift is typically small relative to the frequency of the light and can alter how the light is absorbed by objects within the cavity, this frequency shift can be compensated for during analysis. When diffraction occurs toward the AOM's acoustic source, the frequency shift is a downward shift; when diffraction occurs away from the acoustic source, the frequency shift is an upward shift. As noted above, this effect is minimal.

[0074] Working beam 52, deflected by second optical modulator 68, is focused via focusing lens 72. The laser beam, and thus working beam 52, continues to diverge as it propagates from CO laser 24 or Carbon-13O laser 28. Focusing lens 72 focuses working beam 52 back.

[0075] A mirror 76 on the optics mount then redirects the working beam 52 into a ring-down chamber 80. The two mirrors 64, 76 extend the path length of the working beam 52.

[0076] Ring-down chamber 80 is an elongated tube that defines a resonant cavity therein, referred to as ring-down cavity 84. Front and back cavity mirrors 88a and 88b (alternatively referred to collectively as cavity mirrors 88) are positioned at the longitudinal ends of ring-down cavity 84. Cavity mirror 88 is highly reflective to both light directed at it from outside ring-down cavity 84 and light directed at it from within ring-down cavity 84. As a result, a portion (approximately 0.1%) of working beam 52 directed at front cavity mirror 88a passes through front cavity mirror 88a and enters ring-down cavity 84, while a majority (approximately 99.9%) of the working beam is reflected back toward mirror 76.

[0077] The cavity mirrors 88 are mounted on mirror mounts 92 that are operable to adjust the position and orientation of the cavity mirrors 88. In particular, the front cavity mirror 88a facing the front of the ring-down cavity 84 is mounted on an actuable mirror mount 92 via three mechanical micrometers 96a. The back cavity mirror 88b facing the rear of the ring-down cavity 84 is mounted on an actuable mirror mount 92 via three piezoelectric micrometers 96b, which can be manually adjusted for optical alignment or further adjusted using piezoelectric drivers.

[0078] The angle of each cavity mirror 88 can be changed to ensure that the mirrors are sufficiently aligned so that the light beam does not deviate as it enters the ring-down cavity 84. If one cavity mirror 88 is angled, in particular, some light will be reflected to the side of the ring-down cavity 84, resulting in a loss of light intensity and higher-order modes. Micrometer 96 can also be simultaneously tuned to change the length of the ring-down cavity 84 without affecting the angular alignment. This allows tuning of the ring-down cavity 84 so that it resonates at the frequency of the light entering the ring-down cavity 84.

[0079] Focusing lens 72 focuses the laser light to match the optical mode of ring-down cavity 84 so that the minimum waist of the beam is located at the same location as the minimum beam waist of ring-down cavity 84. The position of focusing lens 72 can be adjusted to match the optical mode of a range of laser wavelengths.

[0080] A light intensity sensor system in the form of a liquid nitrogen-cooled detector 100 is positioned behind the rear cavity mirror 88b to receive the light exiting that mirror. The liquid nitrogen-cooled detector 100 measures the intensity of the light exiting the ring-down cavity 84. Other types of light intensity sensor systems for measuring the intensity of the exiting light can be used in place of the liquid nitrogen-cooled detector 100.

[0081] A thermal desorption tube 104 is used to collect the sample for testing and deposits the sample into the ring-down cavity 84 .

[0082] Referring now to FIG. 2 , an exemplary thermal desorption tube 104 is shown. The thermal desorption tube 104 has a tubular stainless steel case 105 defining openings 106 at both ends. A receiving end 107 of the thermal desorption tube 104 receives a sample. In the illustrated exemplary embodiment, the sample is human exhaled breath collected from a human subject for testing. A foam separator 108 is positioned toward the receiving end and configured to distribute fluid pressure more evenly across the cross-section of the stainless steel case 105. A sorbent material 109 is positioned adjacent to the foam separator 108 and another foam separator 110. The separators may alternatively be made of wire mesh or other suitable materials. The sorbent material 109 is highly porous, has a relatively large surface area, and is selected for compound-specific sampling to capture and retain compounds of interest even in the presence of other compounds. Furthermore, the sorbent material 109 allows the collected compounds to be easily desorbed or extracted for analysis. Additionally, the selected solid sorbent material is non-reactive with the sample. In a specific example, the solid adsorbent is Tenax or a carbonaceous material. As the sample is received at receiving end 107, the sample becomes more concentrated toward receiving end 107 of thermal desorption tube 104. In other embodiments, the composition and configuration of the thermal desorption tube may be different, as will be appreciated by those skilled in the art.

[0083] During sample collection, exhaled breath is collected by a breath collection device fitted with a thermal desorption tube 104. The breath collection device can be relatively small, such as a wearable mask, or relatively large, such as a tabletop unit into which a person breathes. The breath sample is received at receiving end 107 of thermal desorption tube 104. The various small and large molecules contained in the person's breath are then captured by sorbent material 109 in thermal desorption tube 104. In other embodiments, other types of gaseous samples can be collected by the thermal desorption tube.

[0084] Referring again to FIG. 1 , a sample loading system 112 is used to load a sample from the thermal desorption tube 104 into the ring-down cavity 84, and the sample loading system 112, including the ring-down cavity 84, is evacuated. During sample loading, the sample loading system 112 fills the ring-down cavity 84 with at least a portion of the collected sample (i.e., desorbs the sample from the thermal desorption tube 104 and places the sample in the ring-down cavity 84 without introducing contaminants), brings the pressure in the ring-down cavity to 1 atmosphere, brings the temperature to 50° C., and seals the ring-down cavity 84. In this embodiment, a set of sample absorption spectra are determined at this pressure and temperature to be at least indirectly compared to the measured absorption spectra, ensuring consistency between parameters that may affect the results. However, in other embodiments, the pressure and temperature may be fixed at other levels of the known measured absorption spectra. During evacuation of at least a portion of the sample, the sample input system 112 removes the previously provided sample from the ring-down cavity 84 and from the various conduits for directing the sample from the thermal desorption tube 104 to the ring-down cavity 84.

[0085] The sample loading system 112 has an intake section that includes a nitrogen gas source 116. The nitrogen gas source 116 can be a pressurized source of ultra-clean nitrogen gas or can be capable of pressurizing nitrogen gas to 1 atmosphere or greater. In this embodiment, the nitrogen gas source 116 is pressurized to 5 psi above atmospheric pressure, but this can be varied as long as the compression is sufficient to pressurize the ring-down cavity 84 to 1 atmosphere or another atmospheric pressure selected for the analysis. In the illustrated embodiment, the nitrogen gas source 116 is nitrogen gas evaporated from a liquid nitrogen container. The nitrogen gas source 116 is connected via conduit 120 to a gas inlet valve 124a. An auxiliary gas inlet valve 124b allows for the connection of other gases but is not typically used. The gas inlet valve 124a and the auxiliary gas inlet valve 124b communicate with the gas inlet line 120a. Pressure gauges 128 are located along the gas inlet line 120a and the gas inlet valve 124c. A filter 130a is positioned along gas inlet line 120a ahead of cavity inlet valve 124d, which seals gas inlet line 120a from ring-down cavity 84. Filter 130a prevents contaminants from entering ring-down cavity 84, where they can deposit on cavity mirror 88 and interfere with reflectivity.

[0086] The gas inlet valve 124a and auxiliary gas inlet valve 124b communicate with a pathing valve 124e, which enables or disables direct access to the desorption tube line 120b and the sample outlet line 120c.

[0087] The desorption tube line 120b includes a forward valve 124f and a rearward valve 124g. The thermal desorption tube 104 is disposed between the forward valve 124f and the rearward valve 124g, with the receiving end 107 of the thermal desorption tube 104 facing the rearward valve 124g. The thermal desorption tube 104 is disposed within a heater 132 that is controllable to heat the thermal desorption tube 104 to a range of temperatures.

[0088] Sample outlet line 120 c includes sample outlet valve 124 h and mass flow controller 136 .

[0089] The sample loading system 112 also has an outlet portion including a cavity outlet valve 124i in communication with the ring-down cavity 84. An outlet line 140 is in communication with the cavity outlet valve 124i. A pressure gauge 144 is disposed along the outlet line 140. A vacuum shut-off valve 124j is disposed between the pressure gauge 144 and a vacuum pump 148. A vacuum inlet valve 124k is in communication with the vacuum pump 148 and draws air through a pump inlet line 150. A filter 130b is disposed in the pump inlet line 150 to prevent the introduction of contaminants that may interfere with the operation of the vacuum pump 148.

[0090] Valves 124a-124k may alternatively be referred to collectively as valves 124.

[0091] While cavity inlet valve 124d and cavity outlet valve 124i are shown as being suitably coupled to ring-down cavity 84 at particular locations, it should be understood that the locations at which valves 124d, 124i are coupled to ring-down cavity 84 may vary. In a preferred configuration, cavity inlet valve 124d communicates with ring-down cavity 84 toward the end of ring-down cavity 84 adjacent front cavity mirror 88a, and cavity outlet valve 124i communicates with ring-down cavity 84 toward the end of ring-down cavity 84 adjacent rear cavity mirror 88b.

[0092] When a new sample is to be introduced into the ring-down cavity 84, the thermal desorption tube 104 containing the new sample is coupled to the sample introduction system 112 as shown in FIG.

[0093] During the evacuation phase, vacuum inlet valve 124k is opened and vacuum pump 148 is turned on. Then, vacuum inlet valve 124k is closed, and vacuum shutoff valve 124j, cavity outlet valve 124i, cavity inlet valve 124d, gas inlet line valve 124c, and flow path control valve 124e are opened in sequence. The contents of the lines and ring-down cavity 84 along this flow path are evacuated from the CRDS system by vacuum pump 48. Pressure gauge 144, particularly when pressure gauge 122 is isolated from vacuum pump 148, allows for determination of whether the system is sufficiently evacuated. Once it is determined that the system is sufficiently evacuated, the open valves 124j, 124i, 124d, 124c, and 124e are closed in reverse order. Then, during the nitrogen fill phase, valves 124a, 124c, 124d, 124i, and 124j are opened to allow nitrogen gas from nitrogen gas source 116 to fill line 120a and line 140. The nitrogen gas is then evacuated in another evacuation step. The nitrogen filling and evacuation steps can be repeated as desired to clear the lines. In this manner, the previously tested sample is evacuated from the CRDS system 20.

[0094] During the introduction of a new sample, the thermal desorption tube 104 is flushed to remove carbon dioxide and water from the thermal desorption tube 104 and minimize the amount of carbon dioxide and water introduced into the ring-down cavity 104. To flush the thermal desorption tube 104, the gas inlet valve 124a, the gas inlet line valve 124c, and the back valve 124 are opened to provide a path for nitrogen gas to flush the thermal desorption tube 104. The thermal desorption tube 104 is selected to prevent carbon dioxide and water from collecting with the sample, although some carbon dioxide and water will typically still be present within the thermal desorption tube 104.

[0095] 500 ml of nitrogen gas is introduced into the thermal desorption tube 104 to remove any remaining carbon dioxide and water from the original sample. Then, the forward valve 124f and the sample outlet valve 124h are opened to provide a path to the mass flow controller 136. The mass flow controller 136 allows the nitrogen gas and the entrained carbon dioxide and water to be released at a specific flow rate. In this configuration, the flow rate is 500 ml / min. Then, all valves 124 are closed.

[0096] Once the carbon dioxide and water have been removed from the thermal desorption tube 104, the sample loading system 112 is again evacuated using the same process described above to remove the nitrogen gas that was just introduced into the lines of the sample loading system 112. A heater 132 surrounding the thermal desorption tube 104 then heats the thermal desorption tube 104 to a desired temperature to thermally desorb at least a portion of the new sample within the thermal desorption tube 104. The gas inlet valve 124a, flow control valve 124e, front valve 124f, rear valve 124g, and cavity inlet valve 124d are then opened to provide a direct flow path for nitrogen gas from the nitrogen gas source 116 through the thermal desorption tube 104 with the desorbed target compounds and toward the ring-down cavity 84.

[0097] One atmosphere of pressure within ring-down cavity 84 is preferred because all reference data collected and analyzed is at this pressure level to ensure that results are reproducible.

[0098] Gas inlet valve 124a is toggled open and closed by the system, which then waits for the pressure reading on pressure gauge 128 to stabilize and reach 1 atmosphere. If pressure gauge 128 stabilizes but the pressure reading is still less than 1 atmosphere, gas inlet valve 124a is toggled again and the process is repeated until the pressure reading reaches 1 atmosphere. When pressure gauge 128 indicates a pressure level of 1 atmosphere in ring-down cavity 84, all valves are closed.

[0099] If desorption at multiple temperature values ​​is desired, vacuum pump 148 is turned on and cavity outlet valve 124i and vacuum shutoff valve 124j are opened to evacuate ring-down cavity 84. Cavity outlet valve 124i is then closed before repeating the desorption process.

[0100] Complete evacuation is generally not performed between multiple desorptions because there is some sample between rear valve 124g and cavity inlet valve 124d that may be lost.

[0101] In this manner, by pressurizing a fixed volume ring-down cavity containing a sample to a desired pressure level, the surface area within the ring-down cavity to which compounds can adhere can be reduced compared to a variable volume ring-down cavity that can be used to increase the pressure within the cavity to a desired level.

[0102] Additionally, the pressure gauge 128 is located upstream of the sample flow path from the thermal desorption tube 104 to the ring-down cavity 84 to prevent contamination by the sample.

[0103] 3 is a schematic diagram of an electronic control subsystem 200 for various components of an exemplary CRDS system 20. All lines represent electrical or electronic signals, arrows represent one-way communication, voltage settings, etc., and lines without arrows represent two-way communication.

[0104] A control module 204 including one or more processors acts as a computer system that controls the functions of the various components shown in Figure 1. The control module 204 has one or more processors 205 and storage 206 that stores computer-executable instructions that, when executed by the processor 205, cause the processor to direct the other components of the CRDS system 20 of the present disclosure. The control module 204 can be any type of component that includes at least one processor that executes computer-executable instructions to control the operation of the other components of the CRDS system 20 of the present disclosure.

[0105] A pair of RF drivers 208 transmits approximately 40 MHz signals that power the CO2 laser 24 and the Carbon-13O2 laser 28. Each laser 24 and laser 28 is tuned using an output coupler and a diffraction grating. A grating actuator 212 actuates (rotates) the diffraction grating. Another actuator actuates (translates) the output coupler. Each output coupler is driven by a 1000 V output coupler piezoelectric 216. A two-channel high-voltage amplifier 220 powering the output coupler piezoelectric 216 is adjustable between 0 V and 1000 V. The high-voltage amplifier 220 is configured with an analog output signal from a data acquisition ("DAQ") card 224 in the control module 204. The DAQ generates an output between 0 V and 10 V, and the high-voltage amplifier 220 amplifies the signal by 100 times to generate a 0 V to 1000 V signal that powers the output coupler piezoelectric 216. Each grating actuator 212, which changes the angle of the grating, is driven by an actuator driver 228 that receives commands from the control unit 204 via RS-232. Each grating actuator 212 moves in millimeters, which translates into pitch angles for the lasers 24, 28.

[0106] Data signals from the pressure gauges 128, 144 of the sample injection system 112 are received via RS-232.

[0107] A small amplifier 232 and oscilloscope 236, to which a high speed infrared detector 56 is connected, can be used to read the amplitude and frequency of the beat signal used to tune the lasers 24,28.

[0108] The temperature controller 240 for the thermal desorption tube heater 132 is controlled by the control module 204 via RS-232. The tube heater 132 includes a strip of aluminum wrapped with heating tape and a temperature sensor. Both the heating tape and the temperature sensor are connected to the temperature controller 240, which is a PID (proportional integral derivative) controller. The PID controller sets the temperature and provides a readback to the main control module 204 via RS-232.

[0109] A relay board 244 is connected to the control module 204 and is used to turn all solenoid valves 124 and vacuum pumps 148 on and off.

[0110] A three-channel piezoelectric driver 248 drives a piezoelectric actuator 252 that actuates the micrometer 96 to adjust the length of the ring-down cavity 84. Each channel has two components: communication to the piezoelectric driver via RS-232 and an analog input from the DAQ card 224. In other embodiments, more than two piezoelectric drivers may be employed.

[0111] Each optical modulator 60, 68 is driven by an RF driver 256 transmitting an approximately 40 MHz signal. Changing the frequency of the RF driver 256 changes the Bragg angle for a given optical wavelength or changes the optical wavelength to which a given or fixed Bragg angle is tuned. When the RF driver 256 is tuned to a particular frequency and set to full power, it passes most of the working beam 52 (approximately 85%). Adjusting it to 80% or 70% attenuates the optical modulators 60, 68. Setting the RF driver 256 to zero turns the optical modulators 60, 68 completely off. The frequency of the RF driver 256 is set by components via RS232. Analog and digital components can set the on / off conditions and amplitude of the RF driver 256. In particular, the DAQ card 224 sends signals via a digital output (DO) to a timing circuit 260, which generates the four necessary signals needed to enable and set the amplitude of the RF driver. The timing circuit 260 also communicates with the control module 204 via a digitizer 264, allowing the control module 204 to control its operating state to either a ring-down initiation condition or a steady-state condition in which the timing circuit 260 sets the four voltage values ​​to zero and then returns to the previous voltage levels after a predetermined time. The digitizer 264 sends an initiation pulse to the timing circuit 260, causing the optical modulators 60, 68 to extinguish the laser light provided to the ring-down cavity 84 during a ring-down event.

[0112] A method 300 for analyzing a sample using the CRDS system 20 of FIG. 1 is shown in FIG. 4. Traditionally, samples analyzed using CRDS are desorbed at a single temperature. In this method 300, a portion of the sample in the thermal desorption tube may be first desorbed at a first temperature, and then at least one additional portion of the sample may be desorbed from the thermal desorption tube at at least one additional temperature. In some cases, it may be desirable to separate portions of the sample for analysis with the CRDS system 20. For example, acetone and ammonia may be desorbed from the thermal desorption tube at a low temperature. Because acetone and ammonia may dominate the spectrum and obfuscate other substances in the sample, it may be desirable to desorb and analyze portions of the sample at different temperatures to isolate various substances from the other substances. In other embodiments, it may be desirable to desorb the sample more than once at the same temperature because some components of the sample may be more susceptible to desorption at that temperature the first time, while other components of the sample may be less susceptible to desorption at that temperature the first time but equally susceptible the second time. Other components may be more readily desorbed and therefore less present the second time, allowing the other components to be more isolated and separated. The ordered set of desorption temperatures may include a single desorption temperature or may be an ordered set of two or more desorption temperatures.

[0113] 1, 3, and 4, the method 300 begins by setting the temperature of the heater 132 to a first desorption temperature of an ordered set of desorption temperatures (310). This particular configuration may sequence one or more sets of temperatures at which the sample is desorbed from the thermal desorption tube 104. In one preferred approach, the desorption temperatures are ordered from lowest to highest. In other embodiments, the ordered set of desorption temperatures may include two of the same or similar desorption temperatures. A portion of the sample may be desorbed and analyzed at one temperature to release easily desorbed components / compounds, and then another portion of the sample may be desorbed at the same temperature to release less easily desorbed components / compounds for analysis, reducing the amount of easily desorbed components / compounds.

[0114] A first desorption temperature is selected from the set of ordered desorption temperatures, and heater 132 is activated to heat thermal desorption tube 104 to the first desorption temperature. Once thermal desorption tube 104 has been heated to the first desorption temperature by heater 132, a portion of the sample desorbed at the first temperature is introduced into ring-down cavity 84 (320), as described above. Upon introducing the portion of the sample into ring-down cavity 84, a laser beam is generated by one of lasers 24 and 28 tuned to one of a set of wavelengths (330). The set of wavelengths can be selected from among the wavelengths at which lasers 24 and 28 can generate a desired beam. The generated laser beam is directed through first optical modulator 60, reflected by mirror 64, through second optical modulation region 68, and reflected by mirror 72 into ring-down chamber 80. Optical modulators 60 and 68 attenuate working beam 52 and modulate its intensity to some extent.

[0115] When working beam 52 reaches front cavity mirror 88a, a portion of it (approximately 0.1%) passes through front cavity mirror 88a and enters ring-down cavity 84. The majority of working beam 52 (approximately 99.9%) is initially reflected back along the same path to whichever laser 24 or 28 is operating.

[0116] Initially, the ring-down cavity 84 is not illuminated. As light enters the ring-down cavity 84, most of the light in the ring-down cavity 84 is reflected between the two cavity mirrors 88, so that as more light is introduced from the outside via the working beam 52, the amount or power of light in the ring-down cavity 84 begins to increase. A certain percentage of the light leaks out of the cavity mirrors 88. It takes a period of time for the ring-down cavity 84 to "fill" with light, which can occur when the cavity length is equal to the adjacent resonance length of the ring-down cavity 84 for the tuned laser. At that point, the incident and leaking light are in equilibrium.

[0117] Once this equilibrium is achieved, the laser beam directed into the ring-down cavity 84 is extinguished (340). The laser beam incident on the ring-down cavity 84 can be extinguished in several ways. In certain embodiments, the digitizer 264 sends an initiation pulse to the timing circuit 260, causing the optical modulators 60, 68 to extinguish the laser light being provided to the ring-down cavity 84. In other embodiments, the laser is turned off or detuned so that it is not resonant with the configured cavity length, etc.

[0118] The timing circuit 260 simultaneously commands the first light modulator 60 and the second light modulator 68 to attenuate the light beam at or near the attenuation limits of the light modulators 60, 68, reducing the intensity of the light beam from the first light modulator 60. In the CRDS system 20, by commanding both light modulators 60, 68 to shut off simultaneously, the amount of light deflected by the first light modulator 60 in a short period of time is significantly reduced by the second light modulator 68 being shut down.

[0119] The second optical modulator 68 significantly enhances the attenuation achieved by the first optical modulator 60 alone. In the embodiment described herein, the first optical modulator 60 provides 30 dB of attenuation, and the second optical modulator 68 provides an additional 30 dB of attenuation, resulting in a total attenuation of 60 dB achieved by the optical modulators 60, 68. While the ring-down cavity 84 is filled with light, the optical modulators 60, 68 attenuate the working beam 52 and modulate its intensity. In this configuration, each optical modulator 60, 68 attenuates the working beam 52 by 5 dB, resulting in a total attenuation of 10 dB. As a result, each optical modulator 60, 68 can attenuate the working beam 52 by an additional 25 dB for a total attenuation of 50 dB during the extinction of the working beam 52. In a conventional configuration, one optical modulator must attenuate the working beam by 10 dB, leaving an additional 20 dB of attenuation to extinguish the working beam. As can be seen, the working beam 52 can be attenuated much more quickly with an additional 50 dB of attenuation from two optical modulators 60, 68 than with an additional 20 dB of attenuation from one optical modulator. As a result, the amount of additional light introduced into the ring-down cavity 84 after commanding the optical modulators 60, 68 to shut down is less than the additional light introduced in a single optical modulator setup of a conventional CRDS system. By dissipating the working beam 52 more quickly, the measured attenuation of light in the ring-down cavity 84 is less affected by the additional light during the fall times of the optical modulators 60, 68, allowing for greater accuracy in matching the observed decay times to known decay times.

[0120] The loss of laser light provided to the ring-down cavity 84 initiates a ring-down event. In alternative embodiments, the loss of resonant laser light provided to the ring-down cavity 84 can be accomplished in other ways, such as by detuning the laser. By using a threshold to initiate the ring-down event, the ring-down event can be timed to occur during the peak where the ring-down cavity 84 is resonating with the laser light, rather than to one side of the peak. Furthermore, because the bandwidth of the resonance is approximately 10 millivolts, the resolution of the piezoelectric driver 248 is insufficiently fine to adequately track the peak.

[0121] During the ring-down event, the control module 204 records (350) the optical intensity decay data of the light exiting the back end of the ring-down cavity 84 as reported by the liquid nitrogen-cooled detector 100. In this configuration, the ring-down event lasts approximately 10 microseconds, but may last longer or shorter in other embodiments. The optical decay time is approximately 2 microseconds.

[0122] It is then determined whether the ring-down event should be repeated 360. In this embodiment, the CRDS system 20 is configured to collect light intensity decay data for 500 ring-down events.

[0123] Approximately 100 microseconds after initiating the ring-down event, the timing circuit 260 instructs the optical modulators 60, 68 to resume passing the working beam 52 through the ring-down cavity 84. Once light intensity attenuation data has been acquired from 500 ring-down events, the control module 204 deactivates the piezoelectric driver 248 and determines the attenuation rate from the ring-down event data. Ring-down event data may also be referred to interchangeably herein as light intensity attenuation data. Alternatively, if it is determined in 360 that more light intensity attenuation data should be collected, the control module 204 continues to instruct the piezoelectric driver 248 to operate the rear cavity mirror 88b. Once the total voltage reaches a maximum or minimum, it begins to travel in the reverse direction. That is, if the total voltage AV increased before reaching the maximum voltage in its range, then the total voltage AV decreases, returning to the voltage corresponding to the resonance length RL. Alternatively, if the total voltage AV decreased before reaching the minimum voltage in its range, then the total voltage increases, returning to the voltage corresponding to the resonance length RL. In this way, a ring-down event is triggered in both directions.

[0124] The light intensity attenuation data is collected as quickly as possible because the various outputs may drift. For example, the piezoelectric actuator 252 may have a settling time called piezoelectric creep.

[0125] In other embodiments, any number of ring-down events can be used at each wavelength and desorption temperature. In one embodiment, ring-down events can be performed at a wavelength and desorption temperature until a desired level of light intensity decay data is collected. The light intensity decay data collected so far can then be quickly analyzed as soon as it is collected to determine whether it is valid. After valid light intensity decay data has been collected for the desired number of ring-down events, the system can determine that sufficient data has been collected at that wavelength and desorption temperature. This can often reduce the time required to perform the analysis.

[0126] Upon collecting light intensity decay data for a selected wavelength, the light intensity decay data for a particular sample, desorption temperature, and wavelength can be processed by control module 204 to, for example, determine the decay rate for each ring-down event and average these values. Other techniques for processing ring-down event data are also possible.

[0127] It is then determined whether there are other wavelengths in the set that the laser should generate 370. If there are other wavelengths that should be generated, the method 300 returns to 330 to tune the lasers 24, 28 to generate laser beams at the newly selected wavelengths.

[0128] The process is repeated for other wavelengths of light to produce an absorption spectrum of the sample at a particular desorption temperature. The light intensity data from the liquid nitrogen-cooled detector 100 effectively provides, or can be easily converted into, light intensity decay data, which indicates how quickly light of a particular wavelength is absorbed by the portion of the sample introduced into the ring-down cavity 84.

[0129] For example, light produced by a CO laser 24 provides an absorption coefficient over a range of wavelengths. Similarly, absorption coefficients can be generated over a range of wavelengths for light from a Carbon-13O laser 28. In this manner, an absorption spectrum can be developed for the sample.

[0130] Once an absorption spectrum across a set of wavelengths has been generated for the portion of the sample desorbed at the first temperature, the portion of the sample is removed from the ring-down cavity (375). It is determined (380) whether there are other temperatures in the ordered set to which the thermal desorption tube 140 should be heated to desorb other portions of the sample into the ring-down cavity 84 for analysis. If it is determined that there are other temperatures to which the portion of the sample should be desorbed, then the heater is set to the next desorption temperature in 310. In this embodiment, the next desorption temperature to which the thermal desorption tube 104 is heated is higher than the previous desorption temperature to which the thermal desorption tube 104 was heated. If the ordered set of desorption temperatures has a single desorption temperature, then the method 300 does not return to 310.

[0131] As described above, desorbing the sample contained in the thermal desorption tube 104 at a temperature higher than the temperature of the thermal desorption tube 104 at which it was previously desorbed will result in a different subset of compounds / substances in the sample.

[0132] Once light intensity attenuation data is collected at each desorption temperature, the CRDS system 20 analyzes the light intensity attenuation data for that set of wavelengths and desorption temperatures to determine 390 the probability that the patient from whom the sample was drawn has a particular physiological condition or the degree of that pathological condition. The probability is determined at least indirectly using a dataset of light intensity attenuation data for pre-analysis samples for which the presence or absence of the physiological condition (e.g., lung cancer in the exemplary embodiment) or the degree of the physiological condition has been identified. The probability determination may use a predictive model based at least in part on the dataset of light intensity attenuation data for pre-analysis samples.

[0133] The CRDS system 20 can determine the probability that the subject from whom the sample was drawn has one of a set of discrete degrees of that physiological state, representing the most probable degree of that physiological state, or representing the complete set of probabilities. Additionally, the CRDS system 20 can provide a confidence level for each result.

[0134] The light intensity decay data collected at each desorption temperature may be analyzed, at least indirectly, using a dataset of light intensity decay data for pre-analysis samples grouped by the same or similar desorption temperature. Preferably, the dataset of light intensity decay data for pre-analysis samples was collected using the same or similar temperature sequence to desorb portions of the sample. In an alternative embodiment, light intensity decay data from two or more desorption temperatures for a sample may be combined and analyzed using predictive data generated at least in part based on light intensity decay data for pre-analysis samples that have been treated the same or similarly (i.e., light intensity decay data for corresponding desorption temperatures grouped together). Predictive models may be generated from the dataset of light intensity decay data for pre-analysis samples. One preferred approach is through the use of supervised machine learning.

[0135] The dataset of light intensity attenuation data for the pre-analysis sample represents the spectrum for the entire portion of the previously collected sample, not for any one particular volatile organic compound. By adopting this more holistic approach to analyzing samples, the probability that the currently analyzed sample is from a subject with that physiological state is less reliant on the identity of the volatile organic compound under discussion. The subject may be a human or other animal or organism.

[0136] The predictive model is generated using machine learning techniques by at least partially using the dataset of light intensity attenuation data as training data.

[0137] Obtaining a large enough sample data set to predict the presence or absence of a particular physiological condition can be difficult. In these situations, some preprocessing steps can be useful. This is especially true in the case of certain physiological conditions.

[0138] To address missing values ​​in the spectrum, interpolation methods (such as spline interpolation) can be used that use nearby absorption values ​​in the spectrum. If the available wavelengths are not evenly spaced, the average value from comparable subjects identified by k-nearest neighbors can be used, as in VOC regression, to approximate the missing values.

[0139] If large variations are observed in previously collected data, this may be due to natural VOC variation and other confounding variables. Baseline correction can be used to remove undesired trends, for example, using common spectroscopic detrending methods (e.g., linear detrending, quadratic detrending, Savitzky-Golay detrending). Difference spectroscopy can also be used for high-concentration VOCs, such as acetone and ammonia, which are common to all targets and can obscure small, important details in the spectra.

[0140] Similarly, general normalization methods can be used to improve sample consistency. In addition to those based on individual spectra (e.g., max-min, peak, or standard normal normalization), methods that normalize across the entire spectrum (e.g., multiplicative scatter correction) can be employed.

[0141] A comprehensive analysis of spectral and clinical factors is used to identify nontrivial relationships and confounding factors in the light intensity attenuation data dataset for pre-analysis samples to improve data interpretation and identify subgroups that are incorporated into predictive models. Many confounding factors exist in breath VOC analysis, including age, sex, smoking, alcohol consumption, medications, and other diseases. These factors are assessed using unsupervised clustering methods, along with supervised predictive models that classify spectra according to potential confounders. Similarly, spectral data for different subgroups can be used to train individual models for lung cancer detection.

[0142] Spectral-derived features, projection-based features, and wavelet-transform-based features have been employed for feature extraction. Derived features are particularly useful because they provide a unique baseline and scale correction for interindividual variability. Barcode-based features and one-dimensional forms of local binary patterns (1D-LBP) can also be used. Additionally, features derived from the desorption time series of samples at different temperature sequences can also be used.

[0143] A variety of preprocessing and feature extraction methods can be combined with information found in confounding factor and subgroup analyses to generate a final classification scheme. Several feature selection methods can be used to reduce the dimensionality of the scheme and remove useless and redundant features. Statistical filter methods, such as minimum redundancy maximum relevance (mRMR) selection, and classification wrapper methods, such as sequential feature selection (SFS) and genetic algorithm (GA) selection, can be used. Simple support vector machines (SVM), linear or quadratic discriminant analysis (LDA or QDA), and k-nearest neighbor (KNN) classifiers can be used to evaluate the performance of wrapper selection methods and build and validate a large number of classification models.

[0144] A variety of alternatives and methods may be used to generate the predictive model.

[0145] FIG. 5 illustrates a method 400 for generating a predictive model. In this approach, a set of desorption temperature sequences is identified for search. The desorption temperature sequence is an ordered set of temperatures to which the thermal desorption tube is heated to desorb a portion of the sample contained therein. Desorbing the sample using a specific sequence of temperatures allows different selected portions of the sample to be desorbed and analyzed separately, which allows different aspects of the sample to be more easily identified. Specific characteristics of the sample may be more readily apparent from light intensity decay data captured with a specific desorption temperature sequence. For example, if a first component / compound is desorbable from the thermal desorption tube 104 at a particular temperature and is confused with a second component / compound that is desorbable at a slightly higher temperature, a desorption temperature sequence having a desorption temperature equal to the desorption temperature of the first component / compound and then equal to or greater than the desorption temperature of the second component / compound may more readily reveal characteristics of the light intensity decay data of the sample. The relationship between such desorption temperature sequences and the features revealed may not be readily apparent in some cases. As a result, the particular sequence that provides the best correlation to presence or absence may not be readily apparent without exploration. In other aspects, if light intensity attenuation data recorded at two desorption temperature sequences provides substantially equal correlation for a subset of samples taken from subjects having a physiological state or a degree of that physiological state, it may be desirable to select, for example, the shorter desorption temperature sequence.

[0146] Although a general method is described in connection with the CRDS system 20, it should be understood that the method can be performed with multiple local or distributed CRDS systems and other computers using a common set of parameters and standard techniques.

[0147] The method 400 begins with sample selection (404). The sample is selected from a set of samples from subjects who have been identified as having or not having a physiological state. A desorption temperature sequence is then selected (408) from a set of desorption temperature sequences. The set of desorption temperature sequences can be selected based on scientific evidence, randomly selected, etc. The desorption temperature of the heater 132 is set (412) to the next desorption temperature in the sequence. If the sequence remains unprocessed, the first desorption temperature in the sequence is selected.

[0148] A portion of the sample is then desorbed and loaded 416. The heater 132 is heated to a desorption temperature selected from the sequence to desorb the portion of the sample. The sample loading system 112 loads the desorbed portion of the sample into the ring-down cavity 84.

[0149] Control module 204 then activates one of lasers 24 and 28 to generate a laser beam of the selected wavelength, which is directed into ring-down cavity 84, causing it to "fill" with light.

[0150] If the control module 204 determines that there is sufficient light present in the ring-down cavity 84, the control module 204 instructs the light modulator to extinguish the laser beam directed into the ring-down cavity 84 (424).

[0151] The liquid nitrogen-cooled detector 100 records (428) the light intensity of the light exiting the ring-down cavity 84 via the rear cavity mirror 88b. The light detection attenuation data, corresponding to the detected intensity of the light exiting the ring-down cavity 84 over time, is transmitted to the control module 204. Upon completion of the ring-down, the control module 204 determines (430) whether to repeat the ring-down event. The ring-down event may be repeated multiple times to reduce the effects of anomalies. If another ring-down event is to be repeated at the selected wavelength, the method 400 returns to (420) to generate a laser beam at the selected wavelength. It then determines (432) whether there are other wavelengths in the set to use. If there are other wavelengths, the method 400 returns to (420) to generate a laser beam at the next selected wavelength. If, instead, it is determined that there are no other wavelengths in the set, it determines (436) whether there are other desorption temperatures in the sequence. If there is at least one other desorption temperature in the series, the next desorption temperature in the series is selected and heater 132 is heated to that temperature to desorb another portion of the sample (412).

[0152] Alternatively, if it is determined that there are no other desorption temperatures in the sequence (436), it is determined whether there are other desorption temperature sequences to be searched (440). If there are other unsearched desorption temperature sequences, the next desorption temperature sequence is selected (408). If a new desorption temperature sequence is selected, a new thermal desorption tube from the same analyte is used because the sample in the previously used thermal desorption tube 104 has at least partially desorbed.

[0153] Once all desorption temperature sequences have been searched, it is determined whether there are any more samples in the set to analyze (444). If there are any unanalyzed samples in the set, another sample is selected for analysis (404).

[0154] Alternatively, once it is determined that all samples have been analyzed (444), a correspondence between the light intensity attenuation data and the identification of the subject from which the sample was obtained as having a certain physiological condition or a degree of that physiological condition is determined for each desorption temperature sequence (448). The correspondence between the light intensity attenuation data and the identification of the sample as having a certain physiological condition or a degree of that physiological condition is stronger if there is a set of characteristics that are prominent in samples from subjects identified as having that physiological condition or a degree of that physiological condition. That is, if there is a particular set of characteristics that occurs in 85 percent of the subjects with a certain physiological condition and in 20 percent of the subjects without that physiological condition for a particular desorption temperature sequence, then the light intensity attenuation data for that desorption temperature sequence may have a relatively good correlation to the presence or absence of that physiological condition in the subjects from whom the samples were obtained. This is just one example, but any pattern between the light intensity attenuation data and the presence of a physiological condition or the degree (degree) of a physiological condition may indicate the presence of a correlation between the light intensity attenuation data and the presence of a physiological condition or the degree (degree) of a physiological condition.

[0155] A desorption temperature sequence is then selected 452 based on the correspondence between the light intensity attenuation data for each desorption temperature sequence and the identification of the analyte from which the sample was drawn as having the physiological state or degree of that physiological state. The desorption temperature sequence may be selected entirely based on the measured correlation, or may be selected based on a combination of the correlation and other factors. Other factors may determine the selection of the desorption temperature sequence, such as the number of desorption temperatures in the sequence, which may determine the time to perform the analysis.

[0156] A predictive model is then generated (452) using the selected desorption temperature sequence based at least in part on the correlation.

[0157] When a sequence of temperatures is used to desorb a sample from a thermal desorption tube, it may be inconvenient to analyze the portion of the sample that desorbs at a particular temperature: desorption at a particular temperature may be performed to remove less interesting materials than desorption at other temperatures.

[0158] 6 illustrates a method 500 for analyzing a sample with a CRDS. In method 500, a portion of the sample is desorbed at a first temperature and discarded because it may have little correlation to the presence or absence of a physiological condition in a patient. A second portion of the sample is desorbed at a second temperature. The second portion of the sample is deemed to have a relatively high correlation to the presence or absence of a physiological condition or the degree of that physiological condition and is analyzed with the CRDS.

[0159] Method 500 begins by setting 510 the temperature of heater 132 to a first desorption temperature by control module 204. Heating heater 132 to the first desorption temperature heats thermal desorption tube 104 to desorb a first portion of the sample. Control module 204 then instructs sample injection system 112 to release the first portion of the sample without performing CRDS on the first portion. The first portion of the sample may be released to ambient atmosphere, released to a collection vessel, or removed from the path between thermal desorption tube 104 and ring-down cavity 84 in any other suitable manner. In a currently preferred embodiment, the first sample is released through mass flow controller 136 by closing valves 124b and 124d and opening valves 124a, 124c, 124g, 124f, and 124h in much the same manner as for releasing carbon dioxide and water from thermal desorption tube 104.

[0160] After releasing the first portion of the sample, the heater is heated to a second desorption temperature (530). The second desorption temperature is generally higher than the first desorption temperature, resulting in the desorption of additional materials from the thermal desorption tube 104. In other embodiments, the second desorption temperature is equal to or lower than the first desorption temperature. While a significant portion of some components / compounds may have already desorbed at the previous desorption temperature, substantially all other contaminating materials may also have desorbed, facilitating the isolation of these compounds / components. A second portion of the sample is then introduced into the ring-down cavity 84 for analysis (540), as described above. The control module 204 then activates one of the lasers 24 and 28 to generate a laser beam at a selected wavelength (550). The laser beam is directed into the ring-down cavity 84, filling it with light. Once the ring-down cavity 84 is filled, the laser beam is extinguished (560). Liquid nitrogen cooled detector 100 then records 570 the light intensity of the light exiting ring-down cavity 84 through rear cavity mirror 88b. The light exiting ring-down cavity 84 represents light intensity decay data that is recorded by liquid nitrogen cooled detector 100 and sent to control module 204.

[0161] Once the ring-down event is complete, a determination is made as to whether to repeat the ring-down event again (580). As described above, it may be desirable to determine one or more aggregate criteria by repeating the ring-down event multiple times, determining the decay rate of the light intensity decay data for each ring-down, and averaging the decay rates to minimize the effects of anomalies. If another ring-down event is to be repeated, a laser beam is generated at the selected wavelength (550). Alternatively, if no more ring-down events are to be performed, a determination is made as to whether there are other wavelengths in the set of wavelengths to analyze (590). If it is determined that there are other wavelengths to analyze, a laser beam at the next wavelength is generated (550). Alternatively, if there are no more wavelengths at which to generate a laser beam to perform CRDS, the light intensity decay data collected in 570 is analyzed (595), after which method 500 ends. The collected light intensity decay data is then analyzed.

[0162] Although the at least one processor that analyzes the light intensity attenuation data and generates the predictive model is shown and described as being fully integrated with the other elements of the CRDS system, in other embodiments, one or more of the at least one processor may be provided in a separate computer system that has storage and performs some or all of the functions of the control module of the above-described exemplary embodiment.

[0163] FIG. 7A illustrates a system 600 for analyzing samples using CRDS in another embodiment. The system 600 includes a CRDS system 604 similar to that shown in FIG. 1. The CRDS system 604 has a control module 608 having at least one processor for controlling the operation of the various elements of the CRDS system 604. The control module 608 communicates with a computer system 612 over a data communications network 616. The computer system 612 can be a single computer or multiple computers coupled to each other locally and / or remotely to provide the desired functionality. The data communications network 616 can be any suitable medium that allows the control module 608 and the computer system 612 to exchange at least light intensity attenuation data and associated parameters, such as wavelength, the temperature at which at least a portion of the sample desorbs, and other analyte-specific data, such as confounding factors as described above, analyte identifiers, time and location information regarding when and where the sample was analyzed, and the like.

[0164] FIG. 7B illustrates several physical and logical components of computer system 612, including a central processing unit (CPU) 624, random access memory (RAM) 628, input / output (I / O) interface 632, communication interface 636, non-volatile storage 640, and a local bus 644 that allows CPU 624 to communicate with other components. CPU 624 includes one or more processors and is capable of executing at least an operating system and a spectrum evaluation application. RAM 628 provides relatively fast, volatile storage for CPU 624. I / O interface 632 allows input from one or more devices, such as a keyboard or mouse, to be received and information to be output to output devices, such as a display and / or speakers. Communication interface 636 allows communication with other communication devices, such as control module 608, over a computer network, such as data communications network 616. Non-volatile storage 640 stores an operating system and programs, including computer-executable instructions for executing the spectrum evaluation application. During operation of the computer system 612, the operating system, programs and data may be retrieved from the non-volatile storage 640 and placed in the RAM 623 to facilitate execution.

[0165] While the light sources in the above-described embodiment are two lasers producing light in the mid-infrared range, it should be understood that other light sources can be used, such as lasers producing light in the visible spectrum or near-infrared lasers. Furthermore, in some cases, a CRDS system can include more than two lasers, or even a single laser, to generate the working beam.

[0166] Instead of an acousto-optic modulator, an electro-optic modulator can be used.

[0167] The acousto-optic modulator can be configured to upshift or downshift the frequency of the working beam. The net frequency shift achieved by the acousto-optic modulator can shift the frequency of the working beam significantly away from the frequency of the working beam that the laser is generating, minimizing the amount of interference between the reflected light and the generated working beam, as long as the reflected light is outside the bandwidth of the laser light being generated.

[0168] In other embodiments, more than two optical modulators can be used in a CRDS system to provide additional extinction capability to more quickly extinguish the working beam at the onset of a ring-down event. Additionally, in additional embodiments, a single optical modulator can be used.

[0169] In other embodiments, one or more focusing lenses are used, which can be translated to reposition the lenses to allow mode matching for each wavelength of the laser.

[0170] The same approach is applicable to other types of resonant cavities, especially optical resonant cavities.

[0171] Other types of events can be triggered by adjusting the cavity length to be near the resonant wavelength of the cavity for a particular wavelength of choice.

[0172] In other embodiments, the sample can be analyzed at pressure levels other than 1 atmosphere, and the width of the absorption spectrum may vary accordingly.

[0173] While specific advantages have been enumerated above, various embodiments may include some, all, or none of the enumerated advantages.

[0174] In the above embodiment, the resonant cavity is a ring-down cavity, but in other embodiments, other types of resonant cavities can be used.

[0175] Those skilled in the art will recognize that further alternatives and modifications are possible, and that the above examples are merely illustrative of one or more embodiments, the scope of which is therefore to be limited only by the appended claims. [Explanation of symbols]

[0176] 20 CRDS System 24 CO2 laser 28 Carbon-13O2 laser 32 First laser beam 36 Second laser beam 40 Mirror 44 Beam Splitter 48 Sampling beam 52 Output Beam 56 High-speed infrared detector 60 First Optical Modulator 64 Mirror 68 Second Optical Modulator 72 focusing lens 76 Mirror 80 Ring-down chamber 84 Ring-down cavity 88 Cavity Mirror 88a Front cavity mirror 88b Rear cavity mirror 92 mirror mount 96 micrometers 96a Mechanical Micrometer 96b Piezoelectric Micrometer 100 Liquid nitrogen cooled detector 104 Thermal desorption tube 105 Stainless Steel Case 106 Aperture 107 Receiving end 108 Foam separator 109 Adsorbent 110 Foam separator 112 Sample injection system 116 Nitrogen gas source 120 Conduit 120a gas intake line 120b Desorption pipe line 120c sample outlet line 124 Solenoid valve 124a Gas inlet valve 124b Auxiliary gas inlet valve 124c Gas Intake Line Valve 124d Cavity Inlet Valve 124e Flow path control valve 124f forward valve 124g rear valve 124h cavity outlet valve 124i Sample Outlet Valve 124j Vacuum shutoff valve 124k vacuum intake valve 128 Pressure Gauge 130a filter 130b filter 132 Heater 136 Mass flow controller 140 Exit Line 144 Pressure Gauge 148 Vacuum Pump 150 Pump suction line 200 Electronic Control Subsystem 204 Control Module 205 processors 206 Storage 208 RF Driver 212 Lattice Actuator 216 Output coupler piezoelectric 220 High Voltage Amplifier 224 DAQ cards 228 Actuator Driver 232 Amplifier 236 Oscilloscope 240 Temperature Controller 244 Relay Board 248 3-channel piezoelectric driver 252 Piezoelectric Actuator 256 RF Driver 260 Timing Circuit 264 digitizer RL Resonant Length 300 ways 310 Select and set the temperature 320 At least part of the sample is input 330 Generates a laser beam at a selected wavelength 340 Laser beam disappears 350 Record light intensity attenuation data 360 Repeat ring-down events? 370 Are there other wavelengths? 375 Take a portion of the sample 380 Are there any other temperatures in the sequence? 390 Determine the Probability of Physiological States 400 ways Select 404 Sample 408 Select temperature sequence 412 Set heater temperature 416 Add a portion of the sample Generates laser beams at 420 wavelengths 424 Laser beam disappears 428 Record light intensity attenuation data 430 Repeat ring-down events? 432 Are there other wavelengths? 436 Are there other temperatures? 436 Determine the Probability of Physiological States 440 Are there other temperature sequences? 444 Are there any other samples? 448 Identifying highly correlated temperature sequences 452 Select the temperature sequence Generate 456 predictive models 500 ways 510 Set heater to first temperature 520 Release first portion of sample 530 Set heater to second temperature 540 Add second portion of sample 550 Generates a laser beam at a selected wavelength 560 Laser beam disappears 570 Record light intensity attenuation data 580 Repeat ring-down events? 590 Are there other wavelengths? 595 Analyzing the collected light intensity attenuation data 600 System 604 CRDS System 608 Control Module 612 Computer Systems 616 Data Communication Network 624 processor 628 RAM 632 I / O interface 636 Communication Interface 640 Non-Volatile Storage 642 Bus 648 databases

Claims

1. 1. A method for analyzing a sample using cavity ring-down spectroscopy, comprising: heating the thermal desorption tube to a first desorption temperature to desorb a first portion of the sample contained in the thermal desorption tube; depositing a first portion of the sample into a ring-down cavity; for each wavelength in a first set of wavelengths, generating a laser beam of that wavelength using at least one laser that is directed into the ring-down cavity, extinguishing the laser beam incident on the ring-down cavity, and recording optical intensity attenuation data of light exiting the ring-down cavity using an optical intensity sensor system; Removing a first portion of the sample from the ring-down cavity; heating the thermal desorption tube to a second desorption temperature to desorb a second portion of the sample contained in the thermal desorption tube; depositing a second portion of the sample into the ring-down cavity; for each wavelength in the second set of wavelengths, generating a laser beam of that wavelength using the at least one laser that is directed into the ring-down cavity, extinguishing the laser beam incident on the ring-down cavity, and recording light intensity attenuation data of light exiting the ring-down cavity using the light intensity sensor system; analyzing light intensity decay data for desorption at the first desorption temperature and light intensity decay data for desorption at the second desorption temperature.

2. The method of claim 1 , wherein the first set of wavelengths is equal to the second set of wavelengths.

3. The method of claim 2 , further comprising combining the light intensity attenuation data of the second portion of the sample with the light intensity attenuation data of the first portion of the sample.

4. The method of claim 1 , wherein the second desorption temperature is greater than the first desorption temperature.

5. The method of claim 1 , wherein the first desorption temperature is equal to the second desorption temperature.

6. 2. The method of claim 1, wherein the analyzing step comprises determining a probability that the subject from whom the sample was drawn has a particular physiological condition or a degree of a particular physiological condition from the light intensity attenuation data of the first portion of the sample and the light intensity attenuation data of the second portion of the sample, at least indirectly using a dataset of light intensity attenuation data of a pre-analyzed sample in which the presence or absence of the particular physiological condition or the degree of the particular physiological condition has been identified.

7. 7. The method of claim 6, wherein said determining is performed using a predictive model trained at least in part using a dataset of light intensity attenuation data of said pre-analysis samples.

8. 1. A system for analyzing a sample using cavity ring-down spectroscopy, comprising: A ring-down cavity; at least one laser operative to generate a laser beam directed into the ring-down cavity; a sample loading system for loading at least a portion of a sample into the ring-down cavity and removing at least a portion of the sample from the ring-down cavity for analysis, the sample loading system including a heater configured to heat a thermal desorption tube containing the sample; a light intensity sensor system positioned to record light intensity attenuation data of light exiting the ring-down cavity; at least one processor operatively coupled to the sample input system, the at least one laser, and the light intensity sensor system; and storage storing computer-readable instructions that, when executed by the at least one processor, heating the thermal desorption tube to a first desorption temperature to desorb a first portion of a sample contained in the thermal desorption tube and controlling the sample injection system to inject the first portion of the sample into the ring-down cavity; for each wavelength in a first set of wavelengths, operating the at least one laser to generate a laser beam of that wavelength that is directed into the ring-down cavity, extinguishing the laser beam incident on the ring-down cavity, and recording light intensity attenuation data of light exiting the ring-down cavity with the light intensity sensor system; removing a first portion of the sample from the ring-down cavity; heating the thermal desorption tube to a second desorption temperature to desorb a second portion of the sample contained in the thermal desorption tube and controlling the sample injection system to inject the second portion of the sample into the ring-down cavity; for each wavelength in a second set of wavelengths, operating the at least one processor to generate a laser beam of that wavelength directed into the ring-down cavity, extinguishing the laser beam incident on the ring-down cavity, and recording light intensity attenuation data of light exiting the ring-down cavity with the light intensity sensor system; analyzing light intensity decay data for desorption at the first desorption temperature and light intensity decay data for desorption at the second desorption temperature.

9. The system of claim 8 , wherein the first set of wavelengths is equal to the second set of wavelengths.

10. 10. The system of claim 9, wherein the computer-readable instructions, when executed by the at least one processor, cause the at least one processor to combine light intensity attenuation data for the second portion of the sample with light intensity attenuation data for the first portion of the sample.

11. The system of claim 8 , wherein the second desorption temperature is greater than the first desorption temperature.

12. The system of claim 8 , wherein the first desorption temperature is equal to the second desorption temperature.

13. 9. The system of claim 8, wherein the computer readable instructions, when executed by the at least one processor, cause the at least one processor to determine from the light intensity attenuation data of the first portion of the sample and the light intensity attenuation data of the second portion of the sample a probability that the subject from whom the sample was drawn has a particular physiological condition or a degree of a particular physiological condition, at least indirectly using a dataset of light intensity attenuation data of a pre-analyzed sample in which the presence or absence of the particular physiological condition or the degree of the particular physiological condition has been identified.

14. 14. The system of claim 13, wherein said determining is performed using a predictive model trained at least in part using a dataset of light intensity attenuation data of said pre-analysis samples.

15. 1. A method for analyzing a sample using cavity ring-down spectroscopy, comprising: heating the thermal desorption tube to a first desorption temperature to desorb a first portion of the sample contained in the thermal desorption tube; ejecting a first portion of the sample without performing cavity ring-down spectroscopy on the first portion of the sample; heating the thermal desorption tube to a second desorption temperature different from the first desorption temperature to desorb a second portion of the sample contained in the thermal desorption tube; depositing a second portion of the sample into a ring-down cavity; for each wavelength in the second set of wavelengths, generating a laser beam of that wavelength using at least one laser that is directed into the ring-down cavity, extinguishing the laser beam incident on the ring-down cavity, and recording optical intensity attenuation data of light exiting the ring-down cavity using an optical intensity sensor system; analyzing the light intensity attenuation data.

16. 16. The method of claim 15, wherein said analyzing comprises determining from the light intensity attenuation data of the second portion of the sample a probability that the subject from whom the sample was drawn has a particular physiological state or a degree of a particular physiological state, at least indirectly using a dataset of light intensity attenuation data of a pre-analyzed sample in which the presence or absence of the particular physiological state or the degree of the particular physiological state has been identified.

17. 1. A system for analyzing a sample using cavity ring-down spectroscopy, comprising: A ring-down cavity; at least one laser operative to generate a laser beam directed into the ring-down cavity; a sample loading system for loading at least a portion of a sample into the ring-down cavity and removing at least a portion of the sample from the ring-down cavity for analysis, the sample loading system including a heater configured to heat a thermal desorption tube containing the sample; a light intensity sensor system positioned to record light intensity attenuation data of light exiting the ring-down cavity; at least one processor operatively coupled to the sample input system, the at least one laser, and the light intensity sensor system; and storage storing computer-readable instructions that, when executed by the at least one processor, heating the thermal desorption tube to a first desorption temperature to desorb a first portion of the sample contained in the thermal desorption tube; ejecting a first portion of the sample without performing cavity ring-down spectroscopy on the first portion of the sample; heating the thermal desorption tube to a second desorption temperature different from the first desorption temperature to desorb a second portion of the sample contained in the thermal desorption tube; depositing a second portion of the sample into the ring-down cavity; for each wavelength in the second set of wavelengths, generating a laser beam of that wavelength using at least one laser that is directed into the ring-down cavity, extinguishing the laser beam incident on the ring-down cavity, and recording optical intensity attenuation data of light exiting the ring-down cavity using an optical intensity sensor system; analyzing the light intensity attenuation data.

18. 18. The system of claim 17, wherein the computer readable instructions, when executed by the at least one processor, cause the at least one processor to determine from the light intensity attenuation data of the second portion of the sample a probability that the subject from whom the sample was drawn has a particular physiological condition or a degree of a particular physiological condition, at least indirectly using a dataset of light intensity attenuation data of a pre-analyzed sample in which the presence or absence of the particular physiological condition or the degree of the particular physiological condition has been identified.

Images