Methods for determining analyte levels in a subject

Abstract

This document discloses a method for determining analyte levels in a subject. The method includes the following steps: • a) noninvasively detecting a first amount of at least one first volatile organic marker derived from a first source of the subject; • b) noninvasively detecting a second amount of at least the first volatile organic marker derived from a second source of the subject, wherein the second source is different from the first source; and • c) determining the analyte level in the subject based on the first amount and the second amount.

Description

Technical Field

[0001] This disclosure generally relates to the noninvasive determination of a subject's blood glucose level based on the detection of volatile organic markers in the subject's exhaled breath or other gaseous emissions. Background Technology

[0002] Monitoring blood glucose levels is crucial for diabetes management. Typical glucose measurement methods involve pricking the skin, usually a finger, to draw blood and apply it to a chemically active, disposable medium. To avoid the need for repeated skin pricking, various non-invasive blood glucose monitoring techniques have been developed.

[0003] Existing technologies describe the use of measuring volatile organic compounds (VOCs) in different environments, thus originating from different sources, such as breath samples, and detected via different methods, such as electrochemical methods. Furthermore, it is known to combine the detection of different VOCs, for example by simultaneously measuring different VOCs via one or more detection devices.

[0004] US10526633B2 describes an electrochemical sensor for volatile compounds of plants / plant pathogens, a volatile detection system for plants / plant pathogens, and a method for detecting stress-induced volatile compounds of plants and / or volatile compounds emitted by plant pathogens.

[0005] US9551712B2 describes a VOC group for use in breath analysis. Its use in the diagnosis, monitoring, or prognosis of breast cancer, head and neck cancer, prostate cancer, or colon cancer is disclosed.

[0006] US20210341461A1 describes a method for early detection and monitoring of cancer progression by detecting respiratory biomarkers. The method includes assessing the activity of aldehyde-ketone reductase by measuring the concentration of an exogenous substrate of the enzyme and / or measuring the concentration of a metabolite of the substrate in the exhaled breath of a subject. Preferably, the cancer is lung cancer.

[0007] EP2270496A1 describes the identification of biomarkers associated with cystic fibrosis (CF). It describes the diagnostic uses of these biomarkers and novel methods for their identification. Patients with cystic fibrosis are differentiated from healthy individuals by determining the presence of biomarkers selected from this group in exhaled air. These biomarkers have been found to be useful for the early diagnosis of cystic fibrosis.

[0008] WO2022101598A1 describes an apparatus for detecting the presence of one or more subjects infected with one or more viral, bacterial, and / or parasitic diseases in a closed environment, the apparatus comprising: (a) an air sampling unit capable of collecting an air sample of the atmosphere in the closed environment and transferring the sample for sensing; (b) a selected deterministic sensor group comprising at least two sensors reactive to the presence of a specific odor or volatile organic compound (VOC) in an air sample collected from the environment; (c) a processing unit comprising a pattern recognition analyzer that receives output signals from the sensor group, compares them with disease-specific patterns derived from a database of response patterns of the sensor group from the total bodily emissions of a subject with one or more known diseases, wherein each of the disease-specific patterns is a feature of a specific disease selected from bacterial, viral, and parasitic diseases, and selects the closest match between the output signals of the sensor group and the disease-specific pattern; and (d) a control system that triggers sampling of the air space of the environment at predetermined times or intervals to enable the apparatus to operate fully automatically and independently.

[0009] WO2012040318A2 describes compositions, methods, and kits for detecting melanoma and determining melanoma margins. It relates to a group of volatile metabolic biomarkers that can be used to diagnose melanoma, a skin cancer. One method for detecting melanoma is based on volatile byproducts of altered cancer metabolism. Furthermore, a method for identifying molecules that can be used to detect melanoma is provided, laying the foundation for the development of non-invasive detection technologies (biosensors) for melanoma diagnosis.

[0010] CN115297768A describes a method for determining whether a subject has mild cognitive impairment (MCI) or Alzheimer's disease, the method comprising measuring the concentration of one or more VOCs in a subject's exhaled sample and comparing that concentration to a reference concentration.

[0011] EP4085832A1 describes a method for non-invasively determining a subject's blood glucose level by analyzing exhaled breath or other gaseous emissions from the subject. The method includes non-invasively detecting the amount of at least one volatile organic marker in the subject's exhaled breath or other gaseous emissions as marker data, and determining the subject's blood glucose level based on the marker data. The volatile organic marker is selected from a group of markers, such as indole, wherein the amount of the volatile organic marker is negatively correlated with the blood glucose level.

[0012] Despite the advantages and progress achieved through the aforementioned developments, some significant technical challenges remain regarding the accuracy of analyte identification.

[0013] Problems to be solved

[0014] Therefore, it is desirable to provide a method and a system for determining analyte levels in subjects that at least partially address the aforementioned technical challenges. Specifically, a method and a system for determining analyte levels in subjects are desirable, as they increase the accuracy of determining analyte levels in subjects. Summary of the Invention

[0015] This problem is solved by a method for determining analyte levels in a subject, and a system for determining analyte levels in a subject, as described in the independent claims. Advantageous embodiments that can be implemented individually or in any combination are set forth in the dependent claims and throughout the specification.

[0016] As used below, the terms “have,” “contain,” or “include,” or any grammatical variations thereof, are used in a non-exclusive manner. Thus, these terms can refer either to a situation where no further features exist in the entity described in this context besides those introduced by these terms, or to a situation where one or more further features exist. For example, the statements “A has B,” “A includes B,” and “A contains B” can refer either to a situation where no other elements exist in A besides B (i.e., where A is solely and uniquely composed of B), or to a situation where entity A contains one or more further elements (such as element C, element D, or even further elements) besides B.

[0017] Furthermore, it should be noted that the terms "at least one," "one or more," or similar expressions indicating that a feature or element may exist once or more are generally used only once when introducing the corresponding feature or element. In the following text, in most cases, when referring to the corresponding feature or element, the expressions "at least one" or "one or more" will not be used repeatedly, even though the corresponding feature or element may exist only once or more.

[0018] Furthermore, as used below, the terms “preferredly,” “more preferably,” “particularly / specifically,” “more particularly / more specifically,” “specifically,” “more specifically,” or similar terms are used in combination with optional features without limiting the possibility of alternatives. Therefore, features introduced by these terms are optional features and are not intended to limit the scope of the claims in any way. As those skilled in the art will recognize, the invention can be practiced by using alternative features. Similarly, features introduced by “in one embodiment of the invention” or similar expressions are intended to be optional features without limiting alternative embodiments of the invention, without limiting the scope of the invention, and without limiting the possibility of combining features introduced in this way with other optional or non-optional features of the invention.

[0019] In a first aspect of the invention, a method for determining the analyte level in a subject is disclosed.

[0020] As used herein, the term "subject" is a broad term and will be given its common and conventional meaning to those skilled in the art, and is not limited to any particular or customary meaning. The term exemplarily refers to a person who intends to monitor the values ​​of an analyte (such as glucose levels). In embodiments, the term may specifically refer to, but is not limited to, a person applying the method. For example, a subject may be a patient with a condition such as diabetes. A subject may also be referred to as a user or patient. However, in embodiments, the person applying the method differs from a subject.

[0021] As used herein, the term "analyte" is a broad term and is given a common and conventional meaning to those skilled in the art, and is not limited to a specific or customary meaning. Specifically, the term may refer to, but is not limited to, any element, component, or compound that may be present in the body of a subject, and whose presence and / or concentration may be of interest to the subject or medical personnel (such as a physician). In particular, an analyte can be or may include any chemical substance or chemical compound that can participate in the metabolism of a subject, such as at least one metabolite. Specifically, an analyte can be glucose, more specifically, blood glucose. However, additionally or alternatively, other types of analytes and / or any combination of analytes that can identify analytes may be used.

[0022] As used herein, the terms “determining analyte levels” and “detecting analyte” are broad terms and will be given their common and conventional meanings to those skilled in the art, and are not limited to specific or customary meanings. The terms may specifically refer to, but are not limited to, the process of determining the presence and / or amount and / or concentration of at least one analyte. Therefore, detection may be or may include qualitative detection, which simply determines the presence or absence of at least one analyte, and / or may be or may include quantitative detection, which determines the amount and / or concentration of at least one analyte. As a result of detection, at least one signal characterizing the detection result may be generated.

[0023] The term "concentration" can generally refer to the amount of any substance in another medium. Specifically, the term "concentration" can refer to the amount of dissolved or undissolved substance in a fluid medium. Concentration can be described qualitatively, exemplarily as mass concentration, molar concentration, quantity concentration, or volume concentration. Thus, concentration can be quantified, respectively, as the mass of the substance divided by the volume of the other medium, the amount of substance in moles divided by the volume of the other medium, the quantity of the substance divided by the volume of the other medium, and the volume of the substance divided by the volume of the other medium. However, other kinds of descriptions may be feasible, such as by using equivalence concentration, molar concentration, mole fraction, molar ratio, mass fraction, or mass ratio. Thus, concentration can be quantified, respectively, as molar concentration divided by equivalence factor, amount of substance divided by the mass of the solvent in the other medium, amount of substance divided by the mass of the other medium, amount of substance divided by the total amount of all components in the other medium, amount of substance divided by the total amount of all other components in the other medium, mass of the substance divided by the mass of all components in the other medium, and mass of the substance divided by the mass of all other components in the other medium.

[0024] The concentration of glucose in a subject's blood may depend on events that increase or decrease glucose concentration, such as food intake or physical activity. Therefore, the concentration of glucose in the blood can be described as a time-dependent concentration. Generally, the term "time-dependent concentration" refers to the characteristic of concentration changing or altering over time. Thus, when assessing the concentration at a first time point, the concentration may have a first value, and when assessing the concentration at a second time point, the concentration may have a second value that may differ from the first value. The second value may be higher or lower than the first value.

[0025] The method includes the method steps listed below. These method steps can be performed in a given order. However, other orders of these method steps are also possible. Additionally, one or more of these method steps can be performed in parallel and / or with overlapping time. Furthermore, one or more of these method steps can be performed repeatedly. Additionally, there may be additional method steps not listed.

[0026] The method includes the following steps:

[0027] a) Non-invasively detect a first amount of at least one first volatile organic marker originating from a first source in the subject;

[0028] b) Non-invasively detecting a second amount of at least a first volatile organic marker derived from a second source different from the first source; and

[0029] c) Determine the analyte level of the subject based on the first and second measurements.

[0030] As used herein, the term "non-invasive detection" is a broad term and will be given its common and conventional meaning to those skilled in the art, and is not limited to a specific or customary meaning. Specifically, the term may refer to, but is not limited to, a process for determining the presence and / or amount and / or concentration of at least one analyte without inserting a component or part of a detection unit, such as an analyte sensor, into the body of a subject, specifically body tissue. Thus, the detection unit, including the analyte sensor, can remain entirely outside the body (particularly body tissue), particularly entirely outside the body. The detection unit can, exemplarily, be in close contact with the subject's skin. Exemplarily, the detection unit can be attached to the subject's skin via an adhesive or via a wristband. Furthermore, the detection unit can be placed near the subject's nostrils and / or mouth. Further details regarding embodiments of the detection unit are given below.

[0031] The terms "first quantity" and "second quantity" can be considered merely as a nomenclature, without assigning numbers or orders to the named elements, without specifying an order, and without excluding the possibility of several first and second quantities existing. Furthermore, additional quantities, such as one or more third quantities, may exist. Further details regarding additional quantities are given below.

[0032] As used herein, the term "quantity" is a broad term and will be given its common and conventional meaning to those skilled in the art, and is not limited to a specific or customary meaning. The term may specifically refer to, but is not limited to, the amount and / or concentration of a substance. As a result of determining the quantity, at least one signal characterizing the detection result may be generated, such as at least one measurement signal. It should be understood that the detected quantity of each volatile organic marker includes a single value or multiple values ​​changing over time, where each value represents a single measurement or the average of two or more measurements. It should also be understood that the quantity of a volatile organic marker is determined in absolute or relative terms, such as determining the amount of change in marker concentration.

[0033] The term "first volatile organic compound (VOC) marker" can be considered merely a nomenclature, without assigning numbers or orders to the elements, specifying no order, and not excluding the possibility of several first VOC markers existing. Furthermore, additional VOC markers may exist, such as one or more second VOC markers. Further details regarding additional VOC markers are given below.

[0034] As used herein, the term "volatile organic marker," also known as volatile organic compounds (VOCs), is a broad term and will be given its common and conventional meaning to those skilled in the art, and is not limited to any particular or customary meaning. The term may specifically refer to, but is not limited to, organic chemicals having high vapor pressures or, more precisely, low boiling points. Therefore, VOCs are volatile at relatively low temperatures, such as room temperature. The human body is the primary source of VOCs, which originate from various parts and processes within the body. So-called endogenous VOCs originate from metabolic processes within the body. A key organ involved in the production and transformation of such endogenous VOCs is the liver. VOCs originating from the environment are called exogenous VOCs, such as those entering the body through respiration, food ingestion, or diffusion through the skin, or originating from the microbiome, such as the microbiome of the digestive tract.

[0035] VOCs produced or absorbed in various parts of the body enter the bloodstream and are exhaled through gas exchange in the lungs. VOCs from the skin originate from the secretions of exocrine glands, sebaceous glands, or apocrine glands.

[0036] It should be noted that VOC can also refer to a derivative of the compound of interest, i.e., a compound derived from an endogenous process, or a fragment of the compound of interest formed by cleaving the molecule during the analytical process.

[0037] Specifically, the first volatile organic marker can be selected from a group of markers for which the amount of volatile organic markers is negatively correlated with glucose levels. Surprisingly, the negative correlation between the amount of volatile organic markers and blood glucose levels has proven particularly beneficial to the reliability, responsiveness, and sensitivity of this method.

[0038] Negative correlations may only be achieved through compounds that have a functional role in insulin response. The amount of such compounds (i.e., biomarkers) in respiration or other exudates is largely independent of external influences or individual patient characteristics.

[0039] Specifically, the first volatile organic marker can be one of indole (C8H7N), a partially saturated derivative of indole, a fully saturated derivative of indole, or a true fraction of indole.

[0040] Indole is an aromatic heterocyclic organic compound with the formula C8H7N. It is widely distributed in the natural environment and can be produced by various bacteria (such as *Escherichia coli*), typically found in the human gut. Furthermore, indole is the most abundant metabolite produced from tryptophan digestion. Because many bacteria in the human gut can synthesize indole from tryptophan via tryptophanase, the level of indole in the human gut is constant.

[0041] Furthermore, indole is a potential signaling molecule for stimulating glucagon-like protein-1 (GLP-1), which is required for an accurate insulin response. In the presence of glucose, indole diffuses into intestinal cells and blocks K+. + The channel increases GLP-1 secretion. This leads to a decrease in indole concentration in the extracellular space, which is then observed in respiration. Once extracellular glucose levels decline, indole diffuses out of the cell again and binds with K+. + Channel separation. These reversible mechanisms indicate that indole functions solely as a signaling molecule and is not metabolized. The result is an inverse progression of indole relative to blood glucose levels, i.e., a negative correlation. Furthermore, this correlation allows for real-time monitoring of blood glucose levels because any change in blood glucose concentration occurs simultaneously with a change in indole levels. Here, "simultaneously" means that the time interval between changes in the two concentrations is at most five minutes.

[0042] Derivatives are compounds derived from similar compounds through chemical reactions. Some derivatives of indole include aliphatic C8-amines, such as cyclohexylethylamine (C8H). 19 N) or octylamine (C8H) 17 N) or its isomers, showing a corresponding correlation with blood glucose and the aforementioned advantages. This also applies to fragments, such as benzene, where the fragment is named as a fragmented product, i.e., the dissociation of an energy-unstable molecular ion formed by molecules in the ionization chamber of a mass spectrometer. The term "true fragment" is used to indicate that the fragment originates from the dissociation of the claimed compound (i.e., indole).

[0043] Indole has a molecular weight of approximately 117.1 g / mol. Therefore, when indole is detected via PTR-ToF-MS, the additional H+ ions are present. + The mass-to-charge ratio after proton transfer in PTR-ToF-MS is m / z = 118.1.

[0044] Derivatives are, for example, those using H + Cyclohexyl-ethylamine C8H with a mass-to-charge ratio of m / z = 128.14 after protonation 17 N or octylamine C8H with a mass-to-charge ratio of m / z = 130.15 after protonation. 19 N.

[0045] Indole fragments are, for example, when using H... + Benzene C6H6 with a mass-to-charge ratio of m / z = 79.055 after protonation and C7H8 with a mass-to-charge ratio of m / z = 93.069 after protonation.

[0046] In addition, specifically, the first volatile organic compound (VOC) marker can be selected from the group consisting of: formaldehyde, methanol. Alternatively, the first VOC marker can be selected from the group consisting of: acrolein, acetic acid, butanone, propionic acid, phenol, acetone, propionamide, butyric acid. Other VOC markers may also be available.

[0047] As used herein, the term "source" is a broad term and will be given its common and conventional meaning to those skilled in the art, and is not limited to any particular or customary meaning. Specifically, the term may refer to, but is not limited to, any area of ​​the subject's body from which or in which any substance originates or is present, or an area immediately adjacent to the subject's body. Furthermore, the term "source" may refer to the excretion or secretion of a substance from the subject's body, particularly from a part or area of ​​the subject's body and / or from organs of the subject's body (such as skin, nose, and / or mouth). Specifically, the term "source" refers to the excretion or secretion of a part or area of ​​the subject's body or immediately adjacent to the subject's body and / or from the subject's body, specifically at least one first volatile organic marker and / or one or more other volatile organic markers originating from or present in a part or area of ​​the subject's body and / or organs of the subject's body (such as skin, nose, and / or mouth).

[0048] The terms "first source" and "second source" are to be considered merely nomenclature, without numbering or ordering the elements, without specifying an order, and without excluding the possibility that there may be several first and second sources. Furthermore, additional sources may exist, such as one or more third sources. As mentioned above, the second source differs from the first source. Therefore, the first and second sources can be different kinds of sources. The first and second sources can be different regions of the subject's body from which or in which the first volatile organic marker originates, or regions adjacent to the subject's body. Specifically, the first and second sources can refer to different organs of the subject from which or in which the first volatile organic marker originates. Specifically, the first source can be exhaled air, specifically exhaled air originating from the subject's nostrils and / or mouth, and the second source can be another gaseous emission source different from exhaled air, or vice versa. Therefore, the second source can be exhaled air, specifically exhaled air originating from the subject's nostrils and / or mouth, and the first source can be another gaseous emission source different from exhaled air. Specifically, sources of gaseous emissions other than exhaled breath can correspond to secretions from the subject's skin (specifically from skin regions). Exemplarily, skin, specifically skin regions, can be selected from the group consisting of: head, chest, back, armpits, waist, arms, or genital areas. However, other skin regions are also possible.

[0049] As described above, in step c), the subject's analyte level can be determined based on the first and second quantities. Due to the correlation between the entities, the subject's analyte level can be determined based on the amount of the first volatile organic marker detected in exhaled breath and other emissions. It should be understood that the correlation provides a value for the analyte (such as blood glucose) derived from the absolute or relative value of the concentration of the first volatile organic marker in exhaled breath and other emissions. Exemplarily, the first volatile organic data can be normalized based on a reference signal, such as one derived from ambient air or from another marker in the same sample of exhaled breath or from another sample of exhaled breath or other emissions. Exemplarily, the setup for performing the method can be taught and / or trained based on individual metabolic processes, such as creating an individual profile of the subject. For example, training involves selecting volatile organic markers from a set of volatile organic markers that also contain derivatives or fragments of endogenous compounds.

[0050] The method may further include:

[0051] d) Non-invasively detecting a third amount of at least a second volatile organic marker originating from a first source in the subject; and

[0052] e) Non-invasively detect a fourth amount of at least one second volatile organic marker originating from a second source in the subject.

[0053] Specifically, steps d) and e) can be performed before step c). In step c), the subject's analyte level can be detected based on a first, second, third, and fourth quantity. Specifically, the second volatile organic marker can be different from the first volatile organic marker. Therefore, the first and second volatile organic markers can be different types of volatile organic markers. Specifically, the first volatile organic marker can be indole, and the second volatile organic marker can be formaldehyde or methanol. However, other embodiments are also possible. Specifically, the analyte can be glucose, and the method can further include detecting hyperglycemia or hypoglycemia based on the first, second, third, and fourth quantities.

[0054] Specifically, the first and second volatile organic compound (VOC) markers can be formaldehyde and methanol. Alternatively, the first and second VOC markers can be formaldehyde and acrolein. Furthermore, the first and second VOC markers can be formaldehyde and propionamide. Additionally, the first and second VOC markers can be acetone and indole. Using all these groups, a mean absolute relative difference (MARD) of 21% can be achieved for type 1 diabetes and a MARD of 13% for type 2 diabetes.

[0055] The accuracy of blood glucose meters and continuous glucose systems is typically assessed primarily via mean absolute relative difference (MARD). This involves comparing the value determined by the measurement system with a measurement from a reference device, preferably a product considered a standard due to its high accuracy. The difference between the two values ​​gives the relative deviation. A quantity of difference is usually used because this allows for the assumption of both positive and negative values. MARD can be considered the average of each relative deviation.

[0056] The method may further include:

[0057] f) Non-invasive detection of a fifth quantity of at least a third volatile organic marker originating from a primary source in the subject; and

[0058] g) Non-invasively detect at least one third volatile organic marker derived from a second source in the subject.

[0059] Specifically, steps f) and g) can be performed before step c). In step c), the subject's analyte level can be detected based on a first, second, third, fourth, fifth, and sixth quantity. Specifically, the third volatile organic marker can be different from the first and second volatile organic markers. Therefore, the first, second, and third volatile organic markers can be different types of volatile organic markers. Specifically, the first volatile organic marker can be indole, the second volatile organic marker can be formaldehyde, and the third volatile organic marker can be methanol, or vice versa. However, other embodiments are also possible. Specifically, the analyte can be glucose, and the method can further include detecting hyperglycemia or hypoglycemia based on the first, second, third, fourth, fifth, and sixth quantities.

[0060] Specifically, the first, second, and third volatile organic compound (VOC) markers can be formaldehyde, indole, and methanol. Additionally, the first, second, and third VOC markers can be formaldehyde, indole, and acrolein. Furthermore, the first, second, and third VOC markers can be formaldehyde, indole, and acetic acid. Furthermore, the first, second, and third VOC markers can be formaldehyde, indole, and propionic acid. Furthermore, the first, second, and third VOC markers can be formaldehyde, indole, and butyric acid. Furthermore, the first, second, and third VOC markers can be methanol, indole, and acetone. Using all these groups, a mean absolute relative difference (MARD) of 17% can be achieved for type 1 diabetes and a MARD of 10% for type 2 diabetes.

[0061] The method may further include:

[0062] h) Non-invasive detection of the seventh quantity of at least the fourth volatile organic marker originating from the subject's first source; and

[0063] i) Non-invasive detection of the eighth amount of at least one fourth volatile organic marker derived from a second source of the subject.

[0064] Specifically, steps h) and i) can be performed before step c). In step c), the subject's analyte level can be detected based on a first, second, third, fourth, fifth, sixth, seventh, and eighth quantity. Specifically, the fourth volatile organic marker can be different from the first, second, and third volatile organic markers. Therefore, the first, second, third, and fourth volatile organic markers can be different types of volatile organic markers. Specifically, the analyte can be glucose, and the method can further include detecting hyperglycemia or hypoglycemia based on the first, second, third, fourth, fifth, sixth, seventh, and eighth quantities.

[0065] Specifically, the first, second, third, and fourth volatile organic compound (VOC) markers can be formaldehyde, methanol, indole, and propionamide. Additionally, the first, second, third, and fourth VOC markers can be formaldehyde, methanol, indole, and propionic acid. Furthermore, the first, second, third, and fourth VOC markers can be formaldehyde, methanol, indole, and acetic acid. Furthermore, the first, second, third, and fourth VOC markers can be formaldehyde, methanol, indole, and butyric acid. Furthermore, the first, second, third, and fourth VOC markers can be formaldehyde, methanol, indole, and acrolein. Furthermore, the first, second, third, and fourth volatile organic compound (VOC) markers can be formaldehyde, indole, butyric acid, and propionic acid. Using all these groups, a mean absolute relative difference (MARD) of 14% can be achieved for type 1 diabetes and a MARD of 8% for type 2 diabetes.

[0066] The method may further include:

[0067] j) Non-invasive detection of the ninth quantity of at least the fifth volatile organic marker derived from the subject's primary source; and

[0068] k) Non-invasively detect the tenth amount of at least one fifth volatile organic marker derived from a second source of the subject.

[0069] Specifically, steps j) and k) can be performed before step c). In step c), the analyte level of the subject can be detected based on the first, second, third, fourth, fifth, sixth, seventh, eighth, ninth, and tenth quantities. Specifically, the fifth volatile organic marker can be different from the first, second, third, and fourth volatile organic markers. Therefore, the first, second, third, fourth, and fifth volatile organic markers can be different types of volatile organic markers. Specifically, the analyte can be glucose, and the method can further include detecting hyperglycemia or hypoglycemia based on the first, second, third, fourth, fifth, sixth, seventh, eighth, ninth, and tenth quantities.

[0070] Specifically, the first, second, third, fourth, and fifth volatile organic compound (VOC) markers can be formaldehyde, methanol, indole, propionamide, and butyric acid. Alternatively, they can be methanol, acetone, propionamide, butyric acid, and indole. Furthermore, the first, second, third, fourth, and fifth volatile organic compound (VOC) markers can be formaldehyde, methanol, indole, acrolein, and propionamide. (This text is repeated four times in the original.) Furthermore, the first, second, third, fourth, and fifth volatile organic compound (VOC) markers can be formaldehyde, methanol, indole, propionic acid, and butyric acid. Additionally, the first, second, third, fourth, and fifth VOC markers can be formaldehyde, methanol, indole, propionamide, and propionic acid. Furthermore, the first, second, third, fourth, and fifth VOC markers can be formaldehyde, methanol, indole, acetone, and phenol. Using all these groups, a mean absolute relative difference (MARD) of 12% can be achieved in type 1 diabetes and a MARD of 6% to 7% in type 2 diabetes.

[0071] Specifically, in step a), a first amount of at least one first volatile organic compound (VOC) originating from a first source of the subject can be detected continuously or discontinuously. Furthermore, specifically, in step b), a second amount of at least one first VOC originating from a second source of the subject can be detected continuously or discontinuously. Specifically, in step a), the first amount of at least one first VOC originating from the first source of the subject can be detected discontinuously, and in step b), the second amount of at least one first VOC originating from a second source of the subject can be detected continuously. Specifically, the first source can be exhaled air, and in step a), the first amount of at least one first VOC originating from the first source of the subject can be detected discontinuously. Furthermore, specifically, the second source can be another gaseous emission source different from exhaled air, such as a skin area of ​​the subject, and in step b), the second amount of at least one first VOC originating from the second source of the subject can be detected continuously. When the first amount and / or the second amount can be detected continuously, a data stream can be acquired. The term "data stream" can specifically refer to a sequence or set of data elements, such as signals, and more specifically, electrical signals, which are available or provided to any device over time or in a time-dependent manner. A data stream can be a continuous stream. However, a data stream may include or may have one or more gaps, during which no data can be transmitted to the device. Furthermore, the number of data elements or signals transmitted to the device per unit of time can vary over time. When the first quantity and / or the second quantity are detected discontinuously, only a single data element or several data elements may be acquired during a predefined time period.

[0072] This method may further include displaying the subject's analyte levels, such as via at least one display device, such as a smartwatch or smartphone. Thus, the subject can be informed of his or her blood sugar status, such as in the morning. See below for details.

[0073] The method may further include comparing the subject's analyte level with a predefined threshold or with a predefined tolerance range defined by a first threshold and a second threshold different from the first threshold. The term "threshold" can generally refer to a defined, predefined, or determinable numerical value used as a comparison with one or more information items (such as data and / or measurements) to derive at least one secondary information item. As an example, a threshold can define a comparison value for comparison with measured data and / or other information items, where, depending on the comparison, one or more results can be stated. As an example, a threshold can define a comparison value where one or more results are stated once at least one information item reaches, exceeds, or falls below the comparison value. The terms "first threshold" and "second threshold" can be considered merely as nomenclature, without numbering or ordering the named elements, without specifying an order, and without excluding the possibility that several first and second thresholds may exist. Furthermore, additional thresholds, such as one or more third thresholds, may exist. The target range can be defined by the first and second thresholds. The target range can define the desired range of analyte levels, such as a tolerance range. Thus, the subject can be in optimal health regarding the desired analyte level. The term “predefined” can specifically refer to a threshold or tolerance range that is defined (e.g., predetermined and known) before the method is applied.

[0074] A warning signal may be issued if a subject's analyte level exceeds a predefined tolerance range. The warning signal can provide an indication of the subject's required action (such as drug or food intake). As used further herein, the term "signal" can refer to any indication that can be transmitted from one element to another, particularly for the purpose of instruction, warning, guidance, or command. Therefore, the indication may include at least one information item. Specifically, the signal may be or may include at least one of electronic, visual, acoustic, or vibrational signals.

[0075] This method may include acquiring a series of continuous measurements. These measurements can be used to predict the subject's analyte levels in the future, such as within minutes. The method may include predicting the subject's analyte levels using at least one mathematical model. When the analyte is glucose, specifically blood glucose, there may be a time lag between changes in blood glucose levels and changes in volatile organic markers derived from respiration. Typically, changes in volatile organic markers derived from respiration may occur earlier than changes in blood glucose levels. The time lag may be, for example, approximately 5 minutes. The mathematical model may account for this time lag.

[0076] Therefore, the method according to the invention may be advantageous compared to detecting analyte levels in a subject from blood or interstitial fluid. During events known to affect a subject's analyte levels (such as dietary intake), the method according to the invention may have a time advantage compared to detecting analyte levels in blood or interstitial fluid.

[0077] The mathematical model may include at least one machine learning architecture, such as a neural network, specifically to improve the accuracy of glucose readings. The mathematical model may specifically include at least one linear and / or nonlinear time series model, specifically to improve predictive performance. This method may specifically use a recurrent neural network (RNN), more specifically a long short-term memory (LSTM) network. Furthermore, the mathematical model may include at least one double logarithmic function.

[0078] This method may further include field calibration. Field calibration may include a procedure for developing a personalized algorithm for a specific subject. Field calibration may include calibration on individual subjects. Individual subjects may undergo several tests, as described in more detail below. In addition, individual subject characteristics may be considered, as described in more detail below. Personalized mathematical models for subjects may be developed and / or mathematical models may be selected from multiple easily pre-defined mathematical models. Specifically, field calibration may include classifying and / or clustering subjects into specific mathematical models, specifically selected from multiple easily pre-defined mathematical models. Specifically, subjects may be clustered in individual clusters, such as in about 5 individual clusters, and may be clustered in an individual base model developed for each cluster. The model may be enhanced by additional features obtained from the first data point from the individual subject. For field calibration, a single respiratory measurement may be performed by the subject. In addition, additionally or alternatively, for field calibration, a whole series of respiratory measurements during an event with analyte variation may be used. Specifically, the analyte may be glucose, and the event with analyte variation may be an oral glucose tolerance test (OGTT). Furthermore, for field calibration, during events with analyte variations, individual analyte measurements or an entire series of measurements can be performed on the subject by an alternative method, such as analyte detection in bodily fluids (e.g., blood or interstitial fluid) to determine the subject's analyte levels. Additionally, or alternatively, for field calibration, further subject characteristics may be considered. Subject characteristics can be freely selected from the following groups: age, sex, body mass index (BMI), weight, medications, type of diabetes, smoking behavior. Other and / or further subject characteristics are also possible. Field calibration may include providing subjects with feedback on the goodness of fit to a particular mathematical model. In cases of insufficient goodness of fit, field calibration may include providing subjects with options for improving the field calibration, such as considering additional and / or different subject characteristics and / or measurements. The provision of options may also include suggesting the use of different devices with different methods to determine the subject's analyte levels.

[0079] This method may include the use of a database. The database may include multiple different mathematical models. Specifically, the mathematical models may be trained separately for specific patient groups. Specifically, the mathematical models may be trained separately on group-specific training data. This method may include selecting a mathematical model from the database, specifically based on the subject's parameters, and more specifically based on field calibration as described above. Specifically, this method may include parameters input by the subject via a user interface.

[0080] The method may further include detecting events known to affect a subject's analyte levels. Specifically, the method may include detecting events known to increase a subject's analyte levels. Specifically, the analyte may be glucose, and the method may include detecting the subject's dietary intake. Dietary intake may also be detected if the subject has not recorded it. The method may further include estimating dietary volume and / or the glycemic index of the diet. Detection of events known to affect a subject's analyte levels may be specifically based on characteristics in analyte trends. Specifically, detection of dietary intake may be based on characteristic increases in glucose. Furthermore, the method may include considering contextual information, such as events known to affect a subject's analyte levels, exemplarily dietary intake or the subject's physical activity, to predict analyte trends. As described above, contextual information can be detected by this method. However, additionally or alternatively, contextual information may be input by the subject via a user interface. Furthermore, the method may include identifying time periods with reduced accuracy in glucose level prediction, such as time periods with rapid analyte changes, and the method may include providing the subject with information about these time periods.

[0081] The analyte can be glucose, and the method can further include providing an estimate of insulin administration, specifically an estimate of improved insulin administration. The estimate of insulin administration can be based on average glucose levels, specifically during specific time intervals, such as during sleep or at specific intervals throughout the day. Alternatively, the estimate of insulin administration can be based on the rise in glucose after dietary intake. Other options may also be possible.

[0082] The method may further include at least one fail-safe step. As used herein, the term "fail-safe step" is a broad term and should be given its common and customary meaning to those skilled in the art, and should not be limited to a specific or customary meaning. Specifically, the term may refer to, but is not limited to, at least one step that ensures the prevention of the generation and / or determination and / or display of unreliable or even erroneous measurements. During a fail-safe step, the identification of unrealistic measurements may occur.

[0083] In another aspect of the invention, a system for determining the analyte level in a subject is disclosed. This system is configured to perform a method as described above or as will be further described in more detail below.

[0084] As used herein, the term "system" is a broad term and will be given a meaning common and customary to those skilled in the art, and is not limited to a particular or customary meaning. Specifically, the term may refer to, but is not limited to, a group of at least two elements or components capable of interacting to perform at least one common function, in this case, for determining or facilitating the determination of an analyte level in a subject. A system may specifically include components of two or more parts capable of interacting with each other, such as for performing one or more diagnostic purposes, such as for performing medical analysis. A system may also generally be referred to as a component or kit.

[0085] The system includes:

[0086] ● First volatile organic compound (VOC) marker detection unit, configured to non-invasively detect a first amount of at least one first VOC marker originating from a first source of the subject;

[0087] ● A second volatile organic compound (VOC) marker detection unit, configured to non-invasively detect a second amount of at least one first VOC marker originating from a second source in the subject; and

[0088] ● At least one assessment device, configured to determine the analyte level of the subject based on a first quantity and a second quantity.

[0089] The terms "first volatile organic compound (VOC) marker detection unit" and "second VOC marker detection unit" are to be considered merely as nomenclature, without numbering or ordering the elements, without specifying an order, and without excluding the possibility that several first or second VOC marker detection units may exist. Furthermore, additional VOC marker detection units may exist, such as one or more third VOC marker detection units. As used herein, the term "volatile organic compound (VOC) marker detection unit" is a broad term and will be given its common and conventional meaning to those skilled in the art, and is not limited to a specific or customary meaning. The term may specifically refer to, but is not limited to, any element suitable for performing the detection process and / or suitable for use in the detection process. Therefore, a VOC marker detection unit may specifically be suitable for determining the amount of a first VOC marker.

[0090] At least one of the first and second volatile organic compound (VOC) marker detection units can be selected from the group consisting of: gas sensors, exemplarily metal oxide sensors; carbon nanotube chips (CNTs), specifically functionalized carbon nanotube chips; photometric sensors, exemplarily absorption-based light source detectors; mass spectrometers, exemplarily ion mobility spectrometers; electrochemical sensors; and biochemical sensors having functionalized surfaces, wherein the functionalized surfaces may specifically include at least one of amino acids, peptides, proteins, DNA building blocks, other biomolecules, or fragments thereof. Other types of VOC marker detection units are also feasible. Gas sensors may specifically be miniaturized sensors. Regarding gas sensors, reference can be made to commercially available metal oxide sensors that can be used to monitor indoor and outdoor air pollution.

[0091] The system may include at least one vapor distribution member configured to contact a sample from a subject’s exhaled breath or another gaseous source with a first volatile organic compound (VOC) marker detection unit or a second VOC marker detection unit.

[0092] As used herein, the term "evaluation device" is a broad term and is given a common and conventional meaning to those skilled in the art, and is not limited to a specific or customary meaning. Specifically, the term may refer in a non-limiting manner to any component designed to actuate any sensor and / or record signals from the sensor and / or derive at least one piece of information about the analyte from the signal and / or evaluate these signals, wholly or partially. Therefore, an evaluation device may specifically be or may include electronic components. Electronic components may be configured to perform one or more of the following: perform measurements using sensors, perform voltage measurements, perform current measurements, record sensor signals, store measurement signals or measurement data, or transmit sensor signals or measurement data to another device. Therefore, electronic components may specifically include at least one of the following: voltmeter, ammeter, potentiometer, voltage source, current source, signal receiver, signal transmitter, analog-to-digital converter, electronic filter, energy storage device, data processing device, such as a microcontroller. Other embodiments of the electronic components are possible. The electronic components may specifically include at least one circuit board having elements of the electronic components disposed thereon.

[0093] The first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit may be connected to or be connectable to each other, specifically via wireless communication, preferably via wireless far-field communication, and more preferably via radio frequency (RF) transmission. As used herein, the term "wireless far-field communication" generally refers to wireless communication suitable for transmitting data over long distances, such as distances exceeding 10 cm. As an example, wireless far-field communication can be any long-distance communication using electromagnetic waves in the radio frequency range, i.e., it can be radio communication. Therefore, as an example, wireless far-field communication may include at least one radio module having at least one radio antenna for transmitting data to another device via radio transmission.

[0094] Specifically, at least one of the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit may be configured to transmit data to at least one external device, specifically to at least one smartphone or at least one smartwatch. As used herein, the term "external device" can be any device adapted to receive data from the first VOC marker detection unit and / or the second VOC marker detection unit via wireless far-field communication. The at least one external device may be part of the system or may be independent of the system. As an example, the at least one external device may be a portable device, such as a handheld computer and / or a smartphone, capable of communicating via wireless far-field communication. Other examples are also possible.

[0095] The first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit can be placed at different locations on the subject's body. At least one of the first VOC marker detection unit and the second VOC marker detection unit can be disposed on a wristband, a patch such as one that can be placed on the subject's skin via adhesive, glasses, a nose clip, a nose ring, a nose stud, a headband, headphones, a chest strap, clothing, a pillow, or a device on a bedside table. Specifically, one of the first VOC marker detection unit and the second VOC marker detection unit can be configured to establish skin contact with the subject's skin, preferably continuously and / or permanently, and the other of the first VOC marker detection unit and the second VOC marker detection unit can be configured to be disposed continuously and / or permanently and / or discontinuously below the subject's nostrils and / or below the subject's mouth. At least one of the first VOC marker detection unit and the second VOC marker detection unit can be disposed on or integrated into a smartwatch.

[0096] A first volatile organic compound (VOC) marker detection unit may be disposed on a wristband and configured to preferably discontinuously detect a first VOC marker from a first source, namely exhaled breath, of the subject. Furthermore, the system may include a vapor distribution member configured to contact a sample from the subject's exhaled breath or another gaseous emission source with the first VOC marker detection unit. A second VOC marker detection unit may be disposed on the wristband and configured to preferably continuously detect a first VOC marker from a second source of the subject, namely the subject's skin. Specifically, both the first and second VOC marker detection units may be disposed on the wristband.

[0097] Exhaled air samples can be captured using a mouthpiece, nasal cannula, handheld breath analyzer, or any other device suitable for capturing at least a portion of the exhaled air.

[0098] The present invention further discloses and proposes a computer program comprising computer-executable instructions for performing the method according to the invention in one or more embodiments appended herein, when executed on a computer or computer network. Specifically, the computer program may be stored on a computer-readable data carrier. Thus, specifically, one, more, or even all of the methods indicated above as a) to c) can be performed by using a computer or computer network, preferably by using a computer program.

[0099] The present invention further discloses and proposes a computer program product having program code tools so that, when the program is executed on a computer or computer network, the method according to the invention is performed in one or more embodiments appended herein. Specifically, the program code tools may be stored on a computer-readable data carrier.

[0100] Furthermore, the present invention discloses and proposes a data carrier having a data structure stored thereon, which, after being loaded into a computer or computer network, such as after being loaded into the working memory or main memory of the computer or computer network, can perform methods according to one or more embodiments disclosed herein.

[0101] This invention further proposes and discloses a computer program product having program code tools stored on a machine-readable carrier, so that when the program is executed on a computer or computer network, it performs methods according to one or more embodiments disclosed herein. As used herein, a computer program product refers to a program that is a tradable product. This product can generally exist in any format (such as in paper format) or on a computer-readable data carrier. Specifically, the computer program product can be distributed on a data network.

[0102] Finally, the present invention proposes and discloses a modulated data signal containing instructions readable by a computer system or computer network for performing a method according to one or more embodiments disclosed herein.

[0103] Preferably, referring to the computer implementation aspects of the invention, one or more, or even all, of the method steps a) to c) according to one or more embodiments disclosed herein can be performed using a computer or computer network. Therefore, generally, any method steps involving the provision and / or manipulation of data can be performed using a computer or computer network. Typically, these method steps can include any method steps other than those generally requiring manual work (such as providing samples and / or performing actual measurements).

[0104] Specifically, the present invention further discloses:

[0105] - A computer or computer network including at least one processor, wherein the processor is adapted to perform a method according to an embodiment described herein, specifically one or more of method steps a) to c).

[0106] - A computer-loadable data structure adapted to execute a method according to an embodiment described in this specification while the data structure is executed on a computer, specifically one or more of method steps a) to c).

[0107] - A computer program adapted to perform, while the program is executed on a computer, a method according to an embodiment described in this specification, specifically one or more of method steps a) to c).

[0108] - A computer program including program tools for performing, when the computer program is executed on a computer or on a computer network, a method according to an embodiment described in this specification, specifically one or more of method steps a) to c).

[0109] - A computer program, comprising program means according to the foregoing embodiments, wherein the program means is stored on a computer-readable storage medium.

[0110] - A storage medium wherein a data structure is stored on the storage medium, and wherein the data structure is adapted to perform a method according to an embodiment described herein, specifically one or more of method steps a) to c), after being loaded into the main memory and / or working memory of a computer or computer network.

[0111] - A computer program product having program code tools, wherein such program code tools may be stored or stored on a storage medium for performing a method according to an embodiment described in this specification, specifically one or more of method steps a) to c), when the program code tools are executed on a computer or on a computer network.

[0112] The proposed method and apparatus demonstrate many advantages over known apparatus and methods.

[0113] This invention relates to non-invasive analyte determination based on volatile organic compounds (VOCs). Measurements of the same volatile markers or the same combination of volatile markers from at least two different sources can be combined. Combining measurements of volatile organic markers from a first source (such as respiration) with measurements of the same markers from another source (such as emissions (such as sweat)) can significantly increase the accuracy of analyte determination.

[0114] Combining multiple sources of one or more volatile biomarkers can significantly increase the accuracy of non-invasive analyte determination based on volatile biomarkers. Furthermore, it allows for user-friendly measurement procedures, such as combining continuous emission monitoring via non-invasive skin sensors with timely respiratory monitoring via the same or different sensor devices.

[0115] Several detectors, particularly sensors, can be used, and they can be placed at different locations on the subject's body. The detectors can be connected to or interconnected with each other, for example, via radio frequency transmission, and can also be connected to or interconnected with external devices, such as smartphones. Specifically, a single VOC can be detected from all sources. Measurements can be performed continuously or as point monitoring.

[0116] A smartwatch worn on the wrist can be configured to continuously detect VOCs from the skin, especially when glucose levels are not displayed. If a subject wishes to perform a glucose measurement such as spot monitoring, the subject can move the arm wearing the smartwatch close to his / her mouth and blow air onto it 2 to 3 times. The smartwatch can be configured to detect this process and can be configured to assess 2 to 3 exhaled breath cycles. The combined value of VOCs from respiration and skin can be used to calculate the glucose level.

[0117] The glasses may be equipped with sensors on the frame for VOCs originating from the skin. These sensors may have skin contact capability. Additionally, or alternatively, the glasses may have similar sensors behind the ears. Furthermore, sensors for VOCs originating from exhaled breath may be placed below the nostrils.

[0118] Nose clips, nose rings, and / or nose studs can be equipped with sensors for VOCs originating from exhaled breath and can be wirelessly connected or connected to smartwatches and / or smart rings with sensors for VOCs originating from the skin.

[0119] A smartwatch on the wrist, a smart ring on the finger, or a patch at different locations on the skin for skin-derived VOCs can be connected to or can be connected to a smartphone for exhaled VOCs, which can perform point monitoring by exhaling the VOCs.

[0120] It is possible to implement an earphone having a sensor arm for VOCs originating from exhaled breath, such as a microphone arm similar to that of ordinary earphones, and additional sensors for VOCs originating from the skin, located on the cheek or behind the ear. The sensors for VOCs originating from exhaled breath can be coupled to, or can be coupled to, the additional sensors for VOCs originating from the skin, located on the wrist (such as via a smartwatch) or on the finger (such as via a smart ring).

[0121] In addition, clothing can be equipped with detectors for VOCs originating from the skin. Detection systems for VOCs originating from exhaled breath are also possible; these systems are not wearable but can be placed near the subject to monitor blood glucose levels, such as devices on pillows and / or bedside tables.

[0122] This system may include a screening alarm system that specifically does not provide continuous glucose values, but only alerts the subject when the value is "out of range" (i.e., hyperglycemia or hypoglycemia). "Unique clusters," or specialized VOC groups, can be used to specifically detect hyperglycemic or hypoglycemic states. Glucose time-varying curves can be recorded, and instructions regarding action can be provided. Subjects can be informed of their blood glucose status, such as in the morning.

[0123] A continuous series of measurements can be used to support or improve predictions of glucose status, such as trends in the near future (e.g., within minutes). Machine learning, including neural networks, can be used as a mathematical model to improve the accuracy of glucose readings. Other mathematical evaluation models using double logarithmic functions can be applied.

[0124] On-site calibration can be applied, including procedures for developing personalized algorithms for specific subjects. A full series of respiratory measurements can be used during events with glucose variations, such as the oral glucose tolerance test (OGTT). Only a few single respiratory measurements can be applied to classify and / or cluster patients into a specific algorithm selected from multiple easily pre-defined algorithmic options. Clustering or grouping can be supported by other patient characteristics, independent of their breathing, such as age, sex, BMI, weight, and / or medications. Feedback on the goodness of fit to a particular algorithm can be provided to patients, offering improvement options (performing an OGTT or using a different CGM system) if needed. Failure safety, such as for erroneous results, can be achieved using VOCs derived from exhaled breath and VOCs derived from the skin.

[0125] In summary, and without excluding other possible embodiments, the following embodiments are conceivable:

[0126] Example 1: A method for determining the analyte level in a subject, the method comprising the following steps:

[0127] a) Non-invasively detect a first amount of at least one first volatile organic marker originating from a first source in the subject;

[0128] b) Non-invasively detecting a second amount of at least a first volatile organic marker derived from a second source different from the first source; and

[0129] c) Determine the analyte level of the subject based on the first and second measurements.

[0130] Example 2: According to the method described in the foregoing examples, the first source is exhaled air, and the second source is a different gaseous emission source from exhaled air, or vice versa.

[0131] Example 3: The method according to any one of the foregoing examples, wherein the analyte is glucose, specifically blood glucose.

[0132] Example 4: The method according to any one of the preceding examples, wherein the first volatile organic marker is selected from a group of markers in which the amount of the volatile organic marker is negatively correlated with the glucose level.

[0133] Example 5: The method according to any one of the preceding examples, wherein the first volatile organic marker is one of indole (C8H7N), a partially saturated derivative of indole, a fully saturated derivative of indole, and a true fraction of indole.

[0134] Example 6: The method according to any one of the preceding examples, wherein the first volatile organic marker is selected from the group consisting of: formaldehyde, methanol.

[0135] Example 7: The method according to any one of the foregoing embodiments, wherein the method further includes:

[0136] d) Non-invasively detecting a third amount of at least a second volatile organic marker originating from a first source in the subject; and

[0137] e) Non-invasively detect a fourth amount of at least one second volatile organic marker derived from the second source of the subject;

[0138] In step c), the analyte level of the subject is detected based on the first amount, the second amount, the third amount, and the fourth amount.

[0139] Example 8: According to the method described in the foregoing examples, the first volatile organic marker is indole, and the second volatile organic marker is formaldehyde or methanol.

[0140] Example 9: The method according to any one of the preceding two examples, wherein the analyte is glucose, and wherein the method further includes detecting a state of hyperglycemia or hypoglycemia based on the first amount, the second amount, the third amount, and the fourth amount.

[0141] Example 10: The method according to any one of the foregoing three embodiments, wherein the method further comprises:

[0142] f) Non-invasive detection of a fifth quantity of at least a third volatile organic marker originating from a primary source in the subject; and

[0143] g) Non-invasively detect a sixth amount of at least one third volatile organic marker derived from the second source of the subject;

[0144] In step c), the analyte level of the subject is detected based on the first, second, third, fourth, fifth, and sixth quantities.

[0145] Example 11: According to the method described in the foregoing examples, the first volatile organic compound (VOC) marker, the second VOC marker, and the third VOC marker are formaldehyde, methanol, and indole.

[0146] Example 12: The method according to any one of the preceding two embodiments, wherein the analyte is glucose, and wherein the method further includes detecting a state of hyperglycemia or hypoglycemia based on the first amount, the second amount, the third amount, the fourth amount, the fifth amount, and the sixth amount.

[0147] Example 13: The method according to any one of the foregoing embodiments, wherein in step a), the first amount of the at least one first volatile organic marker originating from the first source of the subject is detected continuously or discontinuously.

[0148] Example 14: The method according to any one of the preceding examples, wherein in step b), the second amount of the at least one first volatile organic marker originating from the second source of the subject is detected continuously or discontinuously.

[0149] Example 15: The method according to any one of the preceding embodiments, wherein in step a), the first amount of the at least one first volatile organic marker originating from the first source of the subject is continuously detected, and wherein in step b), the second amount of the at least one first volatile organic marker originating from the second source of the subject is discontinuously detected.

[0150] Example 16: The method according to any one of the foregoing embodiments, wherein the method further includes displaying the subject's analyte level to the subject.

[0151] Example 17: The method according to any one of the foregoing embodiments, wherein the method further includes comparing the subject's analyte level with a predefined tolerance range defined by a first threshold and a second threshold different from the first threshold, wherein a warning signal is output if the subject's analyte level is outside the predefined tolerance range.

[0152] Example 18: A system for determining the analyte level of a subject, wherein the system is configured to perform the method according to any one of the foregoing embodiments, wherein the system comprises:

[0153] ● First volatile organic compound (VOC) marker detection unit, configured to non-invasively detect a first amount of at least one first VOC marker originating from a first source of the subject;

[0154] ● A second volatile organic compound (VOC) marker detection unit, configured to non-invasively detect a second amount of at least one first VOC marker originating from a second source in the subject; and

[0155] ● At least one assessment device, configured to determine the analyte level of the subject based on a first quantity and a second quantity.

[0156] Example 19: The system according to the foregoing embodiments, wherein the system includes at least one vapor distribution member configured to contact a sample from the subject’s exhaled breath or another gaseous emission source with the first volatile organic marker detection unit or the second volatile organic marker detection unit.

[0157] Example 20: The system according to any one of the foregoing two embodiments, wherein at least one of the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit is selected from the group consisting of: a gas sensor, exemplarily a metal oxide sensor; a carbon nanotube chip (CNT), specifically a functionalized carbon nanotube chip; a photometric sensor, exemplarily an absorption-based light source detector; a mass spectrometer, exemplarily an ion migration spectrometer; an electrochemical sensor; a biochemical sensor having a functionalized surface, wherein the functionalized surface may specifically include at least one of amino acids, peptides, proteins, DNA building blocks, other biomolecules, or fragments thereof.

[0158] Example 21: The system according to any one of the three preceding embodiments, wherein the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit can be placed at different locations on the subject's body.

[0159] Example 22: The system according to any one of the foregoing four embodiments, wherein the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit are connected to each other, specifically via wireless communication, preferably via wireless far-field communication, and more preferably via radio frequency transmission.

[0160] Example 23: The system according to any one of the preceding five embodiments, wherein at least one of the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit is configured to transmit data to at least one external device, specifically to at least one smartphone.

[0161] Example 24: The system according to any one of the preceding six embodiments, wherein at least one of the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit is disposed on one of a wristband, a patch such as one that can be placed on the skin of a subject via an adhesive, eyeglasses, nose clip, nose ring, nose stud, headband, headphones, clothing, pillow, or bedside table.

[0162] Example 25: According to the system described in the foregoing embodiments, at least one of the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit is arranged on or integrated into the smartwatch.

[0163] Example 26: A system according to any one of the foregoing two embodiments, wherein the first volatile organic compound (VOC) marker detection unit is disposed on the wristband, wherein the first VOC marker detection unit is configured to preferably discontinuously detect the first VOC marker from the first source of the subject, the first source being exhaled breath, wherein the system includes a vapor distribution member configured to contact a sample from the subject's exhaled breath or another gaseous emission source with the first VOC marker detection unit.

[0164] Example 27: The system according to any one of the preceding three embodiments, wherein the second volatile organic marker detection unit is disposed on the wristband, wherein the second volatile organic marker detection unit is configured to preferably continuously detect the first volatile organic marker from a second source of the subject, the second source being the subject's skin.

[0165] Example 28: The system according to any one of the foregoing four embodiments, wherein both the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit are arranged on the wristband.

[0166] Example 29: The system according to any one of the preceding eleven embodiments, wherein one of the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit is configured to preferably establish skin contact with the subject's skin continuously and / or permanently, and wherein the other of the first VOC marker detection unit and the second VOC marker detection unit is configured to be continuously and / or permanently or discontinuously arranged below the subject's nostrils or is capable of being arranged below the subject's nostrils. Attached Figure Description

[0167] Other optional features and embodiments will be disclosed in more detail in the following description of embodiments, in conjunction with the dependent claims. As those skilled in the art will recognize, each optional feature can be implemented individually and in any feasible combination. The scope of the invention is not limited to the preferred embodiments. Embodiments are schematically depicted in the accompanying drawings. In these drawings, the same reference numerals refer to the same or functionally equivalent elements.

[0168] In the attached diagram:

[0169] Figures 1A to 1D show scatter plots illustrating the correlation between measured blood glucose values ​​and predicted blood glucose values;

[0170] Figures 2A and 2B show further scatter plots (Figure 2A) and consistency error grids (Figure 2B) illustrating the correlation between measured and predicted blood glucose values.

[0171] Figures 3A and 3B show further scatter plots (Figure 3A) and further consistency error grids (Figure 3B) illustrating the correlation between measured and predicted blood glucose values.

[0172] Figures 4A and 4B show a summary of the results of further scatter plots demonstrating the correlation between measured blood glucose values ​​and predicted blood glucose values ​​(Figure 4A) and further consistency error grids (Figure 4B).

[0173] Figures 5A and 5B show further scatter plots (Figure 5A) and further consistency error grids (Figure 5B) illustrating the correlation between measured and predicted blood glucose values.

[0174] Figures 6A and 6B show further scatter plots (Figure 6A) and further consistency error grids (Figure 6B) illustrating the correlation between measured and predicted blood glucose values.

[0175] Figures 7A and 7B show two further scatter plots illustrating the correlation between measured and predicted blood glucose values; and

[0176] Figure 8 shows a summary of the results shown in Figures 2A, 3A, 4A, 5A, 6A, 7A, and 7B. Detailed Implementation

[0177] Figures 1A through 1D show scatter plots illustrating the correlation between measured and predicted blood glucose values. In Figures 1A through 1D, the measured blood glucose values ​​(c) in mg / dL are plotted on the y-axis. mFurthermore, in Figures 1A to 1D, the predicted blood glucose values ​​c in mg / dL are plotted on the x-axis. p A log-log model was applied.

[0178] Figure 1A shows a scatter plot where indole is applied as a volatile organic marker derived from the breath of the subject, for the mathematical model used to determine the predicted blood glucose value. Linear regression was performed, and the correlation coefficient R is 0.727. Therefore, there may be a significant correlation between the predicted and measured blood glucose values.

[0179] Figure 1B shows a scatter plot where indole is applied as a volatile organic biomarker derived from the subject's breath and skin, for the mathematical model used to determine predicted blood glucose levels. Linear regression was performed, and R² was 0.768. Therefore, by detecting indole from two different sources, the correlation between predicted and measured blood glucose levels increased.

[0180] Figure 1C shows a scatter plot where indole, methanol, and formaldehyde were used as volatile organic markers derived from the breath of the subjects for the mathematical model applied to determine predicted blood glucose levels. Linear regression was performed, and R = 0.860. Therefore, by detecting three different types of volatile organic markers, the correlation between predicted and measured blood glucose levels increased.

[0181] Figure 1D shows a scatter plot where indole, methanol, and formaldehyde were used as volatile organic markers derived from the subject's breath and skin, respectively, for the mathematical model applied to determine predicted blood glucose levels. Linear regression was performed, and R² was 0.920. Therefore, by detecting three different volatile organic markers from two different types of sources, the correlation between predicted and measured blood glucose levels was further increased.

[0182] Figure 2A shows a further scatter plot illustrating the correlation between measured and predicted blood glucose values, and Figure 2B shows the consistency error grid.

[0183] In Figure 2A, the measured blood glucose values ​​c, in mg / dL, are plotted on the y-axis. m In addition, the predicted blood glucose value c in mg / dL is plotted on the x-axis. p For mathematical models used to determine predicted blood glucose levels, indole is a volatile organic biomarker derived from the respiration of the subject.

[0184] An alert should be provided when blood glucose levels exceed 200 mg / dL. Therefore, three boxes are shown in Figure 2A. The lower left box is shown with a solid line. The upper left box is shown with a dashed line. The lower right box is shown with a dashed line. The study used 1250 blood glucose counts from 50 patients, with 25 observations per patient.

[0185] The bottom left box contains 757 counts corresponding to normal blood glucose levels of less than 200 mg / dL. The top left box contains 184 counts corresponding to false negative results. Therefore, incorrectly, no alarm was output. The bottom right box contains 62 counts corresponding to false positive results. Therefore, incorrectly, an alarm was output. The remaining 247 counts correspond to correct positive results.

[0186] Therefore, when the subjects were in a hyperglycemic state, the model's prediction accuracy was 57% (247 / (184+247)). Furthermore, the model's prediction accuracy for hyperglycemia was 80% (247 / (247+62)).

[0187] Figure 2B shows the consistency error grid, where indole is a volatile organic marker derived from the breath of the subjects. The number of blood glucose counts in the study was 1250, derived from 50 patients, with 25 observations per patient.

[0188] The consistent error grid, also known as the Parkes error grid, was developed as a novel tool for assessing the accuracy of blood glucose meters. In recent years, blood glucose meter manufacturers have increasingly used the consistent error grid in their clinical studies. The ISO 15197:2013 guideline specifies the use of the consistent error grid in the evaluation of blood glucose monitoring systems. In Zone A, there is generally no clinical impact. In Zone B, there is generally little or no impact on clinical outcomes. In Zone C, it generally may affect clinical outcomes. In Zone D, there may generally be significant medical risks. In Zone E, there may generally be dangerous consequences.

[0189] As shown in Figure 2B, area A contains 684 counts, area B contains 536 counts, area C contains 31 counts, and areas D and E each contain 0 counts.

[0190] Figure 3A shows a further scatter plot illustrating the correlation between measured and predicted blood glucose values, and Figure 3B shows a further uniformity error grid.

[0191] In Figure 3A, the measured blood glucose values ​​c, expressed in mg / dL, are plotted on the y-axis. mIn addition, the predicted blood glucose value c in mg / dL is plotted on the x-axis. p For mathematical models used to determine predicted blood glucose levels, indole is a volatile organic biomarker derived from the breath and skin of the subjects.

[0192] An alert should be provided when blood glucose levels exceed 200 mg / dL. Therefore, three boxes are shown in Figure 2A. The lower left box is shown with a solid line. The upper left box is shown with a dashed line. The lower right box is shown with a dashed line. The study used 1250 blood glucose counts from 50 patients, with 25 observations per patient.

[0193] The bottom left box contains 767 counts corresponding to normal blood glucose levels of less than 200 mg / dL. The top left box contains 154 counts corresponding to false negative results. Therefore, incorrectly, no alarm was output. The bottom right box contains 52 counts corresponding to false positive results. Therefore, incorrectly, an alarm was output. The remaining 272 counts correspond to correct positive results.

[0194] Therefore, when the subjects were in a hyperglycemic state, the model's prediction accuracy was 63% (272 / (159+272)). Furthermore, the model's prediction accuracy for hyperglycemia was 84% ​​(272 / (272+52)).

[0195] Figure 3B shows the consistency error grid, where indole is a volatile organic marker derived from the breath and skin of the subjects. The number of blood glucose counts in the study was 1250, derived from 50 patients, with 25 observations per patient.

[0196] As shown in Figure 3B, area A contains 740 counts, area B contains 484 counts, area C contains 26 counts, and areas D and E each contain 0 counts.

[0197] Figure 4A shows a summary of the results of a further scatter plot demonstrating the correlation between measured and predicted blood glucose values, and Figure 4B shows a further consistency error grid.

[0198] For the results shown in Figures 4A and 4B, indole is a volatile organic biomarker derived from the skin of the subjects, used in the mathematical model applied to determine predicted blood glucose values. The number of blood glucose counts in the study was 1250, derived from 50 patients, with 25 observations per patient.

[0199] An alarm should be issued when blood glucose levels are higher than 200 mg / dL. Therefore, four boxes are shown in Figure 4A. The lower left box contains 761 counts corresponding to normal blood glucose levels of less than 200 mg / dL. The upper left box contains 200 counts corresponding to false negative results. Therefore, incorrectly, no alarm is issued. The lower right box contains 58 counts corresponding to false positive results. Therefore, incorrectly, an alarm is issued. The remaining 231 counts (upper right box) correspond to correct positive results.

[0200] Therefore, when the subjects were in a hyperglycemic state, the model's prediction accuracy was 54% (231 / (231+200)). Furthermore, the model's prediction accuracy for hyperglycemia was 80% (231 / (231+58)).

[0201] Figure 4B shows the consistency error grid. As shown in Figure 4B, area A contains 651 counts, area B contains 567 counts, area C contains 32 counts, and areas D and E each contain 0 counts.

[0202] Figure 5A shows a further scatter plot illustrating the correlation between measured and predicted blood glucose values, and Figure 5B shows a further uniformity error grid.

[0203] In Figure 5A, the measured blood glucose values ​​c in mg / dL are plotted on the y-axis. m In addition, the predicted blood glucose value c in mg / dL is plotted on the x-axis. p For mathematical models used to determine predicted blood glucose levels, indole, formaldehyde, and methanol are volatile organic markers derived from the respiration of the subjects.

[0204] An alert should be provided when blood glucose levels exceed 200 mg / dL. Therefore, three boxes are shown in Figure 5A. The lower left box is shown with a solid line. The upper left box is shown with a dashed line. The lower right box is shown with a dashed line. The study used 1250 blood glucose counts from 50 patients, with 25 observations per patient.

[0205] The bottom left box contains 763 counts corresponding to normal blood glucose levels of less than 200 mg / dL. The top left box contains 116 counts corresponding to false negative results. Therefore, incorrectly, no alarm was output. The bottom right box contains 56 counts corresponding to false positive results. Therefore, incorrectly, an alarm was output. The remaining 315 counts correspond to correct positive results.

[0206] Therefore, when subjects were in a hyperglycemic state, the model predicted it to be 73% (315 / (315+56)). Furthermore, the model's prediction accuracy for hyperglycemia was 85% (315 / (315+116)).

[0207] Figure 5B shows the consistency error grid, where indole, formaldehyde, and methanol are volatile organic markers derived from the breath of the subjects. The number of blood glucose counts in the study was 1250, derived from 50 patients, with 25 observations per patient.

[0208] As shown in Figure 5B, area A contains 875 counts, area B contains 369 counts, area C contains 6 counts, and areas D and E each contain 0 counts.

[0209] Figure 6A shows a further scatter plot illustrating the correlation between measured and predicted blood glucose values, and Figure 6B shows a further consistent error grid.

[0210] In Figure 6A, the measured blood glucose values ​​c in mg / dL are plotted on the y-axis. m In addition, the predicted blood glucose value c in mg / dL is plotted on the x-axis. p For mathematical models used to determine predicted blood glucose levels, indole, formaldehyde, and methanol are volatile organic markers derived from the breath and skin of the subjects.

[0211] An alert should be provided when blood glucose levels exceed 200 mg / dL. Therefore, three boxes are shown in Figure 6A. The lower left box is shown with a solid line. The upper left box is shown with a dashed line. The lower right box is shown with a dashed line. The study used 1250 blood glucose counts from 50 patients, with 25 observations per patient.

[0212] The bottom left box contains 777 counts corresponding to normal blood glucose levels of less than 200 mg / dL. The top left box contains 85 counts corresponding to false negative results. Therefore, incorrectly, no alarm was output. The bottom right box contains 42 counts corresponding to false positive results. Therefore, incorrectly, an alarm was output. The remaining 346 counts correspond to correct positive results.

[0213] Therefore, when subjects were in a hyperglycemic state, the model predicted it to be 80% (346 / (346+86)). Furthermore, the model's prediction accuracy for hyperglycemia was 89% (346 / (346+42)).

[0214] Figure 6B shows the consistency error grid, where indole, formaldehyde, and methanol are volatile organic markers derived from the breath and skin of the subjects. The number of blood glucose counts in the study was 1250, derived from 50 patients, with 25 observations per patient.

[0215] As shown in Figure 6B, area A contains 1039 counts, area B contains 208 counts, area C contains 3 counts, and areas D and E each contain 0 counts.

[0216] Table 1 below summarizes the results for the consistent error grids shown in Figures 2B, 3B, 4B, 5B, and 6B.

[0217] Table 1: Summary of the results for the consistency error grid shown in Figures 2B, 3B, 4B, 5B, and 6B, with the number of counts in each zone given as total and percentage.

[0218]

[0219] This summary suggests that the more diverse the types and sources of volatile organic compounds considered, the higher the percentage of correctly predicted hyperglycemic events can be.

[0220] Figures 7A and 7B show further scatter plots illustrating the correlation between measured and predicted blood glucose values.

[0221] In Figures 7A and 7B, the measured blood glucose values ​​c, in mg / dL, are plotted on the y-axis. m In addition, the predicted blood glucose value c in mg / dL is plotted on the x-axis. p For the mathematical model applied to determine predicted blood glucose values, indole, formaldehyde, and methanol are volatile organic markers derived from the breath and skin of the subjects. For Figure 7A, the upper limit of the 50% confidence interval was considered. For Figure 7B, the upper limit of the 90% confidence interval was considered.

[0222] An alert should be provided when blood glucose levels exceed 200 mg / dL. Therefore, three boxes are shown in Figures 7A and 7B. The lower left box is shown with a solid line. The upper left box is shown with a dashed line. The lower right box is shown with a dashed line. The study used 1250 blood glucose counts from 50 patients, with 25 observations per patient.

[0223] In Figure 7A, the lower left box includes 691 counts corresponding to normal blood glucose levels of less than 200 mg / dL. The upper left box includes 29 counts corresponding to false negative results. Therefore, incorrectly, no alarm was output. The lower right box includes 128 counts corresponding to false positive results. Therefore, incorrectly, an alarm was output. The remaining 402 counts correspond to correct positive results. Therefore, when the subject was hyperglycemic, the model predicted it with 93% accuracy (402 / (402+29)). Furthermore, the model's prediction accuracy for hyperglycemia was 76% (402 / (402+129)).

[0224] In Figure 7B, the lower left box includes 490 counts corresponding to normal blood glucose values ​​of less than 200 mg / dL. The upper left box includes 0 counts corresponding to false negative results. Therefore, incorrectly, no alarm was output. The lower right box includes 329 counts corresponding to false positive results. Therefore, incorrectly, an alarm was output. The remaining 431 counts correspond to correct positive results. Therefore, when the subject was in a hyperglycemic state, the model predicted it to be 100% (431 / (431+0)). Furthermore, the model's prediction accuracy for hyperglycemia was 57% (431 / (431+329)).

[0225] Figure 8 summarizes the results shown in Figures 2A, 3A, 4A, 5A, 6A, 7A, and 7B. Specifically, Figure 8 shows the percentage of hyperglycemic events, where blood glucose concentrations were greater than the correctly predicted 200 mg / dL in a study with 50 patients.

[0226] The arrow marked "1" indicates the results shown in Figure 4A. Therefore, when indole is a volatile organic marker derived from the skin, which is the source for the subject, the percentage of correctly predicted hyperglycemic events is 54% for the mathematical model applied to determine predicted blood glucose values.

[0227] The arrow marked "2" indicates the results shown in Figure 2A. Therefore, when indole is a volatile organic marker derived from the breath of the subject, the percentage of correctly predicted hyperglycemic events is 57% for the mathematical model applied to determine predicted blood glucose values.

[0228] The arrow marked "3" indicates the results shown in Figure 3A. Therefore, when indole is a volatile organic marker derived from the skin and respiration of the subject, the percentage of correctly predicted hyperglycemic events is 63% for the mathematical model applied to determine predicted blood glucose values.

[0229] The arrow marked "4" indicates the results shown in Figure 5A. Therefore, when indole, formaldehyde, and methanol are volatile organic markers derived from the breath of the subjects, the percentage of correctly predicted hyperglycemic events is 73% for the mathematical model applied to determine predicted blood glucose values.

[0230] The arrow marked "5" indicates the results shown in Figure 6A. Therefore, when indole, formaldehyde, and methanol are volatile organic markers derived from the breath and skin of the subjects, the percentage of correctly predicted hyperglycemic events is 80% for the mathematical model used to determine predicted blood glucose values.

[0231] The arrow marked "6" indicates the results shown in Figure 7A. Therefore, when indole, formaldehyde, and methanol are volatile organic markers derived from the breath and skin of the subjects, the percentage of correctly predicted hyperglycemic events is 93% when applied to the mathematical model for determining predicted blood glucose values, and when the upper limit of the 50% confidence interval is applied.

[0232] The arrow marked with "7" indicates the results shown in Figure 7B. Therefore, when indole, formaldehyde, and methanol are volatile organic markers derived from the breath and skin of the subjects, the percentage of correctly predicted hyperglycemic events is 100% when applied to the mathematical model for determining predicted blood glucose values ​​and when the upper limit of the 90% confidence interval is applied.

[0233] The summary also indicates that the more diverse the types and sources of volatile organic biomarkers considered, the higher the percentage of correctly predicted hyperglycemic events can be. Further improvements can be achieved by considering confidence intervals.

Claims

1. A method for determining the analyte level in a subject, the method comprising the steps of: a) Non-invasively detect a first amount of at least one first volatile organic marker originating from a first source of the subject; b) Non-invasively detect a second amount of at least the first volatile organic marker originating from a second source different from the first source; as well as c) Determine the analyte level of the subject based on the first and second quantities.

2. The method according to the preceding claim, wherein the first source is exhaled air, and wherein the second source is another gaseous emission source different from exhaled air, or vice versa.

3. The method according to any one of the preceding claims, wherein the analyte is glucose.

4. The method according to any one of the preceding claims, wherein the first volatile organic marker is selected from a group of markers in which the amount of the volatile organic marker is negatively correlated with glucose levels.

5. The method according to any one of the preceding claims, wherein the first volatile organic marker is one of indole (C8H7N), a partially saturated derivative of indole, a fully saturated derivative of indole, and a true fraction of indole.

6. The method according to any one of claims 1 to 3, wherein the first volatile organic marker is selected from the group consisting of formaldehyde and methanol.

7. The method according to any one of the preceding claims, wherein the method further comprises: d) Non-invasively detect a third amount of at least a second volatile organic marker derived from the first source of the subject; as well as e) Non-invasively detect a fourth amount of at least one second volatile organic marker derived from the second source of the subject; In step c), the analyte level of the subject is detected based on the first amount, the second amount, the third amount, and the fourth amount.

8. The method according to claims 1 to 3, wherein the method further comprises: d) Non-invasively detect a third amount of at least a second volatile organic marker derived from the first source of the subject; as well as e) Non-invasively detect a fourth amount of at least one second volatile organic marker derived from the second source of the subject; In step c), the analyte level of the subject is detected based on the first amount, the second amount, the third amount, and the fourth amount, wherein the first volatile organic marker is indole, and the second volatile organic marker is formaldehyde or methanol.

9. The method according to any one of claims 7 or 8, wherein the method further comprises: f) Non-invasively detect a fifth amount of at least a third volatile organic marker derived from the first source of the subject; as well as g) Non-invasively detect a sixth amount of at least one third volatile organic marker derived from the second source of the subject; In step c), the analyte level of the subject is detected based on the first, second, third, fourth, fifth, and sixth quantities.

10. The method of claim 8, wherein the method further comprises: f) Non-invasively detect a fifth amount of at least a third volatile organic marker derived from the first source of the subject; as well as g) Non-invasively detect a sixth amount of at least one third volatile organic marker derived from the second source of the subject; In step c), the analyte levels of the subject are detected based on the first amount, the second amount, the third amount, the fourth amount, the fifth amount, and the sixth amount, wherein the first volatile organic marker, the second volatile organic marker, and the third volatile organic marker are formaldehyde, methanol, and indole.

11. The method according to any one of the preceding claims, wherein in step a), the first amount of the at least one first volatile organic marker originating from the first source of the subject is detected continuously, and wherein in step b), the second amount of the at least one first volatile organic marker originating from the second source of the subject is detected discontinuously.

12. A system for determining the analyte level in a subject, wherein the system is configured to perform the method according to any one of claims 1 to 6, wherein the system comprises: ● A first volatile organic compound (VOC) marker detection unit, configured to non-invasively detect the first amount of the at least one first VOC marker originating from the first source of the subject; ● A second volatile organic compound (VOC) marker detection unit, configured to non-invasively detect the second amount of the at least one first VOC marker originating from the subject's second source; as well as ● At least one assessment device, the at least one assessment device being configured to determine the analyte level of the subject based on the first quantity and the second quantity.

13. The system according to the preceding claims, wherein the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit can be placed at different locations on the subject's body.

14. The system according to any one of the preceding two claims, wherein the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit are connectable to each other, specifically via wireless communication, preferably via wireless far-field communication, wherein at least one of the first VOC marker detection unit and the second VOC marker detection unit is configured to transmit data to at least one external device.

15. The system according to any one of the preceding three claims, wherein one of the first volatile organic compound (VOC) marker detection unit and the second VOC marker detection unit is configured to establish skin contact with the subject's skin, and wherein the other of the first VOC marker detection unit and the second VOC marker detection unit is configured to be or can be disposed below the subject's nostrils or below the subject's mouth.

Images