Method and device for monitoring disease

Abstract

A method for determining a measure of a disease in a region of a body of a subject is provided. The method includes receiving imaging data obtained from imaging a region of a body of a subject, the imaging data including data representative of a distribution of a marker administered within the region of the subject prior to imaging. The marker binds to a biological target associated with a disease. The imaging data is processed to obtain a metric of the marker signal in the region. The metric of the marker signal is corrected for the effect of the marker signal of the tissue structure within the region. A measure of the disease of the region is determined using the corrected measure of the marker signal. An apparatus configured to perform the method is also provided.

Description

Technical Field

[0001] The present invention relates to medical imaging, particularly medical imaging for disease monitoring. Background Art

[0002] Within the healthcare industry, there is a need for better methods to identify which drugs a patient is likely to respond to before starting treatment. Currently, over 50% of disease-modifying agents fail in clinical practice because patient selection based on blood biomarkers is suboptimal. When a patient presents with a disease, a doctor may prescribe a specific drug and have to wait several months to see if it effectively controls the condition. If the drug does not work, the doctor will prescribe a different drug and repeat the process until the best one is found. This wastes valuable time and money in the healthcare industry. As the cost of new drugs continues to rise, this becomes increasingly important. Additionally, within the pharmaceutical industry, there is a need for better methods to stratify subjects in clinical trials of new drugs such as anti-inflammatory or immunomodulatory therapies.

[0003] For example, chronic inflammation is a core process in the pathophysiology of some of the leading causes of morbidity and mortality worldwide. Cardiovascular and cerebrovascular diseases are increasingly regarded as inflammatory diseases and are the leading causes of death worldwide, followed by chronic obstructive pulmonary disease (COPD), which is also characterized by chronic activation of inflammatory pathways, usually due to repeated inhalation of toxins. In addition, autoimmune inflammatory diseases such as rheumatoid arthritis and inflammatory bowel disease are also highly prevalent chronic inflammatory diseases, causing significant morbidity worldwide, particularly in Europe and North America. Chronic inflammatory diseases are a major burden on healthcare services and there is an urgent need for new disease-modifying therapies. With the introduction of primary and secondary prevention strategies, the outcomes of cardiovascular diseases have improved, but the development of further treatment interventions is still ongoing.

[0004] With the introduction of monoclonal antibody therapies in diseases such as rheumatoid arthritis and Crohn's disease, there have been recent advancements in the treatment of inflammatory diseases. Anti-TNF-α therapies such as infliximab and other biologics including anti-CD20 (rituximab) and anti-IL-6 (tocilizumab or sarilumab) have been introduced as part of the established treatment regimens for rheumatoid arthritis. In lung diseases, advancements in new therapies for asthma have provided monoclonal antibody therapies for selected patient populations with clear evidence of benefit, but the same progress has not been seen in the search for disease-modifying therapies for COPD. Studies targeting anti-IL-5, anti-IL-5Rα, and anti-TNF-α therapies have failed to show significant benefits in the COPD population.

[0005] Heterogeneity within the COPD disease group may be a key factor in the apparent lack of efficacy. The current view is that specific inflammatory phenotypes or endotypes will respond to specific therapies. Similarly, although anti-TNF-α antibodies are an established therapy for rheumatoid arthritis, up to 40% of patients fail to show a long-term response to the therapy due to heterogeneity in the population, and there is currently no available method to predict which patients will respond. 1 , because patient genotyping based on blood biomarkers or clinical characteristics is suboptimal. Treatment of inflammatory bowel disease is also often ineffective.

[0006] Given the few current therapies and interventions that affect the disease outcomes of patients with airway diseases, and the difficulty in selecting appropriate therapies for individuals in other inflammatory diseases, there is a need to better understand the complex immune processes involved. Thus, the key scientific challenge is to better understand the mechanisms involved in these inflammatory pathways and to identify targets for treatment.

[0007] Technologies are now available that allow the use of single-photon emission computed tomography (SPECT) and positron emission tomography (PET) technologies for immune imaging of tissues and organs. These technologies utilize radiopharmaceuticals, such as radiolabeled fluorodeoxyglucose ( 18 F-FDG) or radiolabeled monoclonal antibodies, to visualize processes, cells, or cytokines associated with the immune response. These nuclear medicine technologies have previously been used in the field of oncology to identify, localize, and evaluate the functional status of various cancers 5,6 . This form of imaging allows researchers to visualize active inflammation and pathologic processes in a way that was previously only achievable through tissue biopsy, but which can cover a larger area and does not require invasive procedures 2-4 .

[0008] Extending the same technologies to imaging of the immune system not only provides opportunities for studying diseases such as chronic inflammatory processes in vivo, but also allows prediction of responses to specific therapies, particularly monoclonal antibody therapies. New functional imaging of the immune system can provide means to reveal new knowledge about these processes and has the potential to serve as a tool for predicting responses to disease-modifying therapies.

[0009] A number of early studies have been conducted across a range of chronic inflammatory diseases, some of which have shown that increased uptake of radiolabeled antibodies observed in imaging of inflamed tissues is associated with improved response to monoclonal antibody therapies or indications for systemic therapies 7-9 . However, these studies are far from conclusive, and methods and systems need to be further developed to reliably identify disease markers, such as inflamed tissues, and to evaluate the efficacy of disease treatment. Summary of the Invention

[0010] According to a first aspect, there is provided a method for determining a measure of a disease in a region of a subject's body, comprising: receiving imaging data obtained from imaging a region of the subject's body, the imaging data including data representing the distribution of a marker within the region of the subject to which the marker was administered prior to imaging, wherein the marker binds to a disease-related biological target; processing the imaging data to obtain a measure of the marker signal in the region; correcting the measure of the marker signal for the effect of the marker signal on the tissue structure within the region; and using the corrected measure of the marker signal to determine a measure of the disease in the region.

[0011] In vivo molecular imaging provides the potential to assess diseases in a non-invasive manner, for example, by quantifying the drivers of the inflammatory process in individuals with chronic inflammatory diseases. Correction for the effect of signals on tissue structure eliminates confounding factors that make it difficult to distinguish signals caused by inflammation from signals caused by other factors. This provides a new approach for precision medicine, enabling clinicians to select for the first time disease-modifying therapies that are effective for individual patients without invasive procedures.

[0012] Optionally, the corrected measure of the marker signal represents the level of binding of the marker to the tissue within the region. Markers that bind to tissue are more likely to represent the quantity of biological targets in the tissue compared to markers that are present in the region for other reasons.

[0013] Optionally, the disease is an inflammatory disease and the biological target is a component of an inflammatory pathway. Markers that bind to components of the inflammatory pathway have been tested and represent a promising and valuable application of the technology.

[0014] Optionally, correcting the measure of the marker signal includes correcting for the marker signal caused by blood flow in the region. The marker is typically administered into the patient's blood. This means that free markers in the blood can produce significant signals. Additionally, regions with higher perfusion due to greater blood flow may exhibit higher levels of bound markers simply because more markers are available to interact with the tissue. Thus, the signal in the region can be distorted by the amount of vasculature present. Correction for this can provide a better understanding of how much biological target is present relative to other parts of the region.

[0015] Optionally, correcting for the marker signal caused by blood flow includes determining the vascular density in the region. The vascular density can be used to estimate the proportion of the signal in the region that is due to free markers in the blood rather than markers bound to biological targets.

[0016] Optionally, determining vascular density includes determining the total volume of blood vessels in a region. This can be used to estimate how much signal needs to be corrected for total blood flow in the region.

[0017] Optionally, the imaging data further includes data representing the tissue structure in the region, and determining vascular density includes using the data representing the tissue structure to identify blood vessels in the region. Using another source of structural data, such as from another imaging modality that is more suitable for vascular imaging, can improve the accuracy of vascular density determination.

[0018] Optionally, the data representing the tissue structure can be obtained by one or more of low-dose computed tomography, high-resolution computed tomography, multislice computed tomography, magnetic resonance imaging, plain film radiography, or ultrasound. These imaging modalities may be particularly suitable for determining the tissue structure within a region.

[0019] Optionally, correcting the marker signal due to blood flow in the region includes determining a background marker signal corresponding to the marker signal of the unbound marker in the blood. This allows the signal level caused by the free marker in the subject's blood to be taken into account during measurement and a better measure of relative marker activity to be obtained.

[0020] Optionally, correcting the measure of the marker signal includes correcting for the tissue density in the region. Variations in tissue density in the region may give a false impression of the disease level in the tissue. Even at the same disease level, denser tissue will naturally take up more marker per unit volume than less dense tissue. This can lead to a false estimate that the tissue is more diseased when it is simply denser. Correction for this effect can yield a more accurate estimate of the extent of the disease.

[0021] Optionally, the imaging data further includes data representing the tissue structure in the region, and correcting for tissue density includes using the data representing the tissue structure to identify sub-regions in the region where the tissue density is reduced. Identifying sub-regions where the tissue density is reduced can allow for appropriate correction due to the sub-regions.

[0022] Optionally, the data representing the tissue structure is obtained by one or more of low-dose computed tomography, high-resolution computed tomography, multislice computed tomography, magnetic resonance imaging, plain film radiography, and ultrasound. These imaging modalities are particularly suitable for determining structural information.

[0023] Optionally, data representing tissue structure is obtained using X-rays, and a sub-region is identified as a sub-region with reduced tissue density when the attenuation of the X-rays in the sub-region is lower than a predetermined threshold. A decrease in the attenuation of X-rays is a known method for identifying reduced tissue density, particularly in a respiratory environment. Therefore, using this method can facilitate integration with existing systems and processes.

[0024] Optionally, the correction for tissue density includes correcting the measure of the marker signal based on the proportion of tissue in the region with reduced density. This enables the method to account for signal reduction not due to a decrease in disease level.

[0025] Optionally, correcting the measure of the marker signal based on the proportion of tissue in the region with reduced density is determined using the measures of the marker signal in multiple individuals with different tissue proportions in the region with reduced density. This enables appropriate calibration of the degree of influence of changes in tissue density on the marker signal.

[0026] Optionally, the measure of the marker signal is the average marker signal in the region, optionally the median marker signal. Using the average marker signal allows for a more precise measurement of the signal in the entire region. Compared to other types of average signals, the median signal is less distorted by isolated extreme values.

[0027] Optionally, the measure of the marker signal is normalized by a background marker signal corresponding to the marker signal of the unbound marker in the blood. Normalizing the measure by the background signal enables a better measurement of the relative marker activity in the region.

[0028] Optionally, the background marker signal is identified from the marker signal in the major blood vessel (optionally the aortic arch). The signal from the major blood vessel may be dominated by the signal caused by the unbound marker in the blood, and thus becomes a useful source of the background signal. Optionally, the background marker signal is the average marker signal in the major blood vessel, optionally the median marker signal. Averaging the blood vessel will give a more representative measure of the background activity.

[0029] Optionally, the biological target includes immune and / or inflammatory signaling proteins, such as cytokines, chemokines, cell surface receptors, or extracellular matrix components. These can all indicate diseases in the body and can thus be used to determine a measure of the disease.

[0030] Optionally, the marker includes a radioactive marker, and / or the marker includes a monoclonal antibody. Monoclonal antibodies can be engineered to have a very specific binding effect on specific components of the target, and radioactive markers can be easily detected using non-invasive imaging techniques.

[0031] Optionally, data representing the distribution of markers in a region is obtained using single photon emission computed tomography, SPECT, positron emission tomography, PET, planar scintigraphy, or magnetic resonance imaging. These techniques are all capable of providing adequately resolved imaging of various types of markers.

[0032] Optionally, the body region of the subject includes an organ of the subject, such as one or both lungs of the subject, or a joint of the subject. These are regions of particular interest in common pathological processes such as inflammation and may be of particular interest in the diagnosis and evaluation of treatment regimens of patients.

[0033] Optionally, the imaging data includes data representing the distribution of each of a plurality of markers administered to the subject prior to imaging in a region, wherein each of the plurality of markers binds to a different biological target; processing the imaging data includes processing the imaging data to obtain a measure of the marker signal for each of the plurality of markers; correcting the measure of the marker signal includes correcting the measure of each marker signal; and determining a measure of the disease uses a plurality of corrected marker signal measures. By combining multiple markers, different biological targets can be identified. In addition, the efficacy of different therapeutic targets targeting different biological targets can be evaluated simultaneously. This improves the speed and convenience of patient diagnosis.

[0034] Optionally, the data representing the tissue structure and the data representing the marker distribution have different resolutions and / or volume segmentations, are measured with the subject's body in different states, and / or are measured at different times; and the method further includes aligning the data representing the tissue structure and the data representing the marker distribution. This can allow different measurements to be performed continuously, allowing the patient to rest to improve comfort. This can also allow the synthesis of data obtained by different instruments that may have different specifications.

[0035] Optionally, the alignment includes using a non-rigid registration algorithm. This allows the alignment to take into account deformations of the region of the body, such as the expansion of the lungs during breathing.

[0036] Optionally, the alignment includes warping the data representing the tissue structure to match the data representing the marker distribution. Warping the data representing the marker distribution is more likely to result in distortion of the measure of the marker signal, so it is preferred to match the data representing the tissue structure to the data representing the marker distribution rather than vice versa.

[0037] According to a further aspect, there is provided a computer program comprising instructions which, when executed by a computer, cause the computer to perform the method.

[0038] According to a further aspect, there is provided a computer-readable medium including instructions which, when executed by a computer, cause the computer to perform the method.

[0039] According to a further aspect, there is provided an apparatus for determining a measure of inflammation in a region of a subject's body, the apparatus including a processor configured to perform the steps of the method. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Embodiments of the present invention will now be described, by way of non-limiting example only, with reference to the accompanying drawings, in which corresponding reference numerals represent corresponding parts, and in which, according to some embodiments of the present disclosure:

[0041] Figure 1 is a flowchart showing a method for determining a measure of a disease;

[0042] Figure 2 is a flowchart showing further details of calibrating a measure of a marker signal;

[0043] Figure 3a and Figure 3b shows the alignment of data representing tissue structure and data representing the distribution of a marker;

[0044] Figure 4 is a detailed flowchart of a process used in a specific implementation;

[0045] Figure 5a and Figure 5b is a box plot of median counts at different time points;

[0046] Figures 6a to 6e shows the normalized median counts at different time points;

[0047] Figure 7 shows the difference in blood vessel density calculated using scans at different time points;

[0048] Figure 8a and Figure 8b shows the effect of blood vessel density on median normalized counts;

[0049] Figures 9a to 9c shows the effect of reduced tissue density on median normalized counts;

[0050] Figure 10a and Figure 10b is a box plot of median normalized counts corrected for the effect of blood flow;

[0051] Figures 11a to 11d is a box plot of median normalized counts corrected for the effect of reduced tissue density;

[0052] Figure 12a and Figure 12b shows biomarker signals in the right lungs of two study participants;

[0053] Figure 13 is a flowchart of an exemplary software platform workflow; and

[0054] Figure 14 a and Figure 14 b confirm the potential implementation of the method as a clinical platform. DETAILED DESCRIPTION

[0055] New biotherapies are being used increasingly widely for various diseases, including cancer and inflammatory diseases (such as asthma or rheumatoid arthritis). These therapies can be used both to treat diseases and for diagnostic applications, such as cancer screening, and to identify tumor types or infectious agents, without the need for invasive biopsies.

[0056] Biotherapies can be very effective, but not every patient responds to the same treatment in the same way. Biologics, such as monoclonal antibodies, target specific branches of the pathways driving the disease, and their importance varies from patient to patient. These drugs are expensive, so there is an urgent need for a method to determine which biologic is most suitable for a particular patient in order to provide effective treatment at the first time. In addition to reducing the associated healthcare costs, this would significantly reduce the disease burden on patients by ensuring that they receive the most effective treatment as soon as possible.

[0057] In vivo molecular imaging offers the possibility of evaluating the driving biological processes in individuals with diseases (such as chronic inflammatory diseases). This new approach to precision medicine has the potential to enable clinicians to select disease-modifying therapies and to select the correct therapy at the first time without the need for invasive surgery. However, the main limitation in translating molecular imaging into the clinic is the high complexity of the required image analysis, as well as the ability to quantify the biological signals driven by the abundance of biological targets (such as inflammatory cytokines or molecular immune targets in organs), compared to anatomical medical imaging (such as X-rays, MRI or CT 16 ).

[0058] The present method allows the quantification, visualization and understanding of the biology of in vivo diseases using non-invasive methods. The method can form the basis of an imaging platform capable of determining the activation of specific biological pathways within organ systems. The method also allows the prediction of, or response to, treatment in a clinical setting to optimize the selection of the specific therapy to which an individual will respond.

[0059] Figure 1It is a flowchart showing a method for determining a measure of a disease in a body region of a subject. For example, the disease may be an inflammatory disease and the measure is a measure of inflammation. Alternatively, the disease may be an infection, cancer, or a specific type of cancer.

[0060] The method includes receiving S10 imaging data 10 obtained from imaging a region of the subject's body. The region can be any region of interest where the disease level is to be determined or the treatment effect is to be monitored. For example, the region of the subject's body may include an organ of the subject. This can be, for example, one or both lungs of the subject or a joint of the subject (such as a knee joint or a knuckle joint). Monitoring the lungs may be desirable for evaluating COPD, while monitoring the joints may be desirable for evaluating rheumatoid arthritis (RA).

[0061] The imaging data 10 includes data representing the distribution of a marker within the region of the subject that was administered prior to imaging.

[0062] The data representing the distribution of the marker within the region can be obtained using any suitable imaging technique capable of detecting the marker administered to the subject. For example, the data representing the distribution of the marker within the region can be obtained using any one of single photon emission computed tomography (SPECT), positron emission tomography (PET), planar scintigraphy, or magnetic resonance imaging (MRI). Combinations of these imaging modalities can also be used, such as SPECT-CT, PET-CT, PET-MRI, SPECT-MRI. These options are advantageous because they utilize established imaging methods that are familiar to clinicians and are already available in many settings.

[0063] Optionally, as further discussed below, the imaging data 10 may further include data representing the tissue structure in the region. The data representing the tissue structure can be obtained by one or more of low-dose computed tomography, high-resolution computed tomography, multislice computed tomography, MRI, plain film radiography, ultrasound, and Doppler ultrasound. Any suitable combination of methods can be used to obtain the data representing the tissue structure and the data representing the distribution of the marker. The data representing the tissue structure can be used to determine the volume of the region of the body, particularly if the imaging data covers more of the body than the region for which the measure of the disease is to be determined.

[0064] The choice of marker will depend on the imaging technique used to obtain the data representing the distribution of the marker. For example, the marker can include a radioactive marker (also referred to as a tracer or a radiotracer). When using MRI, the marker can include a magnetic marker or an MRI contrast agent.

[0065] Data representing the distribution of markers within a region may include measurements for each of a plurality of voxels within the region. For example, when the marker includes a radiolabel, the measurement may be the activity count for each voxel over a predetermined time period.

[0066] Bind to a disease-related biological target and a marker. The biological target may be a component of a biological pathway (such as a biological pathway known to be related to a disease). For example, when the disease is an inflammatory disease, the biological target may be a component of the inflammatory pathway. When the disease is cancer, the biological target may be a cancer protein, such as a protein associated with a specific type of cancer to be detected and / or treated. When the disease is an infection, the biological target may be an infectious pathogen, such as a virus or bacterium.

[0067] Biological targets may include immune and / or inflammatory signaling proteins, such as cytokines, chemokines, cell surface receptors, or extracellular matrix components. A specific example used in the following examples is tumor necrosis factor TNF. TNF-α is a pro-inflammatory cytokine released by macrophages and airway epithelial cells in the lungs of COPD patients. It is known that the level of TNF-α in the sputum of COPD patients is elevated 14 , and TNF-α is crucial for many inflammatory features, such as the secretion of matrix metalloproteinases (MMPs) and fibroblast proliferation. TNF-α may also be associated with cachexia and weight loss that are part of the disease.

[0068] Markers may include monoclonal antibodies. Monoclonal antibodies are useful for binding to very specific targets. For example, when evaluating COPD, the marker may include an anti-TNF-α monoclonal antibody.

[0069] The method includes processing S20 imaging data to obtain a measure of the marker signal in the region. The measure of the marker signal may be a measure representing the entire region. Wherein, the data representing the distribution of markers within the region includes measurements for a plurality of voxels within the region, obtaining the measure of the marker signal may include combining measurements from two or more voxels. The measure of the marker signal may be the average marker signal in the region, optionally the median marker signal. Obtaining the measure of the marker signal may further include other standard techniques and corrections applicable to the imaging modality for obtaining the data. For example, when using an imaging modality with a spatial resolution lower than or similar to the feature size being measured, partial volume effects may be corrected. For example, when using PET or SPECT, this type of correction may be used.

[0070] The measure of the marker signal may be normalized by a background marker signal corresponding to the marker signal of the unbound marker in the blood. The use of the target-to-background ratio has been described in previous literature 8,9This can be used to normalize the biodclearance of the marker when measurements are taken at different time points after the marker has been administered to the subject, and / or to allow comparison between different subjects who may have different biodclearance rates.

[0071] The background marker signal can be identified from the marker signal in the major blood vessel (optionally the aortic arch). The background marker signal can be determined using the voxels corresponding to the major blood vessel in the data representing the distribution of the marker within the region. If the major blood vessel is not within the region where the imaging data is acquired, other data can be used to determine the background marker signal, but this should be measured as close in time as possible to the imaging data. The background marker signal can be the average marker signal in the major blood vessel, optionally the median marker signal. The background marker signal can be expressed per unit volume.

[0072] The method includes correcting the measure of the S30 marker signal for the effect of the marker signal on the tissue structure within the region. The corrected measure of the marker signal can represent the level of binding of the marker to the tissue within the region. In this way, the correction determines the "tissue-bound" marker as a meaningful measure of inflammatory activity. The step of correcting the measure of the S30 marker signal will be discussed in further detail below.

[0073] The method includes using the corrected measure of the marker signal to determine a measure of the disease in the S40 region. The measure of the disease can represent the level of the disease in the region. For example, when the disease is an inflammatory disease, the measure of the disease can represent the level of inflammation in the region. The measure of the disease can represent the amount of the biological target. The corrected measure of the marker signal can be directly used as the measure of the disease. Optionally, the corrected measure of the marker signal can be further processed to obtain the measure of the disease, such as normalizing it relative to a standard amount or average amount of the biological target calculated, for example, with respect to measurements from healthy subjects.

[0074] Correcting the measure of the S30 marker signal can include correcting for the marker signal due to blood flow in the region. This allows the method to correct for the vascular density / perfusion of the lungs in the region.

[0075] The present inventors have found that the number of blood vessels per unit volume (or blood vessel density) is related to the signal detected within the region. Since the unbound markers in the blood provide an additional signal, markers with a higher background blood flow through the tissue in the region can produce a higher marker signal in the blood. This additional signal reduces the usefulness of the measure of the uncorrected marker signal as a measure of disease. The additional signal increases the measure of the marker signal without indicating any higher level of the biological component. In addition, even if the potential disease level in the tissue is similar, a higher level of available markers can cause more markers to bind locally to the tissue compared to regions with lower blood flow. Therefore, correcting for the marker signal caused by blood flow enables the method to provide a more accurate measure of disease.

[0076] Figure 2 is a flowchart showing further details of the implementation of the step of correcting the measure of the S30 marker signal for the effect of the marker signal on the tissue structure within the region. Step S110 and step S120 correspond to correcting the marker signal generated by blood flow in the region. Step S210 and step S220 correspond to correcting the marker signal generated by tissue density in the region. After step S110, step S120, step S210, and step S220, the method includes applying the correction to the S230 marker signal.

[0077] In Figure 2 it is shown that step S110 and step S120 are carried out together with step S210 and step S220. However, alternatively, step S110 and step S120 can be carried out before or after step S210 and step S220, or step S110 and step S120 can be completely omitted. After step S120 and step S220, step S230 of applying the correction to the marker signal is carried out. This reflects the fact that step S230 of applying the correction to the marker signal can use one or both of the corrections for blood flow and tissue density depending on the specific implementation.

[0078] In Figure 2 the implementation, the correction for the marker signal generated by blood flow includes determining the blood vessel density in the S110 region. Determining the S110 blood vessel density can include determining the total volume of blood vessels in the region. The blood vessel density can be defined as the total volume of blood vessels in the region divided by the total volume of the region.

[0079] The blood vessel density can be determined by any suitable method based on the imaging data 10 and / or other data sources about the patient, such as known or expected characteristics of undetected microvessels. In particular, the imaging data 10 can further include data representing the tissue structure in the region. In this case, determining the S110 blood vessel density includes using the data representing the tissue structure to identify the blood vessels in the region.

[0080] Blood vessels can be identified using automated segmentation techniques applied to data representing tissue structures. Automated segmentation can also be used to determine the volume of a body region from data representing tissue structures. The automated segmentation techniques can be implemented using machine learning algorithms. The automated segmentation techniques can identify structural features of tissue, such as blood vessels and fissures in the lungs, and can provide a depiction of the vascular tree within a region. The automated segmentation techniques can utilize prior shape models to detect anatomical regions of a particular organ or region of the body, and / or assist in identifying the vascular system in imaging data 10.

[0081] Determining the S110 vascular density can also consider data from other sources, such as scans of previously excised tissue specimens from the same region of the body (which can be taken from the subject or from other individuals). This data can provide further information for predicting microvascular density that may not be detectable in the subject's clinical imaging. This data can be incorporated into the determination of vascular density, for example, by providing an estimate of the usually additional volume of blood vessels that may not be distinguishable from the separate imaging data 10 alone.

[0082] In Figure 2 the implementation, the correction for the marker signal due to blood flow further includes determining S120 the background marker signal corresponding to the marker signal of the unbound marker in the blood. The background marker signal can be obtained as described above, for example, using the marker signal in major blood vessels such as the aortic arch.

[0083] Once the background signal and vascular density are determined, they can be combined to estimate the signal due to the unbound marker in the blood vessels. It can be assumed that the proportion of the area occupied by the blood vessels is equal to the vascular density and multiplied by the background signal to determine the expected signal due to the blood vessels.

[0084] In Figure 2 the implementation, applying the correction S230 to the marker signal includes subtracting the signal generated by the blood vessels from the measure of the marker signal. The specific implementation is provided by the following formula:

[0085]

[0086] where N T ′ is the normalized measure of the marker signal corrected for blood flow, M Ao is the background signal, D v is the vascular density within the body region, M T is the measure of the marker signal, which in this case is the average (median) of the marker signal in the region.

[0087] In this case, the background signal MAo is a value per unit volume, such that

[0088]

[0089] where R Ao is the raw marker signal within the region used to define the background (e.g., aortic arch), and V Ao is the volume of the region used to define the background. As described above, a normalized measure N of the marker signal normalized by the background signal can be used T , i.e., Thus, Equation (1) can be simplified to

[0090] N T ′ = N T - D v (3)

[0091] Expression (3) can be explained by considering the hypothetical case where the entire region is filled with blood and no tissue is present. If this were the case, the measure of the marker signal from the region would be expected to equal the background signal, and thus the normalized measure N of the marker signal T would be 1. In this case, the vessel density D v would also be 1 (since 100% of the region is occupied by blood), and thus the measure of the corrected marker signal would be zero.

[0092] The measure of the corrected S30 marker signal can include a correction for tissue density in the region.

[0093] In some experiments, the presence of emphysema in COPD patients was identified as a possible factor affecting the measure of the marker signal. On this basis, the measure of the corrected S30 marker signal can include a correction for tissue density in the region, such as changes caused by the presence of emphysema. Upon investigation, the presence of emphysema was considered a confounding factor. A strong correlation was found between the marker signal and the degree of emphysema.

[0094] Without wishing to be bound by theory, it is assumed that in regions of reduced tissue density, the marker cannot bind easily or effectively to the biological target, and the blood flow through the tissue is also affected. The correction for tissue density is considered to represent a composite factor including loss of small vessels, loss of small tissue structures (such as airways and airway epithelium in the lungs), and loss of tissue density. All of these factors can potentially be mediators of the marker signal, either due to the presence of biological targets within the tissue or due to background movement of the marker (e.g., marker diffusion into the tissue).

[0095] As described above, in Figure 2Among them, step S210 and step S220 correspond to correcting the marker signal generated by blood flow in the region. In Figure 2 step S210 and step S220 in are shown to be carried out together with steps S110 and S120. However, as discussed for steps S110 and S120 above, alternatively, steps S210 and S220 can be carried out before or after steps S110 and S120, or steps S210 and S220 can be completely omitted. The step S230 of applying the correction to the marker signal can use one or both of blood flow and tissue density corrections according to the specific implementation.

[0096] In Figure 2 the implementation of, the imaging data 10 further includes data representing the tissue structure in the region, and the correction for tissue density includes using the data representing the tissue structure to identify the sub-regions in the S210 region where the tissue density is reduced. The data representing the tissue structure can be any suitable data as described above.

[0097] For example, when using X-rays to obtain the data representing the tissue structure, if the attenuation of the X-rays in the sub-region is lower than a predetermined threshold, then the sub-region can be identified as a sub-region where the tissue density is reduced. The predetermined threshold can be selected based on the typical attenuation of healthy tissue in the region of the body, for example, determined by the measured values of other healthy subjects.

[0098] In Figure 2 the implementation of, the correction for tissue density includes correcting the measure of the marker signal based on the proportion of the tissue in the region where the density is reduced. This is achieved by determining the proportion of the tissue in the S220 region where the density is reduced and correcting the measure of the marker signal based on this proportion. In the case of emphysema, a common measure of the proportion of the tissue in the region where the density is reduced is the percentage of the low attenuation region of -950 Hounsfield units (%LAA -950 ).

[0099] In Figure 2 the implementation of, applying the correction S230 to the marker signal includes adding a correction ratio of the proportion of the tissue in the region where the density is reduced to the measure of the marker signal. The specific implementation is provided by the following equation:

[0100] N T ″ = N T + mP ld (4)

[0101] where N T ″ is the normalized measure of the marker signal after correction for tissue density, N T is the normalized measure of the marker signal, and P ldis the proportion of tissue in the region of decreased density, and m is a correction factor. If all tissues are normal (i.e., P ld = 0), this allows the median normalized count to be corrected to the expected level.

[0102] The correction of the measure of the marker signal based on the proportion of tissue in the region of decreased density can be determined using measurements of the marker signal in multiple individuals with different proportions of tissue in the region of decreased density. The correction factor can be determined by fitting (e.g., using linear regression) the (uncorrected, but optionally normalized) measure of the marker signal as a function of the proportion of tissue in the region of decreased density. For example, using linear regression on the normalized measure of the marker signal will result in a linear fit of the form:

[0103] N T = b + aP ld (5)

[0104] The coefficient a can be used for the correction factor m in equation (4) above.

[0105] Depending on which measures from multiple individuals are used to obtain the correction factor, different correction factors may be applied. Specifically, the time at which measurements are made after administering the marker to an individual may affect the correction factor due to variations in the level of unbound marker in the subject's blood.

[0106] Measurements made at a later time point (e.g., 24 hours after administration) are generally preferred for determining the correction factor because this represents a time point at which the marker is more likely to be bound to the biological target or at least evenly distributed throughout the body region. However, in some cases, measurements made at an earlier time (e.g., 6 hours after administration) may be preferred because higher levels of unbound marker in the blood may allow for more accurate background correction.

[0107] As discussed above, in addition to using data representing the distribution of the marker, the method can also use imaging data 10 that includes data representing the tissue structure. In this case, different imaging techniques or modalities may be required to obtain data representing the tissue structure and data representing the marker distribution. In addition, the scans for acquiring the imaging data 10 are typically time-consuming and may require providing the subject with rest between scans for different parts of the imaging data 10.

[0108] For these reasons, the data representing the tissue structure and the data representing the marker distribution may have different resolutions and / or volume segmentations, be measured in different states of the subject's body, and / or be measured at different times. In such cases, the method may further include aligning the data representing the tissue structure and the data representing the marker distribution. This can allow the method to account for factors such as the positioning of the subject relative to the imaging device (e.g., if the subject stands or moves around between scans) and organ movement as part of a physiological process (e.g., respiratory movement, cardiac movement).

[0109] The alignment preferably includes deforming the data representing the tissue structure to match the data representing the marker distribution. This can be achieved by segmenting regions of the body into voxels in the data representing the tissue structure and then mapping these voxels to the data representing the marker distribution. Preferably, the data representing the tissue structure is deformed rather than the data representing the marker distribution, as this results in less distortion of the measure of the marker signal. Once the regions are mapped in this way, the marker signal within each region can be quantified to obtain a measure of the marker signal and further analyzed to determine a measure of the disease, as described above. The alignment may include using a non-rigid registration algorithm.

[0110] Figure 3 shows an example of the alignment of two sets of data. Figure 3a Two sets of overlapping data of the lungs are shown, one set acquired during tidal breathing of the subject and the other set acquired at full inspiration. Figure 3b It shows how the two sets of data are aligned after applying a non-rigid registration algorithm.

[0111] The alignment may include other processes, such as downsampling one of the data representing the tissue structure and the data representing the marker distribution. This can be particularly useful when the resolutions of the two types of data are different.

[0112] The imaging data 10 may also include data obtained from contemporaneous tissue samples of the subject under study. For example, when measuring lung diseases, it can be achieved by using bronchoscopy and lung biopsy from the subject, or by including patients undergoing lung resection surgery and imaging these participants before collecting tissue for laboratory studies during the surgery. These additional data, optionally including other data such as immunohistochemical detection of a biological target (suitable for the biological target under consideration) within the same tissue, can be used to calibrate correction factors for blood flow and tissue density. For example, they can be used to improve the estimation of the proportion of tissue with reduced density, or to train an algorithm for segmentation to identify blood vessels in body regions.

[0113] Another variant of the method allows for the simultaneous use of multiple markers. In such an embodiment, the imaging data includes data representing the regional distribution of each of the multiple markers administered to the subject prior to imaging, wherein each of the multiple markers binds to a different biological target. The processing S20 of the imaging data includes processing the imaging data to obtain a measure of the marker signal for each of the multiple markers. For example, each marker may include a different radioisotope such that the activity from each marker can be separated in the imaging data 10. For example, SPECT-CT is able to independently determine the activity of different radioisotopes by detecting the specific energy of the photons emitted by each marker (the "windowing" process of the specific energy), and thus is able to detect the specific binding activity of each marker, thereby detecting the most active biological targets and / or pathways of the subject.

[0114] The correcting S30 the measure of the marker signal includes correcting the measure of each marker signal; determining S40 the measure of the disease uses the measures of the multiple corrected marker signals.

[0115] The determining S40 the measure of the disease using the measures of the multiple corrected marker signals may include combining the multiple corrected measures into a single measure of the disease. For example, the increase in activity of a biological target due to disease can vary between different diseases and / or subjects. Thus, determining a combined measure of the disease based on the multiple corrected measures can provide an indication of the overall activity of the relevant biological pathway, or a better indication of a specific subtype of the disease.

[0116] Alternatively or additionally, the determining S40 the measure of the disease using the measures of the multiple corrected marker signals includes determining a separate measure of the disease for each biological target. Multiple markers are used simultaneously, with each marker binding to a different biological target (examples include various cytokines or cell surface receptors), such that the method is able to simultaneously detect and quantify the activity of each biological target. This potentially enables the simultaneous assessment of multiple diseases or disease subtypes.

[0117] This also enables the simultaneous assessment of the effectiveness of different therapies in binding to biological targets, for example, when the marker is attached to a therapeutic agent (such as a monoclonal antibody). The treatment of a disease may be very effective, but not every patient will respond to the same drug in the same way. This may be due to different levels of different biological targets, especially the specific targets targeted by the treatment. Using multiple different markers that bind to different targets to determine the measure of the disease can be used to determine which targets the treatment for an individual subject should target.

[0118] Another potential application of simultaneous injection of multiple markers is to study non-specific signals from the markers that may have diffused into tissues within a body region but have not bound to the target component. One of the multiple markers may be a specific marker targeting a biological target of interest, while another of the multiple markers may be a "negative control" marker that is not expected to show specific binding to any target within the body region. The negative control marker will provide data on how the markers that have not bound to the biological target behave in the tissue and thus provide how much of the signal detected from the specific marker may be bound to the biological target.

[0119] The method can be carried out by a device for determining a measure of a disease in a region of a subject's body, the device including a processor configured to perform the steps of the method. The device includes a receiving unit configured to receive imaging data obtained by imaging a region of the subject's body, the imaging data including data representing the distribution of markers within the region administered to the subject prior to imaging, wherein the markers bind to components of an inflammatory pathway. The device includes a processing unit configured to process the imaging data to obtain a measure of the marker signal in the region. The device includes a correction unit configured to correct the measure of the marker signal for the effect of the marker signal on the tissue structure within the region. The device includes a determination unit configured to use the corrected measure of the marker signal to determine a measure of inflammation of the region. The device can be a general-purpose computer adapted to perform the method. The configuration of the device can be appropriately modified to correspond to any details of the above method.

[0120] Examples

[0121] The following examples verify the applicability of the method in the specific case of using a marker that binds to TNF-α to determine a measure of inflammation in subjects with COPD.

[0122] Participant Selection for SPECT-CT Study

[0123] In this study conducted at University Hospital Southampton, UK, 5 participants with severe to very severe COPD (according to the GOLD criteria) and 5 participants without any history of underlying lung disease were recruited. According to the GOLD criteria, participants with COPD had a previous clinical diagnosis of COPD, a smoking history of > 10 pack-years, severe or very severe COPD, a predicted post-bronchodilator FEV1 / FVC < 0.7 and FEV1 < 50%. Healthy volunteers had no history of lung disease and were defined as non-smokers with a smoking history < 1 pack-year. The characteristics of the participants are outlined in Tables 1 to 3.

[0124] Participants will be excluded if they have a history of other respiratory diseases, risk factors for pneumonia, chronic kidney disease (CKD) with an estimated glomerular filtration rate <60, autoimmune diseases, are taking immunosuppressive drugs or other monoclonal antibodies, have a history of adverse reactions to any monoclonal antibody drug, have a body mass index (BMI) outside the range of 18 - 30, have an indwelling catheter or are taking any conventional antibacterial drugs, or have taken antiviral or respiratory investigational drugs within 30 days prior to the enrollment visit. In addition, subjects with COPD should not have had an exacerbation requiring oral corticosteroid treatment within 30 days prior to enrollment.

[0125] After recruitment, all participants underwent pre - bronchodilator and post - bronchodilator spirometry to measure the diffusing capacity of the lungs for carbon monoxide (DLCO) and lung volumes by plethysmography. Participants were asked to discontinue their regular bronchodilator medications 12 hours prior (if applicable). During this visit, physical measurements and bioelectrical impedance analysis were also performed to measure body composition. Prior to imaging, participants had blood sampling for safety assessment and measurement of TNF - α concentration. Before and after imaging, sputum induction was also performed to collect sputum TNF - α samples whenever possible and within the scope of COVID - 19 infection control policies.

[0126]

[0127] Table 1

[0128]

[0129]

[0130] Table 2

[0131]

[0132] Table 3

[0133] Use 99m Tc radioactively labeled anti-TNF-α

[0134] Radioactive labeling was performed by radiopharmacology at University Hospital Southampton. Infliximab, an anti - TNF - α monoclonal antibody (Remicade, Janssen Biotech), was labeled with 99m Tc using the direct labeling method described previously 10,11 . The anti - TNF - α was first purified by gel filtration chromatography and then separated based on the optical density at 280 nm using an ultraviolet spectrophotometer. The antibody was then reduced using a molar excess of 2 - mercaptoethanol, leaving free sulfhydryl groups available for labeling. After incubation with 2 - mercaptoethanol for 30 minutes, the reduced antibody was separated using a PD - 10 desalting column.

[0135] Reconstitute methylene diphosphate (MDP) bone scan kits (Draximage and Polatom) in 5 ml of 0.9% sodium chloride to provide a solution containing 5 mg of methylene diphosphate (medronate), 0.34 mg of stannous fluoride, and 2 mg of para-aminobenzoic acid. Technetium-99m is eluted from 99 Mo / 99m Tc generators and obtained in the form of sodium pertechnetate. 500 mcg of reduced antibody is added to the solution, and then 450 MBq of pertechnetate is added to generate 99mTc-anti-TNF-α.

[0136] Radiochemical purity is determined by instant thin layer chromatography (iTLC) strips using the cut counting method. ITLC is the stationary phase and 0.9% NaCl is the mobile phase. These conditions give an Rf value of Rf = 0 for the labeled antibody and Rf = 1 for the impurities. Radiochemical purity remained >95% for all but one participant, who had a radiochemical purity of 91% using this method (Table 4).

[0137]

[0138]

[0139] Table 4

[0140] Dosimetry Study

[0141] Dosimetry was performed on the first two participants in the study (one per group). The estimated whole body effective dose was 1.6 mSv to 1.7 mSv (1.6 mSv to 1.9 mSv when converted to 370 MBq). This is still well below the upper limit of 3.6 mSv allowed by the study protocol. Whole body retention was calculated to match most closely the expected behavior of a full antibody of size 150 kDa (such as infliximab 1 / 2 ) with a biological t of 23.7 hours (4.8 hours effective t 1 / 2 ) for the healthy group and 19.5 hours (4.6 hours effective t 1 / 2 ) for the COPD group. 12 )

[0142] SPECT-CT Imaging and Biodistribution Study

[0143] Imaging was performed 6 hours (+ / - 60 minutes) and 24 hours (+ / - 4 hours) after the start of injection. Participants in the healthy volunteer and COPD groups underwent whole-body planar imaging, chest SPECT imaging, and chest low-dose CT imaging at each time point. Participants in the COPD group had additional chest high-resolution CT (HRCT) at the 6-hour time point. For the first participant in each group, whole-body planar imaging was also performed 3 hours (+ / - 30 minutes) after injection for additional dosimetry calculations.

[0144] All imaging was obtained using a Symbia Intevo Bold (Siemens Healthcare GmbH, Germany) 16-slice dual-headed gamma camera system. Whole-body planar imaging was acquired with a 256x1024 matrix, at a collection speed of 15 cm / min at 3 hours, 10 cm / min at 6 hours, and 5 cm / min at 24 hours. SPECT imaging at 6 hours was acquired with a 256x256 matrix, 60 views, 20 seconds per view. SPECT at 24 hours was acquired with a 256x256 matrix, 60 views, 40 seconds per view.

[0145] SPECT imaging was performed with the patient supine, hands above the head. Immediately after SPECT imaging, the participants remained in that position for CT imaging. All participants underwent free-breathing low-dose CT (LDCT) at both time points. Inspiratory and expiratory HRCT imaging was obtained only in COPD patients at the 6-hour time point.

[0146] Imaging Processing and Analysis

[0147] After imaging, the images were first stored using the hospital Picture Archiving and Communication System (PACS), and then all personal identifying information was removed and transferred to a workstation for further analysis.

[0148] CT data, including MSCT (multi-slice CT) and low-dose, were first analyzed using Southampton Pulmonary Radiomics (SPR) software (version β-e3b7ef8254). This analysis consisted of:

[0149] 1) Manually identifying seed points near the top of the trachea;

[0150] 2) Automatic airway segmentation, followed by semi-automatic editing to ensure adequate segmentation of all necessary branches;

[0151] 3) Automatic segmentation of the left and right lungs;

[0152] 4) Manually labeling airway branches, especially the left and right main bronchi;

[0153] 5) Semi - automatically segment lung lobes and fissures;

[0154] 6) Automatically analyze lung lobes based on a - 950HU threshold to identify regions of emphysema;

[0155] 7) Segment blood vessels using a fully automated, AI - driven method.

[0156] The administration of the corresponding low - dose CT scan enables the operator to manually identify the location of the aortic arch and place a spherical region of interest (RoI). In each case, the size and position of the RoI are manually adjusted to ensure that as much of the aortic arch volume as possible is captured.

[0157] Due to the different voxel sizes of the SPECT and low - dose CT scans, the regions identified from the low - dose CT scan are reduced in pixel sampling to match the size of the corresponding SPECT scan. These reduced pixel sampling regions can be used to directly mask out the regions of interest from the SPECT data.

[0158] However, the use of these regions (corresponding to lung lobes and the whole left and right lungs) is complex because the fact that these scans are obtained at full inspiration pause while the SPECT scans are obtained during tidal breathing. This results in significant differences in the size and shape of the lungs between the two scans, meaning that it is not possible to directly use the MSCT regions. This problem is solved by adopting a non - rigid registration algorithm to align the regions identified by the MSCT scan with the corresponding features identified by the low - dose CT scan. Since the low - dose CT scan and the SPECT scan are automatically spatially aligned by the scanner, it can be assumed that the displayed MSCT regions will be aligned with the corresponding regions in the SPECT scan.

[0159] Non - rigid registration is performed separately for each lobe to ensure maximum alignment accuracy and save computational resources. The binary masks generated by the segmentation of each lobe are used to separate the MSCT voxels that only correspond to that lobe. Then the volume corresponding to the separated lobe is extracted from the entire MSCT scan, resulting in a smaller and more manageable volume on which all further processing is performed. A zero - padding of 10 voxels in each dimension is added to the separated volume. Then an initial Affine rigid registration is calculated using a 1 + 1 evolutionary optimizer and the Mattes mutual information registration metric. After the initial rigid registration, non - rigid registration is performed using the demons algorithm, which has cumulative field smoothing, 5 pyramid levels and 100 iterations per level. It has been found that this process produces excellent alignment between the features identified in the MSCT scan and the corresponding regions in the low - dose CT and SPECT scans.

[0160] All further processing (including correction of vascular and tissue density effect marker signal metrics) was performed using Matlab (version 2021a, the Mathworks Inc, Natick, MA) according to the above method. Figure 4 A detailed flow chart of the processing used is shown.

[0161] Statistical Analysis

[0162] Since the voxel counts were non-normally distributed, median-normalized counts were used as the summary statistic for each patient. Student's t-test was used for between-group comparisons. Pearson correlation coefficients were used for the correlation between median-normalized counts and other factors.

[0163] Results

[0164] The raw counts were first determined from the SPECT images within the region of interest defined by the LDCT images taken at the time of SPECT scanning. The raw counts in this example corresponded to the metrics of the marker signal discussed above. The counts were averaged over the entire RoI. The count distribution was non-normally distributed over the volume examined, so the median count was used as the summary statistic for each participant. Figure 5 shows a box plot representing these results.

[0165] Figure 5a It was shown that the mean + / - SD for the healthy group at the 6-hour time point was 5725.0 + / - 1121.3, and for the COPD group was 3182.0 + / - 631.7. Figure 5b It was shown that at the 24-hour time point, the mean for the healthy group was 2900.0 + / - 471.5, and for the COPD group was 1864.0 + / - 282.6. The raw counts for the healthy group were consistently higher at both time points.

[0166] Analysis of the MSCT scans produced a much higher level of detail in the regions compared to the corresponding low-dose CT scans. The ROI area / volume was normalized. 15 .

[0167] A reduced-pixel-sampled aortic arch RoI was used to isolate the SPECT voxels corresponding to the blood within the aortic arch. The sum and mean of these voxel values were calculated and used as a measure of the "background" activity level.

[0168] The sum, mean, and median are calculated in a similar manner for each reduced pixel sampling region identified from the low-dose CT scans (corresponding to the lung lobes and the entire left and right lungs). In addition, normalized counts are calculated by forming a target-background (T / B) ratio to normalize for biological clearance. This is done by dividing each voxel value by the mean activity identified in the aortic arch. The raw voxel count values are then converted to MBq units by multiplying the count value at each voxel by the spatial volume of that voxel and dividing by one million.

[0169] At the 6-hour and 24-hour time points, normalized counts are calculated on a regional and whole-lung basis. The distribution of the normalized count for each voxel follows a non-normal distribution, and thus the median count is calculated as the summary statistic for each participant.

[0170] Figure 6a Box plots showing the distribution of median normalized counts at the 6-hour time point for the healthy and COPD groups are presented. The mean (+ / - standard deviation) of the summary statistics for the healthy group is 0.182 + / - 0.027, while for the COPD group it is 0.087 + / - 0.019.

[0171] Figure 6b Box plots showing the distribution of median normalized counts at the 24-hour time point (B) for the healthy and COPD groups are presented. The mean of the summary statistics for the healthy group is 0.250 + / - 0.089, and the mean of the summary statistics for the COPD group is 0.147 + / - 0.056.

[0172] Figure 6c The difference in normalized counts between healthy subjects (left) and subjects with COPD (right) is confirmed.

[0173] There is a significant difference between the groups, with the healthy group having a higher median normalized count at the 6-hour time point (p < 0.001), but not at the 24-hour time point. The differences observed are due to factors discussed above, such as blood flow and tissue density.

[0174] Figure 6d It is confirmed that the absolute difference in median normalized counts detected between the 6-hour and 24-hour time points is not significantly different between the groups. Figure 6e It is shown that the increase in median normalized counts is greater in the COPD group relative to the 6-hour scan, but the difference does not reach statistical significance.

[0175] The difference in median normalized counts between scans at 6 hours and 24 - hour time points was also calculated for each participant. At the 24 - hour time point, the median normalized counts increased in both groups. The absolute difference in median normalized counts + / −SD for the healthy group was 0.067+ / -0.073, and for the COPD group was 0.059+ / -0.040. The difference between groups was not statistically significant. When the difference in median normalized counts was calculated as a proportion of the 6 - hour scan for each patient, a greater increase in median normalized counts was observed in the COPD group. The median normalized counts in the healthy group increased on average by 35.38%+ / -34.33, and in the COPD group by 64.88%+ / -31.04, but the difference between groups did not reach statistical significance.

[0176] The higher difference in median normalized counts in the COPD group, when calculated as a proportion of the 6 - hour scan, provides a potential measure of more specific binding to TNF - α at the 24 - hour time point in this group.

[0177] Previous literature has shown that as airflow obstruction and emphysema progress, the disappearance of small vessels detectable on CT 13 . Therefore, variability in vascular density was expected in the population selected for this study.

[0178] Vascular density was determined using the above - mentioned automated segmentation technique in scans performed at 6 hours and 24 - hour time points. Some variation was observed between the vascular densities determined from scans at 6 - hour and 24 - hour time points, although this was generally small. At the 6 - hour time point, the mean + / −SD for both lungs was 0.109+ / -0.028 (range 0.069 - 0.158), and at the 24 - hour time point was 0.112+ / -0.033 (range 0.070 - 0.171).

[0179] Figure 7 Indicating that when examining the individual components calculated, the difference in vascular volume was small compared to the total lung volume measured by CT, and the difference in vascular density between scans was attributable to changes in the overall lung volume.

[0180] The mean vascular density detected by CT was plotted against the median normalized counts detected by SPECT to investigate the effect of blood flow detectable through lung tissue on SPECT activity. Figure 8a A plot for the 6 - hour time point is shown, Figure 8bA graph showing the 24-hour time point is presented. A significant positive correlation is observed at both time points. The Pearson correlation coefficient at the 6-hour time point is 0.824 (p = 0.003), and at the 24-hour time point is 0.862 (p = 0.001). This supports the conclusion that vascular density affects the measurement of the marker signal and validates the above method.

[0181] The population selected for the study also differed in the proportion of areas showing reduced tissue density. In a specific embodiment, the reduced tissue density in COPD patients is due to emphysema. Figure 9a A graph showing the change in emphysema score among subjects with COPD is presented, expressed as %LAA -950 (a CT measurement of tissue density indicating the presence of emphysema), which is determined by high-dose inspiratory CT scan.

[0182] The %LAA -950 values are plotted against the median normalized counts detected by SPECT. Figure 9b The results at the 6-hour time point are presented, Figure 9c The results at the 24-hour time point are presented. A significant negative correlation is identified in the HRCT images at both time points. HRCT data are only used for COPD participants. At the 6-hour time point, the Pearson correlation coefficient is -0.884 (p = 0.047), and at the 24-hour time point, the Pearson correlation coefficient is -0.954 (p = 0.012).

[0183] The above method is applied to control the effect of vascular density. Figure 10a A box plot of the median normalized counts at the 6-hour time point after correcting for blood flow using the above method is presented. Figure 10b The corresponding corrected counts at the 24-hour time point are presented. The median normalized counts of the participants decreased on average by 0.110 + / - 0.030, but the uptake in the healthy group remained generally higher compared to the COPD group. With the continuous improvement of the segmentation algorithm, a greater effect can be detected as a higher proportion of the overall vascular tree can be detected from LDCT images.

[0184] To correct for the effect of reduced tissue density due to emphysema, first, an inspiratory MSCT analysis is performed on each participant in the COPD group using the density mask of LAA -950 to determine the degree of emphysema.

[0185] The density mask is used on the lung regions identified in the HRCT images to determine the %LAA -950 values. HRCT data are only applicable to COPD patients (HRCT is not part of the scanning protocol for healthy volunteers). Then the %LAA-950 Results were plotted relative to median-normalized counts and correlations were plotted by linear regression. As described above, if %LAA -950 is zero, the median-normalized counts were adjusted to the desired value using the resulting regression coefficients. The effect of this correction method depends on whether the regression line is determined from the results at the 6-hour or 24-hour time points.

[0186] Figure 11a and Figure 11b box plots indicate that if the 6-hour regression line equation is used, the corrected counts in the healthy group are still higher at the 6-hour ( Figure 11a ) and 24-hour ( Figure 11b ) time points. Using the 6-hour regression line, at the 6-hour time point, the corrected counts (mean + / - SD) in the healthy group were 0.182 + / - 0.027 and in the COPD group were 0.129 + / - 0.009. At the 24-hour time point, in the healthy group they were 0.250 + / - 0.089 and in the COPD group were 0.189 + / - 0.041.

[0187] However, Figure 11c and Figure 11d box plots indicate that if the 24-hour regression line is used, the corrected counts in the COPD group are higher at both the 6-hour ( Figure 11c ) and 24-hour ( Figure 11d ) time points. Using the 24-hour regression line, at the 6-hour time point, the corrected counts in the healthy group were 0.182 + / - 0.027 and in the COPD group were 0.225 + / - 0.038. At the 24-hour time point, in the healthy group they were 0.250 + / - 0.089 and in the COPD group were 0.285 + / - 0.017.

[0188] Discussion

[0189] The imaging platform provides evidence of proper quantification of the signal from the radiopharmaceutical and this allows for quantitative analysis on a regional and bilateral lung basis. Figure 12 is a three-dimensional representation of the quantification of the radiopharmaceutical signal in the right lung at the 6-hour and 24-hour time points for two study participants. Figure 12a shows Subject 1 (healthy), Figure 12b shows Subject 10 (COPD). Counts were determined on a regional basis but could also be analyzed using the median count of the entire bilateral lung RoI. Due to the presence of emphysema in the COPD group, changes in tissue density between the groups are visible.

[0190] The study results indicate that the uncorrected signal is higher in the healthy group but when using the empirical method to correct for emphysema (by %LAA on HRCT -950When using the regression line of the counts detected at the 24-hour time point in the calibration equation in the presence of (as defined), the median normalized count in the COPD group was higher.

[0191] The biodistribution and clearance in this study were consistent with previous study data, with higher vascular compartment activity at earlier time points, metabolism and clearance through the liver and kidney systems, and uptake by the liver and spleen. 8,11 . At later time points, as the T / B ratio increased, more specific binding to the target occurred, which was consistent with what was previously observed in studies of rheumatoid arthritis 9 and sarcoidosis 8 .

[0192] In addition, in the COPD group, a higher increase in the T / B ratio relative to the 6-hour scan seemed to indicate that in these patients, the binding of the radiopharmaceutical to its target TNF-α was more specific. This was consistent with previous literature on patients with sarcoidosis, in whom the percentage change in the T / B ratio was greater in patients with indications for systemic treatment 8 . In studies of rheumatoid arthritis 9 , there was also a higher increase in the absolute T / B value in anti-TNF-α treatment responders. Compared with synovial tissue, lung tissue was more heterogeneous, which might be the reason for the lack of an increase in absolute counts observed in this study. Using relative changes might be more relevant.

[0193] Given the central role of TNF-α as a pro-inflammatory cytokine in COPD, in this study, 99m the quantifiable increase in the uptake of TC-infliximab relative to the 6-hour scan, and the higher T / B ratio after applying the correction factor, would correspond to the expected increase in TNF-α expression in the lungs of COPD patients.

[0194] The evidence provided by this study indicates that using technetium-labeled TNF-α as an inflammatory marker can image active inflammation in the lungs of COPD patients. Previous studies in other disease groups have shown that uptake can predict response to treatment, indicating that these techniques can be used to evaluate the efficacy of different therapies. Due to the lack of COPD disease-modifying therapies, there is a particular need for tools to identify the response of COPD to therapies in the clinical trial setting and the clinical setting.

[0195] This study provides evidence for the effectiveness of the above-mentioned novel quantification method for non-invasive detection of target cytokine activity, where target cytokine activity serves as a measure of active inflammation in the lungs of patients with airway diseases. Previous studies have investigated SPECT-CT and scintigraphy methods in other inflammatory diseases, such as rheumatoid arthritis, Crohn's disease, and sarcoidosis 8,9,11A method for measuring the activity of a target cytokine. This supports the conclusion that the method is also applicable to other inflammatory diseases and to detecting other components of the inflammatory pathway.

[0196] Applications

[0197] In the above embodiments, the method was applied to and tested in the lungs, but can be readily applied to other diseases and / or organs, such as inflammatory diseases of the joints, such as RA. For example, the method can be applied to quantify components such as IL-5, IL-5Rα, and IgE in asthma, or to quantify IL-6 activity in rheumatoid arthritis. Other potential applications include inflammatory diseases of the gut or liver, or other classes of diseases, such as cancer or infection.

[0198] The method can be used for a variety of purposes. Since the measure of the disease can be determined based on specific biological targets, the method can be used to elucidate the underlying active mechanisms involved in the disease. Disease imaging using the method can also be used as a biomarker to distinguish specific endotypes or phenotypes within a disease group. For example, the method can be used to assist in the analysis and identification of cancer subtypes or occult infections. The method can provide initial stratification of study subjects and generation of final outcomes in clinical trials, for example, aimed at selecting a specific subgroup of patients with a particular disease who are most likely to respond to the investigational medicinal product (IMP).

[0199] As described above, the method can also predict the probability of an individual subject's response to a specific therapy based on the level of the biological target targeted by the therapy. This provides a precision medicine approach to determining the best therapy for a patient and can quickly identify the best drug for a specific patient. This enables a doctor to find the correct therapy at the first instance, reducing the time and money spent on trying different treatment methods to find a therapy effective for a specific subject.

[0200] The method will also facilitate the research of new drugs and their application in other disease areas by providing a means to rapidly quantify the effect of a drug on a disease using a non-invasive method. The method can also be applied to assist in the stratification of study participants, thereby increasing efficiency and obtaining results more quickly. This is highly valuable for disease groups such as COPD, where the heterogeneity of the population may be the reason why no disease-modifying therapies have been proven successful in clinical trials to date.

[0201] This method can be implemented as part of the deployment of a platform through a Software as a Service (SaaS) model, which will enhance current imaging capabilities. This method can ultimately be developed into a clinical imaging platform integrated into existing imaging processes, which is advantageous because the analysis required for this process can be automated. The software platform can be connected to hospital computer systems to access imaging data 10 and feed the results back to clinicians to assist in guiding treatment. From the patient's perspective, they only need to attend a routine imaging appointment at their local hospital. Some parts of the platform may be provided locally to the hospital, while other parts can be provided remotely. For example, as Figure 13 shown in the example flowchart of, basic data validation of the imaging data 10 can be performed locally, such as during imaging. The analysis for determining the metrics of the disease can be performed by a remote system. The partitioning of the processing can be selected based on the availability of computing resources and other considerations, such as restrictions for data protection considerations.

[0202] Figure 14 The potential implementation of this method as a clinical platform was confirmed. Figure 14 a shows the entire platform process and targets specific biological targets, namely components of the COPD inflammatory pathway. Figure 14 b shows the expected clinical applications of using a SPECT-CT imaging platform to achieve a precision medicine approach for selecting biotherapies. By injecting several monoclonal antibodies, multiple biological pathways can be studied simultaneously, where each monoclonal antibody is labeled with a different radioactive isotope. For example, if it is determined that three monoclonal antibodies provide potential benefits to a patient, each monoclonal antibody can be labeled with a different radioactive isotope and injected simultaneously. The uptake of each antibody is quantified, and the most active pathway is identified, thus selecting the most effective therapy for the patient.

[0203] References

[0204] 1. Bek, S. et al. Systematic review and meta-Analysis: Pharmacogenetics of anti-TNF treatment response in rheumatoid arthritis. Pharmacogenomics J. 17, 403 - 411 (2017).

[0205] 2. Jiemy, W.F. et al. Positron emission tomography (PET) and single photon emission computed tomography (SPECT) imaging of macrophages in large vessel vasculitis: Current status and future prospects. Autoimmun. Rev. 17, 715 - 726 (2018).

[0206] 3. Jamar, F., Versari, A., Galli, F., Lecouvet, F. & Signore, A. Molecular Imaging of Inflammatory Arthritis and Related Disorders. Semin. Nucl. Med. 48, 277 - 290 (2018).

[0207] 4. Jones, H.A., Marino, P.S., Shakur, B.H. & Morrell, N.W. In vivo assessment of lung inflammatory cell activity in patients with COPD and asthma. Eur. Respir. J. 21, 567 - 573 (2003).

[0208] 5. Mariani, G. et al. A review on the clinical uses of SPECT / CT. Eur. J. Nucl. Med. Mol. Imaging 37, 1959 - 1985 (2010).

[0209] 6. Israel, O. et al. Two decades of SPECT / CT - the coming of age of a technology: An updated review of literature evidence. Eur. J. Nucl. Med. Mol. Imaging 46, 1990 * 2012 (2019).

[0210] 7. Pacilio, M., Lauri, C., Prosperi, D., Petitti, A. & Signore, A. New SPECT and PET Radiopharmaceuticals for Imaging Inflammatory Diseases: A Meta - analysis of the Last 10 Years. Semin. Nucl. Med. 48, 261 - 276 (2018).

[0211] 8. Vis, R. et al. 99mTc - anti - TNF - α antibody for the imaging of disease activity in pulmonary sarcoidosis. Eur. Respir. J. 47, 1198 - 1207 (2016).

[0212] 9. Conti, F. et al. Role of scintigraphy with 99mTc - infliximab in predicting the response of intraarticular infliximab treatment in patients with refractory monoarthritis. Eur. J. Nucl. Med. Mol. Imaging 39, 1339 - 1347 (2012).

[0213] 10. Mather, S. J. & Ellison, D. Reduction - mediated technetium - 99m labeling of monoclonal antibodies. J. Nucl. Med. 31, 692 - 697 (1990).

[0214] 11. D’Alessandria, C. et al. Use of a 99m - Tc labeled anti - TNF alpha monoclonal antibody in Crohn’s disease: in vitro and in vivo studies. Q. J. Nucl. Med. Mol. Imaging 51, 334 - 342 (2007).

[0215] 12. Valentin, J. Radiation dose to patients from radiopharmaceuticals. Addendum 3 to ICRP Publication 53. ICRP Publication 106. Approved by the Commission in October 2007. Ann. ICRP 38, 1 - 197 (2008).

[0216] 13. Matsuoka, S. et al. Quantitative CT Measurement of Cross - sectional Area of Small Pulmonary Vessel in COPD. Acad. Radiol. 17, 93 - 99 (2010).

[0217] 14. Keatings, V.M., Collins, P.D., Scott, D.M. & Barnes, P.J. Differences in interleukin - 8 and tumor necrosis factor - alpha in induced sputum from patients with chronic obstructive pulmonary disease or asthma. Am. J. Respir. Crit. Care Med. 153, 530 - 534 (1996).

[0218] 15. Galli, F., Lanzolla, T., Pietrangeli, V., Malviya, G., Ricci, A., Bruno, P., Ragni, P., Scopinaro, F., Mariotta, S., & Signore, A. (2015). In vivo evaluation of TNF - alpha in the lungs of patients affected by sarcoidosis. BioMed Research International, 2015. https: / / doi.org / 10.1155 / 2015 / 401341

[0219] 16.Dammes,N.,&Peer,D.(2020).Monoclonal antibody-based molecularimaging strategies and theranostic opportunities.In Theranostics(Vol.10,Issue2,pp.938-955).Ivyspring International Publisher.https: / / doi.org / 10.7150 / thno.37443。

Claims

1. A method for determining a measure of a disease in a region of a subject's body, comprising: receiving imaging data obtained from imaging a region of the subject's body, the imaging data including data representing the distribution of a marker within the region of the subject to which the marker was administered prior to the imaging, wherein the marker binds to a disease-related biological target; processing the imaging data to obtain a measure of the marker signal in the region; correcting the measure of the marker signal for the effect of the marker signal of the tissue structure within the region; and using the corrected measure of the marker signal to determine a measure of the disease in the region.

2. The method according to claim 1, wherein The corrected measure of the marker signal represents the binding level of the marker to the tissue within the region.

3. The method according to claim 1 or 2, wherein The disease is an inflammatory disease and the biological target is a component of an inflammatory pathway.

4. The method according to any one of the preceding claims, wherein, Correcting the measure of the marker signal includes correcting for the marker signal caused by blood flow in the region.

5. The method according to claim 4, wherein, Correcting for the marker signal caused by blood flow includes determining the vascular density in the region.

6. The method according to claim 5, wherein Determining the vascular density includes determining the total volume of blood vessels in the region.

7. The method according to claim 5 or 6, wherein The imaging data further includes data representing the tissue structure in the region, and determining the vascular density includes identifying blood vessels in the region using the data representing the tissue structure.

8. The method according to claim 7, wherein, The data representing the tissue structure is obtained by one or more of low-dose computed tomography, high-resolution computed tomography, multi-slice computed tomography, magnetic resonance imaging, plain film radiography, and ultrasound.

9. The method according to any one of claims 4 to 8, wherein Correcting for the marker signal caused by blood flow in the region includes determining a background marker signal corresponding to the marker signal from unbound markers in the blood.

10. The method according to any one of the preceding claims, wherein, Correcting the measure of the marker signal includes correcting for the tissue density in the region.

11. The method according to claim 10, wherein, The imaging data further includes data representing the tissue structure in the region, and correcting for the tissue density includes using the data representing the tissue structure to identify sub-regions in the region where the tissue density is reduced.

12. The method according to claim 11, wherein, The data representing the tissue structure is obtained by one or more of low-dose computed tomography, high-resolution computed tomography, multi-slice computed tomography, magnetic resonance imaging, plain film radiography, and ultrasound.

13. The method according to claim 12, wherein, The data representing the tissue structure is obtained using X-rays, and when the X-ray attenuation in the sub-region is below a predetermined threshold, the sub-region is identified as a sub-region where the tissue density is reduced.

14. The method according to any one of claims 10 to 13, wherein Correcting for the tissue density includes correcting the measure of the marker signal based on the proportion of tissue in the region with reduced density.

15. The method according to claim 14, wherein, Correcting the measure of the marker signal based on the proportion of tissue in the region with reduced density is determined by using measured values of the marker signal in a plurality of individuals with different tissue proportions in the region with reduced density.

16. The method according to any one of the preceding claims, wherein, The measure of the marker signal is the average marker signal in the region, optionally the median marker signal.

17. The method according to any one of the preceding claims, wherein the measure of the marker signal is normalized by a background marker signal corresponding to the marker signal of the unbound marker in the blood.

18. The method according to any one of claims 9 or 17 or their dependent claims, wherein The background marker signal is identified from the marker signal of the major blood vessels, optionally the aortic arch.

19. The method according to claim 18, wherein, The background marker signal is the average marker signal in the major blood vessels, optionally the median marker signal.

20. The method according to any one of the preceding claims, wherein, The biological target includes immune and / or inflammatory signaling proteins, such as cytokines, chemokines, cell surface receptors, or extracellular matrix components.

21. The method according to any one of the preceding claims, wherein The marker includes a radiolabel, and / or the marker includes a monoclonal antibody.

22. The method according to any one of the preceding claims, wherein, Data representing the distribution of the marker within the region is obtained using single photon emission computed tomography, SPECT, positron emission tomography, PET, planar scintigraphy, or magnetic resonance imaging.

23. The method according to any one of the preceding claims, wherein, The region of the subject's body includes the subject's organs, such as one or both lungs of the subject, or the joints of the subject.

24. The method according to any one of the preceding claims, wherein: The imaging data includes data representing the distribution within the region of each of a plurality of markers administered to the subject prior to the imaging, wherein each of the plurality of markers binds to a different biological target; The processing of the imaging data includes processing the imaging data to obtain a measure of the marker signal of each of the plurality of markers; The correction of the measure of the marker signal includes correcting the measure of each of the marker signals; and The determination of the measure of the disease uses the measures of the plurality of corrected marker signals.

25. The method according to any one of claims 7 or 11 or its dependent claims, wherein: The data representing the tissue structure and the data representing the marker distribution have different resolutions and / or volume segmentations, are measured with the subject's body in different states, and / or are measured at different times; and The method further includes aligning the data representing the tissue structure and the data representing the marker distribution.

26. The method according to claim 25, wherein, The alignment includes using a non-rigid registration algorithm.

27. The method according to claim 25 or 26, wherein The alignment includes deforming the data representing the tissue structure to match the data representing the marker distribution.

28. A computer program comprising instructions that, when executed by a computer, cause the computer to perform the method according to any one of claims 1 to 27.

29. A computer-readable medium comprising instructions that, when executed by a computer, cause the computer to perform the method according to any one of claims 1 to 27.

30. A device for determining a measure of a disease in a region of a subject's body, comprising a processor configured to perform the steps of the method according to any one of claims 1 to 27.

Images