Computer-implemented method for quantifying lung function and means thereof

Abstract

The invention relates to a computer-implemented method for determining a so-called PUV value for a region of interest (ROI) on a lung scan that is a PET / CT or a SPECT / CT lung scan image, wherein the PUV value for the region of interest is determined according to the formula: PUVROI = (CROI x ALV) / A0 with ALV being the Anatomical Lung Volume, A0 being the total radioactivity inside the region defined by the ALV and CROI being the concentration of radioactivity inside the ROI. The invention also relates to a data processing apparatus configured for carrying out the method, to a computer program product and a non-transitory computer readable medium having stored thereon the such a computer program product, a device for determining a PUV value for a region of interest on a PET / CT or SPECT / CT lung scan and a medical image apparatus comprising such a device for determining a PUV value for a region of interest on a PET / CT or SPECT / CT lung scan as well as a method for monitoring the condition of a patient over time, where the method for determining a PUV for a ROI is reiterated over time.

Description

[0001] COMPUTER-IMPLEMENTED METHOD FOR QUANTIFYING LUNG FUNCTION AND MEANS THEREOF

[0002] The present invention concerns computer-implemented methods aimed at being of assistance in lung nuclear medicine, notably for quantifying lung function, and more precisely pertains to the field of methods for evaluating, notably quantifying, perfusional and / or ventilatory lung function when using lung nuclear medicine techniques that are scintigraphy coupled with computed tomography (SPECT / CT) or lung positron emission tomography coupled with computed tomography (PET / CT).

[0003] Generally, throughout the present text, SPECT / CT stand for “scintigraphy coupled with computed tomography” and PET / CT stands for “lung positron emission tomography coupled with computed tomography”.

[0004] Lung function is a term conventional in the field, which is used to describe how well the lungs work in helping a person breathe. Lung function is also called “pulmonary function”. Several tests are known and used by physicians or in clinical settings to evaluate the same. Depending upon the test used, particular aspects of lung function can be investigated.

[0005] Present invention relates to a computer-implemented method for determining a so-called PUV value (acronym for “Pulmonary Uptake Value”, which is a parameter intrinsic to the present invention - the term value being doubled for pronunciation, and can be transposed to “PUV” alone or “PUV” indicia or similar terms, as used herein. All these terms are manner to transcribe the parameter that is calculated, as computed by the formula described herein for a region of interest) for a Region of Interest (ROI) on a lung scan, and a data processing apparatus, a computer program product, a non-transitory computer-readable medium, a device or a medical image apparatus suited to this end. In the context of present invention, which is applied to the lung, a ROI is considered in a tri-dimensional (3D) space, hence use is also made of the term “voxel” herein, to qualify the said ROI.

[0006] Lung scintigraphy (SPECT) or lung positron emission tomography (PET) are imaging examinations that assess the regional distribution of perfusion and ventilation in the lungs. Perfusion images are obtained after injection of albumin macroaggregates (MAA), which embolize in pulmonary capillaries. Ventilation images are obtained after inhalation of radioactive gases or aerosols. Lung perfusion positron emission tomography coupled with computed tomography (PET / CT) is an imaging modality for assessing regional lung function that is relatively new, especially in clinical settings (1 ). Like lung perfusion scintigraphy, lung perfusion positron emission tomography assesses regional lung perfusion distribution by embolizing albumin macroaggregates (MAA) in pulmonary capillaries (2). Lung ventilation PET can also be performed after inhalation of radioactive gases or aerosols. Lung PET / CT imaging has already shown promising results in various clinical contexts, such as the diagnosis of pulmonary embolism (6)(7), assessment of pulmonary function prior to lung surgery (7)(8) and radiotherapy (RT) planning (9).

[0007] However, a current limitation to the use of either lung SPECT / CT or lung PET / CT, especially in clinical settings is the absence of a tool enabling to reliably and reproducibly quantify tracer uptake within the lung. More precisely, considering a single image uptake when performing ventilation or perfusion SPECT or PET scans of the lung, there is currently no method for reliably and reproducibly quantifying the intensity of uptake of radioactivity in a given voxel. By “voxel”, it is meant a “volume element” representing some numerical quantity, i.e., used in the visualization and analysis of three-dimensional (especially scientific and medical) data. The word “voxel” originated, in a literary way, by analogy to a “pixel” and is construed as the “three- dimensional analogue” of a pixel. However, reference to a “voxel” in the present application follows a broad definition of the term as a three-dimensional Region of Interest representing a numerical quantity (a volume element, about which an information is quantified). Reference is also made to a “region of interest” (ROI) in the present text, which can be defined by a voxel in the context. Alternatively, it can also be said that a parameter is measured or considered “at the voxel level”, i.e., considering a region in the three-dimensional space.

[0008] In patients with pulmonary diseases, lung function is intrinsically heterogeneous, with some regions exhibiting little or no function while others are hyperfunctional. Lung SPECT and PET imaging enable visualization of this heterogeneous distribution in both ventilation and perfusion. Indeed, the radioactive concentration in a voxel is correlated with the intensity of the physiological phenomenon observed, namely pulmonary perfusion or ventilation.

[0009] However, a major current technical limitation of these imaging modalities is that no validated tool currently exists to provide a reliable and reproducible quantitative assessment of intrapulmonary functional distribution. Indeed, image interpretation remains qualitative, relying on visual assessment by the physician who describes whether a given lung region shows low or high function. In other words, there is no method allowing expressing whether, within a given ROI, the uptake intensity is low or high as compared to the rest of the lung.

[0010] So far, it has been known, in the absence of dedicated parameter, the use of the so-called Standardized Uptake Value (SUV) metric in PET, particularly in oncology, as a parameter for assessing response to treatment, for example. As explained hereafter, SUV is the ratio of the concentration of activity measured over an region of interest expressed in kilobecquerels per liter (kBq / L), to the theoretical concentration (ratio of radioactivity injected (expressed in kBq) to the volume of distribution (expressed in L)(17) for the said radioactivity. To calculate the SUV parameter, the volume of distribution is commonly estimated using the patient's weight expressed in kilograms (kg), on the assumption that the patient's weight is equal to a homogeneous volume of water. Present invention arose from cautiously and purposively designed experiments aimed at assessing the relevance of a quantitative approach to radiolabeled MAA uptake in regions of interest within the lung, exploring different parameters with a logical, progressive and adjusted approach depending on the findings found along the process, which is described further in the present application, notably in the Experimental Section. The invention therefore addresses the need for a method for quantifying the intensity of uptake in pulmonary ventilation perfusion SPECT / CT or PET / CT, and other needs as described herein, which was unmet so far.

[0011] The invention is intended to enable professionals having interest in quantifying regional lung function and monitoring its evolution, especially in clinical settings, to have an actual and reliable manner to do it. By quantifying regional lung function, aka quantifying local or locally, lung function, is it meant to quantitatively express or show whether the uptake intensity (implicitly, of a tracer), which reflects pulmonary perfusion or ventilatory function, is low or high, or at least determined or determinable within a range of values.

[0012] By “regional”, aka local or locally herein, reference is a made to a ROI where the PUV indicia is calculated.

[0013] In the context of nuclear medicine imaging such as PET / CT or SPECT / CT scans of the lung, the term "uptake" refers to the accumulation or concentration of a tracer, in particular a radiotracer, within a specific region of tissue, such as a region of interest (ROI) in the lung. This tracer is typically a radioactive substance administered to the patient (i.e., a radiotracer), which distributes within the body according to physiological processes (e.g., perfusion or ventilation in the lungs). Uptake is quantitatively measured as the amount of radioactivity detected in a given region, and it reflects the physiological activity or function of that region. For example, higher uptake in a lung region indicates greater accumulation of a tracer, which may correspond to higher perfusion or ventilation in that area, depending on the type of scan performed. In turn, the invention brings a possibility of quantification of the uptake of a radiotracer at the level of a ROI of a patient, the ROI being determined on PET / CT or SPECT / CT scans.

[0014] The invention is therefore of particular interest for pre-therapeutic assessment and followup of patients treated with thoracic radiotherapy, follow-up of patients with Chronic Obstructive Pulmonary Disease (COPD) and emphysema, notably those treated with endo-bronchial valve or rheoplasty, and for regional lung function assessment of lung cancer patients before surgery. More generally, the invention promises to be a true asset in all operations involving image contouring of the distribution of regional lung function, i.e., activity (radiotherapy, pulmonary embolism, chronic post embolic pulmonary hypertension).

[0015] In diseased patients, observing the distribution of regional lung function is meaningful. Indeed, notably in patients with pulmonary diseases, lung function is intrinsically heterogeneous, with some regions exhibiting little or no function while others are hyperfunctional. Lung SPECT and PET imaging enable visualization of this heterogeneous distribution in both ventilation and perfusion. The invention is aimed at providing reliable data, and is an easily implementable method, that can be used for both PET / CT and SPECT / CT, whereas, notably, no sound uptake quantification method is available to date in the field.

[0016] As shown in the experimental section herein, the method of the invention enables the provision of a pertinent quantitation tool, especially indicia and method of using the same, in lung imaging, an aspect that is critical for enhancing the precision, reproducibility, and diagnostic value of the scans, and thus answers unmet needs. The PUV indicia allows expressing whether, within a given ROI, the uptake intensity is low or high as compared to the rest of the lung of the patient. This parameter is therefore patient specific, and specific of the purpose sought, i.e., to quantitatively assess the pulmonary uptake of a tracer in a lung of a given patient. Determining a PUV indicia is, per se, equivalent to quantifying, in the considered region, tracer uptake value.

[0017] More generally, there is currently no validated tool to quantitatively describe intrapulmonary functional distribution.

[0018] The invention relates to a computer-implemented method for determining a PUV value for a region of interest on a lung scan, the lung scan being a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan image, said method comprising:

[0019] (a) aligning an anatomical image and a functional image such that both images have corresponding coordinates;

[0020] (b) determining a lung anatomical contour on the anatomical image and determining the Anatomical Lung Volume (ALV) of the region defined by the said lung anatomical contour, for example expressed in liter (L) unit;

[0021] (c) transferring the lung anatomical contour on the functional image;

[0022] (d) obtaining the total radioactivity (Ao) inside region defined by the lung anatomical volume (ALV) determined in (b), for example expressed in kBq unit;

[0023] (e) determining a region of interest (ROI) within the ALV and obtaining the radioactivity concentration inside the ROI (CROI), for example expressed in kBq / L unit;

[0024] (f) determining a PUV value for the region of interest (PUVROI), according to the formula:

[0025] PUVROI = (CROI X ALV) / Ao

[0026] The invention is concerned by the exploration of the organ that is the lung.

[0027] According to a particular embodiment, the computer-implemented method of the invention where a PUV indicia is determined, is a computer-implemented method for quantifying the intensity of uptake, , i.e., tracer uptake, in pulmonary ventilation perfusion scintigraphy coupled with computed tomography (SPECT / CT) or lung positron emission tomography coupled with computed tomography (PET / CT) on the basis of a lung scan that is a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan image. The invention defines steps enabling the computation of a so-called PUV value for a region of interest (ROI) in the scanned lung(s) of an individual, which can be an animal, especially a mammal, in particular an human. The imaged person can be a patient. The patient can be, in a non-limitative manner, treated with thoracic radiotherapy, or have, treated or not, lung cancer, Chronic Obstructive Pulmonary Disease (COPD) and emphysema, notably treated with endobronchial valve or rheoplasty. The invention also applies to follow-up patients (meaning that the method of the invention can be reiterated over time, and the obtained PUV values compared between them over time, for a same patient).

[0028] In PET / CT and SPECT / CT imaging, the anatomical image is provided by a Computer Tomography (CT) image. The functional image is provided by either a Positron Emission Tomography (PET) image or a Single-photon emission computed tomography (SPECT) image. According to a particular embodiment of the invention, the method requires that a lung anatomical image and a lung functional image be provided, for a same patient. These images can be subsequently aligned in step a) of the method of the invention. The manner such an alignment is carried out is conventional in the field, for example it is readily done by PET / CT and SPECT / CT image treatment devices or apparatuses necessarily used in the context of present invention.

[0029] An anatomical image is a computer tomography image.

[0030] A functional image is an image obtained after intravenous administration of a radiotracer that is accumulated in different concentration in different regions of the lung to visualize functional activities. A radiotracer is a biomolecule injected in low chemical quantities and labelled so that it can be detected by an external device such as a PET scanner or a SPECT scanner. The label is often a radionuclide. The radiotracer may be any tracer known to be able to be images by an imaging device such as PET or SPECT.

[0031] In the present invention, the functional image is provided by the PET or SPECT acquisition and therefore is a Single photon emission computed tomography image or is a Positron Emission Tomography image.

[0032] A functional PET or SPECT image provide primary images that are visualized values of radiation count values. In such images, sites having high-concentration of radiopharmaceutical are illuminated. In other words, the intensity of the pixels for a specific layer displayed on a 2D screen provide information regarding radiation count values. Radioactivity count values, however, are subject to various factors, and thus higher pixel values of specific pixels or in a region of interest do not always indicate abnormality in the corresponding tissue. There is therefore a need for standardization, which is the purpose of present invention.

[0033] According to a particular embodiment, the functional images are obtained after intravenous administration of macro-aggregated albumin (MAA) particles which are labeled with Technetium 99m or Gallium-68, a p+ isotope.

[0034] According to a particular embodiment, the functional images may be obtained after inhalation of a radioactive gas (e.g. Krypton 81 m) or aerosolized particles labeled with Technetium 99m or Gallium-68, which distributes through the airways into the lungs according to the regional ventilation.

[0035] In step a) of the method of the invention, the anatomical image and the functional image are aligned such that both images have corresponding coordinates. In other words, the anatomical image and the functional image are co-registered such that both images have corresponding pixels and voxels. Co-registering of the anatomical image and the functional image may be carried out by any suitable technique.

[0036] In step b), a lung anatomical contour is determined on the anatomical image and the Anatomical Lung Volume (ALV) of the region defined by the said lung anatomical contour is then determined.

[0037] Determining an anatomical contour on an anatomical CT image of a lung is conventional in the field. It can be done manually, i.e., the lung anatomical contour is manually delineated on the anatomical image, or it can be done by semi-automatic or automatic segmentation. For example, the lung contours in CT images can be extracted using an automated method based on Hounsfield units to find the two areas representing left and right lung. In addition, the contours may be manually corrected. An example of the lung anatomical contour is shown in Figure 1 . The shape of the lung in CT images is represented by contours drawn along the lung perimeter.

[0038] Accordingly, the volume (ALV) of the region defined by the said determined lung anatomical contour is determined in step b). This can be done conventionally in the field. For instance, the ALV is made of a plurality of voxels, and calculation of the volume of the ALV is made by summing up all the volumes of the voxels constituting the ALV. The ALV is for example expressed in liter (L) unit. Other units may be used, as long as they concern a volume. The skilled person will readily apprehend the fact that another volume unit used in the PUV formula of the invention will match the volume unit used for ALV, and vice-versa.

[0039] In step c), the lung anatomical contour is transferred on the functional image. In particular, the information related to the lung anatomical contour is transferred to the functional image without altering or modifying the intensity levels of the functional images, as conventionally done in the field. The manner such a transfer is carried out is conventional in the field, for example it is readily done by PET / CT and SPECT / CT image treatment devices or apparatus necessarily used in the context of present invention.

[0040] Once the lung anatomical contour has been transferred on the functional image, it is possible, in a step d), to obtain the total radioactivity (Ao) inside the region defined by the lung anatomical volume (ALV) previously determined in step b).

[0041] It is understood that the total radioactivity (Ao) inside the ALV, is a measure, which is done at the voxel level, i.e., at the level of the entire considered volume. For instance, for a particular layer displayed on a 2D image, each pixel of the anatomical image may have an intensity value corresponding to the radioactive concentration of the tracer measured in the pixel. The total radioactivity Aocorresponds to the sum of all pixel intensities for all the 2D layers within the anatomical contour. In other words, all voxels having a non-zero intensity are counted in the total radioactivity. Thus, Aocorresponds to the total radioactivity calculated within the anatomical contour, at the voxel level. For a particular layer displayed on a 2D image, the intensity of a pixel is denoted Ai and the total radioactivity is expressed in: i=N

[0042] ^0 in 2D layer = ; (threshold = 0) i=l wherein the number N represents the number of pixels contained in the considered 2D layer of the anatomical contour. The calculation at the voxel level takes into account all layers of the ALV.

[0043] The total radioactivity (Ao) is for example expressed in kBq unit. Other units may be used, as long as they concern an unit for measuring the level of radioactivity. The skilled person will readily apprehend the fact that another unit for measuring the level of radioactivity used in the PLIV formula of the invention will match the unit for measuring the level of radioactivity used for Ao, and vice-versa.

[0044] In step e), it is determined a region of interest (ROI) within the ALV and the concentration of radioactivity inside the ROI (CROI) is then obtained. The ROI can for example be anatomical regions of the lungs, such as lung segment(s) or lung lobe(s), or isodose regions of the lungs, i.e., region(s) of the lung, which have received a peculiar, identical, dose of radiotherapy. These examples are not limitative and are provided to illustrate the nature of a region of interest (ROI) within the ALV, the concentration of radioactivity of which is determined for implementing the method of the invention. According to a particular embodiment, it can be appreciated that the region of interest (ROI) can be a so-called pathological region, if the method is carried out for evaluating the same.

[0045] The concentration of radioactivity inside the ROI (CROI) is for example expressed in kBq / L unit. Other units may be used, as long as they concern a concentration over a volume ratio. The skilled person will readily apprehend the fact that other unit(s) used for a concentration over a volume ratio used in the PUV formula of the invention will match the units for the concentration or volume used for Ao and / or ALV as described herein, and vice-versa.

[0046] According to the invention, a ROI is to be found within the ALV region defined by the anatomical contour.

[0047] ROI contour delineation can be carried out, conventionally, in different ways. ROI contour delineation can be done manually (a ROI can be any region, examples of the nature of which are provided above). It can be done using a semi-automatic or an automatic method, noting that manual correction of automatically determined contours is generally possible

[0048] Once step e) has been completed, it is possible to determine, in a step f), a PUV value for the region of interest within the ALV, according to the formula:

[0049] PUV OI = (CROI X ALV) / Ao Expressed in layman terms, a PUV (Pulmonary Uptake Value) is a ratio of the concentration of activity measured in a region of interest (ROI) in the lung (for example expressed in kBq / L), to a theoretical homogeneous activity concentration in the lung (ratio of the total activity measured in the anatomical lung volume (for example expressed in kBq) over the anatomical lung volume (for example expressed in L).

[0050] The skilled person will readily appreciate that the PUV value has no unit. It is an indicator (similarly to the SUV parameter of the prior art), or differently said, an indicium (indicia as synomym). For this reason, the units in which CROI, ALV and Ao are expressed will match so that the formula is mathematically correct. While being an indicator however, it has an ability to appropriately quantify, quantitatively determine the uptake at the level of a voxel in a considered region in the lung, which makes it an unequaled parameter for assessing the same, in light of the prior art. Regarding quantitation, if a radiotracer distribution within the lungs is perfectly homogeneous, obtained PUV values equal 1 across all voxels of the pulmonary volume. A PUV value of 2 in a ROI then indicates that the local radioactivity concentration is twice that of the theoretical homogeneous distribution.

[0051] A Pulmonary Uptake Value, i.e., a Pulmonary Uptake Value indicia, is calculated for a region of interest (ROI) by the equation set above, i.e., PUVROI = (CROI X ALV) / Ao, with the explanations provided herein regarding the components of the equation.

[0052] As explained in the experimental section herein, the “volume of distribution”, which could have been assimilated to the patient's weight in a different implementation, is replaced by the anatomical lung volume measured by CT. The administered activity is replaced by the activity measured in the anatomical lung volume. This is in contrast to the use of the SUV value: in lung PET, using a so-called SUV value is of little relevance because the radiopharmaceutical's volume of distribution is mainly the lung, rendering senseless to use, after consideration (see experimental section) the patient's weight as the volume of distribution. Also, estimating the activity administered is more complicated in perfusion settings (where there is residual activity in the syringe) and impossible in ventilation settings. The invention also proves useful in this respect. Furthermore, a variable proportion of the activity injected during perfusion when perfusion settings is at stake, passes through the lungs and is distributed to other areas outside the field of acquisition, which is the lung. For at least these reasons, the invention provides outstanding advantages over the state of the art in the field. Drastic changes in calculations have been made.

[0053] In a particular embodiment, the method of the invention, as described in any embodiment disclosed herein, is, alternatively or in combination, for determining an indicia termed PUVRO for a region of interest on a lung scan, the indicia enabling quantifying the intensity of uptake of a tracer in the region of interest, when emission tomography coupled with computed tomography (PET / CT) or scintigraphy coupled with computed tomography (SPECT / CT) is carried out on a patient. According to a particular embodiment, combinable with any other embodiment described herein, the method of the invention is reiterated over at least one further region of interest, i.e., the steps a) to f) described above are carried out for a ROI that is distinct from the first ROI. According to a particular embodiment, the anatomical image can remain the same between the iterations made, and / or the values determined in step b) and / or d) can remain the same between the reiterations.

[0054] By “at least one further region of interest” it is meant an integer number from 1 to a higher integer number as meaningful, reasonable or appropriate. It can be 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10 or more. It can be 10, 20, 30, 40, 50 or more, or any interval of integer numbers between the numbers indicated in the present paragraph taken as boundaries, according to all possible combinations thereof.

[0055] According to a particular embodiment, the method is aimed at evaluating perfusional and / or ventilatory lung function. In a particular embodiment, the method is aimed at evaluating perfusional lung function. In a particular embodiment, the method is aimed at evaluating ventilatory lung function. The lung function evaluation is made possible by the calculation of a so- called PUV value for one or more ROI as described herein. Accordingly, a last or further step of the method as described in any embodiment herein, can be a step of evaluating perfusional and / or ventilatory lung function on the basis of a lung scan that is a PET / CT or a SPECT / CT lung scan image. The skilled person can readily use his / her common knowledge for such an evaluation, which is based on the inherent skills of a practitioner in the field of lung imagery especially in clinical settings. Examples (notably with Figures, especially interpreted Figures) are provided herein regarding manner of evaluating perfusional and / or ventilatory lung function based on calculation of the PUV indicia, i.e., manners of interpreting PUV indicia, their numerical values or evolution over time for these values, in a meaningful manner once determined. In a particular embodiment, this evaluation is available as soon as the PUV parameter is calculated (determined), given the inner significance it bears.

[0056] According to a particular embodiment, the method is a method for quantifying the intensity of uptake in pulmonary ventilation perfusion SPECT / CT or PET / CT, making use to this end of the calculation of a PUV value for one or more ROI as described herein. The quantification is made possible by the calculation of a so-called PUV value for one or more ROI as described herein. Accordingly, a last or further step of the method as described in any embodiment herein, can be a step of quantifying the intensity of uptake in pulmonary ventilation perfusion SPECT / CT or PET / CT on the basis of a lung scan that is a PET / CT or a SPECT / CT lung scan image, through the calculation of a PUV value for one or more ROI as described herein. As explanations of the meaning of the calculated PUV parameter, if the tracer, especially radiotracer, distribution within the lungs is perfectly homogeneous, the PUV equals 1 across all voxels of the pulmonary volume. Conversely, a PUV value higher than 1 indicates a higher lung function compared to homogenous distribution. For example, a PUV value of 2 indicates that the local radioactivity concentration is twice that of the theoretical homogeneous distribution. In contrast, a PUV value lower than 1 indicates a lower uptake. The closer the value is to 0, the lower the uptake is. This offers to the skilled person a quantitative manner of measuring uptake value, which is reliable, meaningful, patient-centered and patient-related, specifically. The determination if therefore specific.

[0057] It is reminded that there is no sound uptake quantification method available to date in the field, so that the invention enabling quantifying regional lung function of a patient, is of interest in monitoring the evolution of the condition of the patient, especially in clinical settings. The invention is therefore of particular interest for pre-therapeutic assessment and follow-up of patients treated with thoracic radiotherapy, follow-up of patients with Chronic Obstructive Pulmonary Disease (COPD), follow-up of patients with emphysema, notably those treated with endo-bronchial valve or rheoplasty, and for regional lung function assessment of lung cancer patients before surgery.

[0058] According to a particular embodiment, combinable with any other embodiment described herein, the method is carried out at a first point in time and then reiterated at least a second point in time for a same patient that is imaged over time, the at least second point in time being separate from the first point in time and at a later time with respect to the first point in time.

[0059] By “at least a second point in time” it is meant at least 2, 3, 4, 5, 6, 7, 8, 9, 10 or more. It can be 10, 20, 30, 40, 50 or more, or any interval of integer numbers between the numbers indicated in the present paragraph taken as boundaries, according to all possible combinations thereof. The described embodiment is meaningful when a follow-up of a patient is carried out, notably to determine whether the patients’ condition has changed, improved or not.

[0060] Accordingly, the invention also relates to a method for monitoring the condition of a patient over time, in particular a patient treated with thoracic radiotherapy, or diagnosed or treated for Chronic Obstructive Pulmonary Disease (COPD), or diagnosed or treated for emphysema, notably those treated with endo-bronchial valve or rheoplasty, and for assessing the condition of a lung cancer patient before surgery, and before and after surgery. According to a particular embodiment, such a method comprises a step of carrying out a method of determining a PUV value for a region of interest of the invention according to any embodiment disclosed herein. The method for monitoring can also comprise a step of reiterating carrying out a method of determining a PUV value for a region of interest of the invention according to any embodiment disclosed herein, at least a second point in time per the definition provided herein. The method for monitoring can also comprise a step of comparing the result of these methods carried out at several points in time, so as to conclude about the evolution of the condition of the monitored patient.

[0061] The invention also relates to a data processing apparatus configured for carrying out the method of the invention according to any embodiment described herein, in particular when provided with a PET or SPECT lung scan, especially a PET / CT or SPECT / CT lung scan.

[0062] The invention also relates to a computer program product comprising software code adapted to cause a computer to perform a method of the invention according to any embodiment described herein, in particular when provided with a PET or SPECT lung scan, especially a PET / CT or SPECT / CT lung scan.

[0063] The invention also relates to a non-transitory computer readable medium having stored thereon the computer program product described herein, notably in the above paragraph.

[0064] The invention also relates to a device for determining a PUV value for a region of interest on a PET / CT or SPECT / CT lung scan, said device comprising module(s) adapted to carry out steps (a) to (f) of a method of the invention according to any embodiment described herein, or comprising the data processing apparatus described herein, notably in the preceding paragraphs.

[0065] The invention also relates to a medical image apparatus comprising an anatomical image device, a functional image device and the device described herein, notably in the above paragraph, for determining a PUV value for a region of interest on a PET / CT or SPECT / CT lung scan.

[0066] In these embodiments involving “a PUV”, i.e., a pulmonary uptake value, it is meant that either one PUV can be determined, or more than one values can be determined, in accordance with the embodiments described herein where more than one PUV is determined. According “a “pulmonary uptake value” or “PUV” is used generically in this context.

[0067] Other examples and features of the invention will be apparent when reading the examples and the Figures, which illustrate the experiments conducted by the inventors, in complement to the features and definitions given in the present description.

[0068] Legend of the Figures

[0069] Fig. 1 Automatic contouring of anatomical lung volume (ALV) on CT images using an algorithm based on Hounsfield unit values by image thresholding and subtraction technique. The anatomical lung volume (ALV) is represented by the combination of the volume contoured in blue on the left part of the Figure (thin gray line in grey levels) and the volume contoured in yellow on the right part of the Figure (thin white line in grey levels).

[0070] Fig. 2 depicts measurement of ALV depending upon patient’s weight. Correlation coefficient was 0.1439.

[0071] Fig. 3 depicts the distribution of the percentage of retention of 68Ga-MAA within the lungs. The mean (SD; min; max) percentage of retention of 68Ga-MAA within the lungs was 58.2% (7.7%; 39,3%; 76,3%).

[0072] Fig. 4 depicts the difference of lung retention between MO and M3 in patients who had a scan at 3 months. There was a wide variability of lung retention in a same patient, ranging from -22,9% to 33,7%. Fig. 5 shows the baseline 68Ga-MAA PET / CT scans of two patients included in the PEGASUS trial. This figure depicts a comparison of quantification metrics in patients with decorrelated ALV and weight. Visual analysis of68Ga-MAA PET / CT images of patient 1 showed preserved lung function in the right upper lobe parenchyma surrounding the lung nodule, with high68Ga-MAA uptake in this region (see arrows) as compared with the lower lobes. In contrast, visual analysis of68Ga-MAA PET / CT images of patient 2 showed altered lung function in the right lung parenchyma surrounding the lung nodule (see arrows). Quantitative metrics (SUV, SUVp, PUV) were calculated in the 50 Gy isodose for each patient. While the regional lung function was preserved in patient 1 and strongly altered in patient 2, SUV was inconsistently low in patient 1 (SUV 4,84) and high in patient 2 (SUV 15.73), In contrast, the PUV value was low (0.79) for patient 2, while it was higher for patient 1 (1 .20), consistent with the visual analysis. In the present cases, wide differences between SUV and PUV were due to wide differences in the ALV and the patient’s weights. ALV of patients 1 and 2 was 6155 ml and 2132 ml, while patient’s weight was 51 kg and 85 kg, respectively. The PUV method, which is normalized to the volume of distribution of interest of the tracer, provided quantitative results consistent with the visual analysis. Of note is that the SUVp were not inconsistent in these cases, with lower values in patient 2. This could be explained by a close percentage of tracer retention within the lungs (48% vs 53%). The percentage of activity in the 50 Gy isodose was 0.15% and 1 .24%, respectively, not reflecting the low relative uptake in patient 2. This was explained by different isodose volume (10 vs 28 ml).

[0073] Fig. 6 shows68Ga-MAA PET images images at M0 and M3 of a patient treated with radiotherapy of a right lower lobe lung carcinoma. This figure depicts a comparison of quantitative metrics in a patient with different percentage of [68Ga]Ga-MAA lung retention between M0 and M3. Figure 4 shows68Ga-MAA PET images at M0 and M3 of a patient treated with radiotherapy of a right lower lobe lung carcinoma. Visual analyses of68Ga-MAA PET images at 3 months demonstrated a reduction of lung perfusion in the irradiated lung region (arrow) as compared with the baseline68Ga-MAA PET performed at M0. Quantitative metrics (SUV, SUVp, PUV) were calculated in the 50 Gy isodose. SUV and SUVp values inconsistently increase (+ 5% and + 4%) between M0 and M3. In contrast, there was a 39% decrease of the PUV value, consistent with the visual analysis. In the present case, the main explanation for the failure of the SUV and SUVp measurements was a wide difference in tracer lung retention between M0 and M3 between M0 (47%) and M3 (81%). This unpredictable variation does not invalidate the PUV method since the metrics is normalized to the total activity in the ALV and not to the injected activity.

[0074] Fig. 7 shows68Ga-MAA PET images images at M0 and M3 of a patient treated with radiotherapy of a right upper lobe carcinoma. This Figure depicts a comparison of quantitative metrics in a patient with different percentage of [68Ga]Ga-MAA lung retention between M0 and M3. Figure 5 shows68Ga-MAA PET images at MO and M3 of a patient treated with radiotherapy of a right upper lobe lesion. Visual comparison of68Ga-MAA PET images a MO and M3 is challenging because of the heterogeneity of lung perfusion in this region, illustrating the complexity of a visual subjective analysis and the need for quantitative objective metrics. Nevertheless, no major changes or at most a slight decrease of lung perfusion was observed. Quantitative metrics (SUV, SUVp, PUV) were calculated in the 50 Gy isodose. SUV and SUVp methods found a significant decrease of lung perfusion between MO and M3, inconsistent with the visual analyses. PUV metrics were more coherent with a 9% decrease between MO and M3. Again, the main explanation for the failure of the SUV and SUVp measurements was a difference in the lung retention of the injected activity between MO (64%) and M3 (47%). This unpredictable variation does not invalidate the PUV method since the metrics is normalized to the total activity in the ALV and not to the injected activity.

[0075] Experimental Section and Results

[0076] A. Introduction - Experimental progression

[0077] Lung perfusion positron emission tomography coupled with computed tomography (PET / CT) is a new imaging modality for assessing regional lung function. (1 ). Like lung perfusion scintigraphy, this test assesses regional lung perfusion distribution by embolizing albumin macroaggregates (MAA) in pulmonary capillaries (2). However, MAA particles are labeled with gallium-68 (68Ga-MAA), a B+ emitter, instead of technetium-99m.68Ga has a short half-life of 68 minutes and is readily available from an on-site generator. (3). PET technology can be considered as more interesting than the scintigraphy, including single-photon emission computed tomography (SPECT), with notably much better sensitivity and resolution, as well as more favorable contrast due to more effective corrections, enabling more accurate assessment and quantification of regional lung functionality (4)(5). Lung PET / CT imaging has already shown promising results in various clinical contexts, such as the diagnosis of pulmonary embolism (6)(7) assessment of pulmonary function prior to lung surgery (7)(8) and radiotherapy (RT) planning (9).

[0078] As stated before, a current limitation to the use of lung PET / CT or lung SPECT / CT in clinical settings is the absence of a tool to reliably and reproducibly quantify tracer uptake within the lung.

[0079] We sought to assess the relevance of a quantitative approach to radiolabeled MAA uptake in regions of interest within the lung.

[0080] The most commonly used metric in PET is the Standard Uptake Value (SUV), particularly in oncology as a parameter for assessing response to treatment, for example. SUV is the ratio of the concentration of activity measured over a region of interest, expressed in kilobecquerels per liter (kBq / L), to the theoretical concentration (ratio of radioactivity injected (expressed in kBq) to the volume of distribution (expressed in L)(17).

[0081] Activity Concentration measured in the region of interest -7— 0 x Distribution Volume (L) SUV = - - - -

[0082] Injected Activity (kBq)

[0083] In the context of therapeutic evaluation in oncology, particularly when using 18F-FDG as a radiopharmaceutical, the volume of distribution is commonly estimated using the patient's weight expressed in kilograms (kg), on the assumption that the patient's weight is equal to a volume of homogeneous water, and therefore that 1 L of volume is equal to 1 kg. According to this simplification, taking into account the correction for radioactive decay, if we were to measure the average SUV throughout the body, it would be equal to 1 .

[0084] However, in the case of68Ga-MAA PET scans, virtually all the radiopharmaceutical is fixed in the lungs. It is therefore legitimate to ask whether it is appropriate to consider the patient's volume, estimated by weight, as the dilution volume of the drug, rather than the lung volume itself.

[0085] Furthermore, since we are not interested in the non-lung portion of68Ga-MAA uptake in the assessment of lung function, and since PET imaging is a quantitative imaging technique, we may also ask whether it is relevant to consider the activity measured in the syringe pre-injection, rather than the activity directly measured in the lung areas.

[0086] We sought to assess the relevance of a quantitative approach based on normalizing the activity measured to the anatomical lung volume (ALV) and to the total activity measured in the ALV.

[0087] B. Hypothesis 1: Is It appropriate to use SUV normalized to patient weight as used in oncology in lung PET?

[0088] As a reminder, SUV is the expression of the concentration of activity measured over an region of interest (expressed in kBq / L), normalized with respect to the radioactivity injected (expressed in kBq) and the volume of distribution (expressed in L). Applied to imaging, particularly in oncology, it is assumed that the patient's weight is equal to a homogeneous volume of water, and therefore that 1 L of volume is equal to 1 kg, hence the expression of SUV as the activity concentration measured in the region of interest, normalized to the injected radioactivity and the patient's weight expressed in kg.

[0089] Activity Concentration measured in the region of interest -7— 0 x Patient's Weight(kg) SUV = - -

[0090] Injected Activity (kBq) The aim of this first working hypothesis was to assess the applicability of this measurement to lung PET.

[0091] B.1. Materials and methods

[0092] \Ne were interested in lung perfusion PET / CT scans performed in patients participating in the PEGASUS trial (NCT04942275), a pilot study evaluating the possibility of sparing lung functional areas by integrating functional mapping obtained with68Ga-MAA perfusion PET / CT into the planning of stereotactic radiotherapy of the lung (SBRT) in patients referred for radiotherapy of primary or secondary tumors at Brest University Hospital (16). Patients who agreed to take part in the study had a lung perfusion PET / CT scan with68Ga-MAA as part of their pre-therapy assessment and 3 months after completion of radiotherapy. The study involved 60 patients.

[0093] Lung perfusion PET / CT was performed on a Biograph vision 600 PET / CT digital device (Siemens Healthineers, Knoxville, TN, USA), immediately after intravenous administration of a fixed 50 MBq activity of68Ga-MAA. PET acquisitions were coupled with a CT scan.

[0094] Images were analyzed using MIM software (MIM 7.3.3; MIMSoftware). For example, the MIM Encore® software (https: / / go. mimsoftware.com / nuclear-medicine / oncologic-pet ) is an End- to-End Solution for Nuclear Medicine, which provides a vendor-neutral solution for reconstruction, image processing, reading, and reporting. It thus assists Nuclear Medicine imaging workflow. It includes several tools such as SPECTRA Recon®, which supports SPECT / CT image reconstruction and CT attenuation correction for virtually any camera manufacturer, or SPECTRA Quant® which allows to convert image counts to, for instance, SUV. It also provides a comprehensive suite of standard Nuclear Medicine processing tools, including lung quantification for SPECT / CT and liver functional analysis. Automated processing is available with MIM Assistant®. Generally, the MIM software, along with other software in the field, is an example of software which can be used for implementing present invention.

[0095] The first step was to apply segmentation to obtain automatic contouring of the anatomical lung volume (ALV) on CT images according to an algorithm based on Hounsfield unit values using an image thresholding and subtraction technique, as illustrated in Figure 1. On Figure 1 , the anatomical lung volume (ALV) is represented by the combination of the volume contoured in blue on the left part of the Figure (thin gray line in grey levels) and the volume contoured in yellow on the right part of the Figure (thin white line in grey levels). ALV contours were examined visually and modified if relevant (in particular to integrate tumors with much higher Hounsfield densities).

[0096] Voxel values were converted to SUV using the injected dose (collected from the PET DICOM data and entered manually during injection, with automatic decay calculation) and the patient's recorded weight (measured using a bathroom scale). DICOM data is raw PET data available in a so-called DICOM file, generally having a “.dem” extension file suffix. The anatomical volumes (ALV) were transferred to the PET images (the PET and CT images were previously merged to have matching coordinates). The mean SUV value within the ALV, as well as the standard deviation and extrema, were then measured in the ALV. The correlation between the patient’s weight and the ALV was also assessed.

[0097] B.2. Results

[0098] Measuring the mean SUV based on a calculation according to the activity injected in kBq and the patient's weight in kg (SUV), we find a mean SUV within the anatomical lung volume of 9.90 (standard deviation 5.24) with a minimum of 0.11 and a maximum of 22.52.

[0099] Figure 2 shows the ALV according to the patient’s weight. Correlation coefficient was 0.1439.

[0100] B.3. Discussion

[0101] Applying the SUV formula yields heterogeneous mean SUVs, with a mean of 9.90 and an inconsistently wide range of values from 0.11 to 22.52. These results are probably largely explained by the normalization step to patient weight.

[0102] This normalization variable is not necessarily relevant to lung perfusion PET, as most of the68Ga-MAA injected is embolized in the pulmonary capillaries. Studies assessing the biodistribution of labeled MAA particles in animals describe almost complete retention of activity in the lungs from 5 min to 30, 45, 60 min or 4 h after injection of68Ga-MAA particles, and very low activity in other organs, notably the liver. Experiments in rats and mice have shown that over 80% of the injected activity was localized in the lungs from 15 min to 4 h post-injection (18). Ament et al (19) who carried out an exploratory study on five patients with clinical suspicion of PE who underwent V / Q PET / CT (V / Q stands for Ventilation / Quantity, with “Quantity” being a synonym for “Perfusion”), observed that perfusion imaging was homogeneous in most cases, with no significant retention or visual uptake of68Ga-MAA particles in the liver.

[0103] Consequently, the volume of distribution of68Ga-MAA would correspond to the anatomical lung volume, and not to the patient's volume or weight. Above all, as illustrated in Figure 2, lung volume is not correlated with patient weight.

[0104] This seems to support the hypothesis that SUV is normalized according to anatomical lung volume (ALV), which can be measured using a thoracic CT scan.

[0105] B.4. Conclusion The use of SUV does not appear to be relevant for PET quantification of lung perfusion, as it is normalized to the patient's weight, which is decorrelated from the radiotracer's actual volume of distribution, i.e. the lung volume.

[0106] Since PET acquisitions are combined with a thoracic CT scan, a reliable measurement of anatomical lung volume is available when performing this examination.

[0107] Our second hypothesis was therefore to normalize SUV, not to the patient's weight, but to the anatomical lung volume (ALV) that can be measured using a thoracic CT scan.

[0108] C. Hypothesis 2: Is It appropriate to use SUV normalized to anatomical lung volume (ALV) measured with thoracic CT?

[0109] Knowing the theoretical injected activity corrected for radioactive decay, and assuming that all68Ga-MAA are embolized within the volume of distribution represented by the anatomical lung volume, the average SUV within the anatomical lung volume should be equal to 1.

[0110] SUV is the expression of the activity concentration measured in a voxel (kBq / L), normalized with respect to the injected activity (kBq) and the volume of distribution (L). We have already seen that it seems more appropriate to take the anatomical lung volume as the volume of distribution, given that most of the injected68Ga-MAA- is embolized in the lung volume and that the lung volume is not correlated with patient weight.

[0111] Based on the assumption that all injected68Ga-MAA- are embolized within the lung volume, and knowing the theoretical injected activity corrected for decay, the second objective was to measure the average SUVp within the lung volume as a function of injected activity according to the following formula (SUVp).

[0112] Activity Concentration x Anatomic Lung Volume(L) SUVp = - - - - -

[0113] Injected Activity (kBq)

[0114] C.1. Materials and methods

[0115] For this purpose, we again used lung PET / CT scans from the 60 patients in the PEGASUS study, as described in the first working hypothesis. Voxel values were converted to SUVp using the injected dose (collected from the PET DICOM data and entered manually during injection, with automatic decay calculation) and the anatomical lung volume (ALV). As described above, the ALV was automatically contoured on CT images according to an algorithm based on Hounsfield unit values using an image thresholding and subtraction technique. \Ne calculated the mean SLIVp within the ALV of all lung PET scans, with standard deviation and extrema. A software such as the MIM software cited herein can be used for SUVp calculation.

[0116] The percentage of retention of68Ga-MAA within the lungs was calculated as follows: (Total Pulmonary Activity measured in the ALV (Bq) / Injected activity (Bq))*100. It was calculated at MO and M3.

[0117] C.2. Results

[0118] Measuring the mean SUVp based on a calculation according to the activity injected in kBq and the anatomical lung volume measured on the thoracic CT image, we find a mean total lung SUV of 0.54 (standard deviation 0.09) with a minimum at 0.12 and a maximum at 0.65.

[0119] Fig 3 shows the distribution of the percentage of retention of 68Ga-MAA within the lungs. The mean (SD; min; max) percentage of retention of 68Ga-MAA within the lungs was 58.2% (7.7%; 39,3%; 76,3%).

[0120] Fig 4 shows the difference of lung retention between M0 and M3 in patients who had a scan at 3 months. There was a wide variability of lung retention in a same patient, ranging from - 22,9% to 33,7%.

[0121] C.3. Discussion

[0122] In this working hypothesis, SUVp was expressed as the concentration of activity (kBq / L), normalized with respect to injected activity (kBq) and ALV (L). The mean SUVp within the anatomical lung volume of the lung PET scans was 0.54 (standard deviation 0.09), with a minimum at 0.12 and a maximum at 0.65. This should be equal to or very close to 1 , on the assumption that the decay-corrected administered activity is known and that all68Ga-MAA in the injection syringe are embolized in the lung volume. The average SUV was therefore well below 1.

[0123] The mean (SD) percentage of retention of 68Ga-MAA within the lungs was 58.2% (7.7%). Therefore, a significant part of the injected activity is not localized in the lungs.

[0124] A first explanatory factor is the actual activity injected, possibly lower than the theoretical activity. The administered activity is in fact collected from the PET DICOM data, entered manually during injection with an automatic decay calculation. One explanation for this error could be that some of the68Ga-MAA may have become trapped in the injection circuit. To verify this, we measured the residual activity of empty syringes on a number of patients, and found a residual activity corresponding to around 10% of the theoretical injected activity.

[0125] Another explanation could be the biodistribution of68Ga-MAA once injected. Previous studies have shown that size is an important factor in particle biodistribution. If particles were larger than 0.08 m and smaller than 10 pm, the target organs were the reticuloendothelial system (RES) and bone instead of lung (18). This size depends in part on the preparation of the MAA, in particular the centrifugation technique used (18). Finally, another explanation could be the presence of free68Ga within the preparation, which could contribute to the difference between the injected activity and the total activity measured in the lungs. In any event, it appears that factors impact the use of a parameter such as the SUVp parameter investigated in the present study.

[0126] Another important result is that the percentage of retention of the tracer within the lungs varies greatly from one patient to another, but also from one scan to another performed on the same patient (See Figure 4).

[0127] Of note is that these results could be extrapolated to perfusion SPECT / CT imaging using99mTc-MAA, given that99mTc-MAA and68Ga-MAA particles have the size and physical properties. Similarly, for lung ventilation SPECT / CT or PET / CT imaging, using either aerosols or inert gases, it is impossible to precisely measure the exact radioactivity inhaled by the patient. It is therefore impossible to reliably normalize the activity to the activity administrated to the patient. For this further reason, the method of the invention is further pertinent in lung ventilation imaging where it is impossible to reliably determine the dose inhaled by the patient.

[0128] C.4. Conclusion

[0129] This work led us to question the reliability of the measurement of the activity actually injected and the assumption that all the activity administered was distributed to the lungs.

[0130] In order to overcome the unreliability of injected activity for SUV calculation, we propose an approach consisting in replacing the administered activity by the activity actually present within the anatomical lung volume, which can be measured thanks to the contribution of PET or SPECT technology enabling absolute measurement of this activity (in kBq / L). The PUV is then expressed according to the following formula:

[0131] Activity measured in the region of interest -7— 0 x Anatomic Lung Volume (L) PUV = - - - -

[0132] Total Pulmonary Activity measured in the ALV (kBq)

[0133] The Anatomic Lung Volume is discussed in the “C. Hypothesis 2” section above and corresponds to the ALV parameter in the PUVROI = (CROI X ALV) / Aoformula defined herein.

[0134] The Activity measured in the region of interest corresponds to the CROI parameter in the PUVROI = (C OI X ALV) / Aoformula defined herein.

[0135] The Total Pulmonary Activity measured by PETin the ALV corresponds to the Ao parameter in the PUV OI = (C OI X ALV) / Aoformula defined herein.

[0136] The table below provides a summary of the results of the average uptake measurement in the ALV measured using the 3 methods: SUV, SUVp and PUV. Table 1. Measurement of fixation in VP A by 3 methods

[0137] By definition, the average PUV in the volume used for normalization is always equal to 1 . On the other hand, PUV variations within the ALV are more closely correlated with local lung function. Definition of “local” is provided in the above description. As a concrete explanation based on examples: if the radiotracer distribution within the lungs is perfectly homogeneous, the PUV equals 1 across all voxels of the pulmonary volume. However, a PUV value higher than 1 indicates a higher lung function compared to homogenous distribution. A PUV value of 2 indicates that the local radioactivity concentration is twice that of the theoretical homogeneous distribution. In contrast, a PUV value lower than one indicate low uptake. The closer the value is to 0, the lower is the uptake.

[0138] D. Illustrative cases assessing the consistency of the PUV method in various scenarios in radiation therapy

[0139] A current limitation to the use of lung SPECT or PET imaging, especially in clinical settings is the absence of a tool enabling to reliably and reproducibly quantify tracer uptake within the lung. Thus, tracer uptake within a region of interest is currently only described qualitatively based on a visual and subjective interpretation (e.g.”low”, “mild”, “high” uptake).

[0140] Having quantitative measurement tools in lung imaging is critical for enhancing the precision, reproducibility, and diagnostic value of the scans. Here are some key reasons:

[0141] Increased accuracy: Quantitative measurements would provide objective numerical values, reducing the subjectivity involved in the current qualitative visual interpretations. This would improve the accuracy in assessing regional lung function.

[0142] Reproducibility: Quantitative data allows for more consistent results between different scans and practitioners, which is crucial for long-term patient follow-up.

[0143] - Tracking changes: Quantitative measurements enable precise monitoring of changes over time, such as treatment response.

[0144] - Standardization and comparison: Quantitative tools allow comparisons between different patients, institutions, and imaging devices, facilitating the establishment of clinical norms. Improved communication: Quantitative results provide clearer communication between healthcare professionals (Nuclear medicine physicians, pneumologists, etc.) and increase confidence in diagnostics and treatment planning.

[0145] - Clinical decision support: Quantitative data could be integrated into predictive models and Al algorithms, improving clinical decision-making by providing more detailed and accurate information.

[0146] In order to assess the reliability of the PUV measurement in clinical practice, we assessed the method in patients included in the PEGASUS trial (NCT04942275),

[0147] As described above, the PEGASUS trial included patients treated with radiotherapy of primary or secondary lung tumors. Patients had68Ga-MAA perfusion PET / CT as part of their pretherapy assessment (MO) and 3 months after completion of radiotherapy (M3). The aim of this follow-up scan at 3 months was to evaluate the effect of radiation on regional lung perfusion. Indeed, a decrease of regional lung function has been reported in some patients after radiotherapy. An early alteration of regional lung function may be predictive of long-term lung toxicity and fibrosis. An early prediction of the risk of long term lung fibrosis may improve the prognosis of these patients thanks to early dedicated treatments.

[0148] However, such an approach requires reliable and reproducible quantitative metrics of regional lung function.

[0149] In the following illustrative cases from the PEGASUS trial, we assessed the consistency of the PUV method in different scenarios, as shown below.

[0150] Illustrative case 1 - Figure 5

[0151] Figure 5 shows the baseline68Ga-MAA PET / CT scans of two patients included in the PEGASUS trial.

[0152] Visual analysis of68Ga-MAA PET / CT images of patient 1 (see Figure 5) showed preserved lung function in the right upper lobe parenchyma surrounding the lung nodule, with a high68Ga- MAA uptake in this region (see arrows) as compared with the lower lobes.

[0153] In contrast, visual analysis of68Ga-MAA PET / CT images of patient 2 (see Figure 5) showed altered lung function in the right lung parenchyma surrounding the lung nodule (see arrows).

[0154] Quantitative metrics (SUV, SUVp, PUV) were calculated in the 50Gy isodose for each patient - see Tables 2 and 3 below.

[0155] While the regional lung function was preserved in patient 1 and strongly altered in patient 2, SUV was inconsistently low in patient 1 (SUV 4,84) and high in patient 2 (SUV 15.73) (Table 2). Table 2

[0156] In contrast, the PUV value was low (0.79) for patient 2, while it was higher for patient 1 (1.20), consistent with the visual analysis (Table 3). The value of 1 ,20 for the PUV parameter expresses that the fixation is increased: this corroborates with what is visually seen by a practitioner, here expressed as a meaningful number, which is consistent with the observation.

[0157] Table 3

[0158] In the present cases, wide differences between SUV and PUV were due to wide differences in the ALV and the patient’s weights. ALV of patients 1 and 2 was 6155 ml and 2132 ml, respectively, while patient’s weight was 51 Kg and 85 Kg, respectively.

[0159] The value of 0,79 for the PUV parameter expresses that the fixation is weakened (diminished): this corroborates with what is visually seen by a practitioner, here expressed as a meaningful number, which is consistent with the observation.

[0160] The PUV method, which is normalized to the volume of distribution of interest of the tracer, provided quantitative results consistent with the visual analysis. The PUV method applies to and is specific to a patient, which is also a further advantage in the daily practice.

[0161] Of note is that the SUVp were not inconsistent in these cases, with lower values in patient 2. This could be explained by a close percentage of tracer retention within the lungs (48% vs 53%).

[0162] Illustrative case 2 - Figure 6

[0163] Figure 6 shows68Ga-MAA PET images at MO and M3 of a patient treated with radiotherapy of a right lower lobe lung carcinoma.

[0164] Visual analyses of68Ga-MAA PET images at 3 months clearly demonstrated a reduction of lung perfusion in the irradiated lung region (arrow) as compared with the baseline68Ga-MAA PET performed at MO.

[0165] Quantitative metrics (SUV, SUVp, PUV) were calculated in the 50Gy isodose (Table 4).

[0166] Table 4 SUV and SUVp values inconsistently increase (+5% and +4%) between MO and M3.

[0167] In contrast, there was a -39% decrease of the PUV value, consistent with the visual analysis.

[0168] In the present case, the main explanation for the failure of the SUV and SUVp measurements was a wide difference in the lung retention of the injected activity between MO (47%) and M3 (81%). This unpredictable variation does not invalidate the PUV method since the metrics is normalized to the total activity in the ALV and not to the injected activity.

[0169] Illustrative case 3 - Figure 7

[0170] Visual comparison of 68Ga-MAA PET images a MO and M3 on Figure 7 is challenging because of the heterogeneity of lung perfusion in this region, illustrating the complexity of a visual subjective analysis and the need for quantitative objective metrics. Nevertheless, no major changes or at most a slight decrease of lung perfusion was observed.

[0171] Quantitative metrics (SUV, SUVp, PUV) were calculated in the 50Gy isodose (Table 5).

[0172] Table 5

[0173] SUV and SUVp values found a significant decrease of lung perfusion between MO and M3, inconsistent with the visual analyses.

[0174] PUV metrics were more coherent with a 9% decrease between MO and M3.

[0175] Again, the main explanation for the failure of the SUV and SUVp measurements was a difference in the lung retention of the injected activity between MO (64%) and M3 (47%). This unpredictable variation does not invalidate the PUV method since the metrics is normalized to the total activity in the ALV and not to the injected activity.

[0176] Conclusion

[0177] Quantification of lung perfusion on PET / CT is feasible with an approach based on PUV calculated from anatomical lung volume measured by thoracic CT and total lung activity measured by PET in the ALV. The method showed consistent results as compared to the current visual analysis, especially in challenging scenarios with variable rate of tracer retention in the lungs or variable anatomical lung volumes. The PUV method may enhance the precision, reproducibility, and diagnostic value of lung imaging. References

[0178] 1 . Le Roux PY, Hicks RJ, Siva S, Hofman MS. PET / CT Lung Ventilation and Perfusion Scanning using Galligas and Gallium-68-MAA. Seminars in Nuclear Medicine. Jan 1 , 2019;49(1 ):71 -81 .

[0179] 2 Blanc-Beguin F, Hennebicq S, Robin P, Tripier R, Salaun PY, Le Roux PY. Radiopharmaceutical Labelling for Lung Ventilation / Perfusion PET / CT Imaging: A Review of Production and Optimization Processes for Clinical Use. Pharmaceuticals. May 2022;15(5):518.

[0180] 3. Blanc-Beguin F, Masset J, Robin P, Tripier R, Hennebicq S, Guilloux V, et al. Fully Automated 68Ga-Labeling and Purification of Macroaggregated Albumin Particles for Lung Perfusion PET Imaging. Frontiers in Nuclear Medicine [Internet]. 2021 [cited May 30, 2023];1 . Available from: https: / / www.frontiersin.Org / articles / 10.3389 / fnume.2021 .778191

[0181] 4. Hicks RJ, Hofman MS. Is there still a role for SPECT-CT in oncology in the PET-CT era? Nat Rev Clin Oncol. Dec 2012;9(12):712-20.

[0182] 5. Le Roux PY, Robin P, Salaun PY. New developments and future challenges of nuclear medicine and molecular imaging for pulmonary embolism. Thromb Res. March 2018;163:236-41 .

[0183] 6. Hofman MS, Beauregard JM, Barber TW, Neels OC, Eu P, Hicks RJ. 68Ga PET / CT Ventilation- Perfusion Imaging for Pulmonary Embolism: A Pilot Study with Comparison to Conventional Scintigraphy. Journal of Nuclear Medicine. Oct 1 , 201 1 ;52(10):1513-9.

[0184] 7. Le Roux PY, Iravani A, Callahan J, Burbury K, Eu P, Steinfort DP, et al. Independent and incremental value of ventilation / perfusion PET / CT and CT pulmonary angiography for pulmonary embolism diagnosis: results of the PECAN pilot study. Eur J Nucl Med Mol Imaging. Jul 1 , 2019;46(8):1596-604.

[0185] 8. Siva S, Hardcastle N, Kron T, Bressel M, Callahan J, MacManus MP, et al. Ventilation / Perfusion Positron Emission Tomography-Based Assessment of Radiation Injury to Lung. International Journal of Radiation Oncology*Biology*Physics. 1 Oct 2015;93(2):408-17.

[0186] 9. Siva S, Thomas R, Callahan J, Hardcastle N, Pham D, Kron T, et al. High-resolution pulmonary ventilation and perfusion PET / CT allows for functionally adapted intensity modulated radiotherapy in lung cancer. Radiotherapy and Oncology. May 1 , 2015;115(2) :157-62.

[0187] 10.Bucknell NW, Hardcastle N, Bressel M, Hofman MS, Kron T, Ball D, et al. Functional lung imaging in radiation therapy for lung cancer: A systematic review and meta-analysis. Radiotherapy and Oncology. Nov 1 , 2018;129(2):196-208.

[0188] 1 1 Lucia F, Rehn M, Blanc-Beguin F, Le Roux PY. Radiation Therapy Planning of Thoracic Tumors: A Review of Challenges Associated With Lung Toxicities and Potential Perspectives of Gallium-68 Lung PET / CT Imaging. Frontiers in Medicine [Internet]. 2021 [cited May 30, 2023];8. Available from: https: / / www.frontiersin.Org / articles / 10.3389 / fmed.2021 .723748

[0189] 12. De Bari B, Deantonio L, Bourhis J, Prior JO, Ozsahin M. Should we include SPECT lung perfusion in radiotherapy treatment plans of thoracic targets? Evidences from the literature. Critical Reviews in Oncology / Hematology. June 1 , 2016;102:1 1 1 -7.

[0190] 13 Eslick EM, Stevens MJ, Bailey DL. SPECT V / Q in Lung Cancer Radiotherapy Planning. Seminars in Nuclear Medicine. Jan 1 , 2019;49(1 ):31 -6.

[0191] 14.Roux PYL, Siva S, Steinfort DP, Callahan J, Eu P, Irving LB, et al. Correlation of 68Ga Ventilation- Perfusion PET / CT with Pulmonary Function Test Indices for Assessing Lung Function. Journal of Nuclear Medicine. Nov 1 , 2015;56(1 1 ):1718-23. 15.Le Roux PY, Siva S, Callahan J, Claudic Y, Bourhis D, Steinfort DP, et al. Automatic delineation of functional lung volumes with 68Ga-ventilation / perfusion PET / CT. EJNMMI Res. 10 Oct 2017;7(1 ):82.

[0192] 16. Pinot F, Bourhis D, Bourbonne V, Floch R, Mauguen M, Blanc-Beguin F, et al. New Automated Method for Lung Functional Volumes Delineation with Lung Perfusion PET / CT Imaging. Cancers. Jan 2023;15(7) :2166.

[0193] 17.Thie JA. Understanding the Standardized Uptake Value, Its Methods, and Implications for Usage. Journal of Nuclear Medicine. Sep 1 , 2004;45(9):1431 -4.

[0194] 18.Shanehsazzadeh S, Jalilian AR, Lahooti A, Geramifar P, Beiki D, Yousefnia H, et al. Preclinical Evaluation of 68Ga-MAA from Commercial Available 99mTc-MAA Kit. Iran J Pharm Res. 2017;16(4) :1415-

[0195] 23.

[0196] 19. Ament SJ, Maus S, Reber H, Buchholz HG, Bausbacher N, Brochhausen C, et al. PET Lung Ventilation / Perfusion Imaging Using 68Ga Aerosol (Galligas) and 68Ga-Labeled Macroaggregated Albumin. In: Baum RP, Rbsch F, editors. Theranostics, Gallium-68, and Other Radionuclides. Berlin, Heidelberg: Springer; 2013. p. 395-423 (Recent Results in Cancer Research).

[0197] 20. Bajc M, Schumichen C, Gruning T, Lindqvist A, Le Roux PY, Alatri A, et al. EANM guideline for ventilation / perfusion single-photon emission computed tomography (SPECT) for diagnosis of pulmonary embolism and beyond. European journal of nuclear medicine and molecular imaging. 2019;46:2429-51 . doi : 10.1007 / S00259-019-04450-0.

[0198] 21. Le Roux PY, Le Pennec R, Salaun PY, Zuckier LS. Scintigraphic Diagnosis of Acute Pulmonary Embolism: From Basics to Best Practices. Seminars in nuclear medicine. 2023;53:743-51 . doi : 10.1053 / j.semnuclmed.2023.04.002.

[0199] 22. Blanc-Beguin F, Elies P, Robin P, Tripier R, Kervarec N, Lemarie CA, et al. (68)Ga-Labelled Carbon Nanoparticles for Ventilation PET / CT Imaging: Physical Properties Study and Comparison with Technegas(R). Molecular imaging and biology : MIB : the official publication of the Academy of Molecular Imaging. 2020. doi:10.1007 / s1 1307-020-01532-6.

[0200] 23. Tulchinsky M, Fotos JS, Wechalekar K, Dadparvar S. Applications of Ventilation-Perfusion Scintigraphy in Surgical Management of Chronic Obstructive Lung Disease and Cancer. Seminars in nuclear medicine. 2017;47:671 -9. doi :10.1053 / j.semnuclmed.2017.08.001 .

[0201] 24. Weber WA, Ziegler SI, Thodtmann R, Hanauske AR, Schwaiger M. Reproducibility of metabolic measurements in malignant tumors using FDG PET. Journal of nuclear medicine : official publication, Society of Nuclear Medicine. 1999;40:1771 -7.

[0202] 25. Damen EM, Muller SH, Boersma LJ, de Boer RW, Lebesque JV. Quantifying local lung perfusion and ventilation using correlated SPECT and CT data. Journal of nuclear medicine : official publication, Society of Nuclear Medicine. 1994;35:784-92.

[0203] 26. Lucia F, Bourhis D, Pinot F, Hamya M, Goasduff G, Blanc-Beguin F, et al. Prediction of Acute Radiation-Induced Lung Toxicity After Stereotactic Body Radiation Therapy Using Dose-Volume Parameters From Functional Mapping on Gallium 68 Perfusion Positron Emission Tomography / Computed Tomography. International journal of radiation oncology, biology, physics. 2024;1 18:952-62. doi : 10.1016 / j.ijrobp.2023.10.004.

[0204] 27. Lucia F, Hamya M, Pinot F, Goasduff G, Blanc-Beguin F, Bourhis D, et al. A Feasibility Study of Functional Lung Volume Preservation during Stereotactic Body Radiotherapy Guided by Gallium-(68) Perfusion PET / CT. Cancers. 2023;15. doi:10.3390 / cancers15061726. 28. Shanehsazzadeh S, Jalilian AR, Lahooti A, Geramifar P, Beiki D, Yousefnia H, et al. Preclinical Evaluation of (68)Ga-MAA from Commercial Available (99m)Tc-MAA Kit. Iranian journal of pharmaceutical research : IJPR. 2017;16:1415-23.

[0205] 29. Amor-Coarasa A, Milera A, Carvajal D, Gulec S, McGoron AJ. Lyophilized Kit for the Preparation of the PET Perfusion Agent [(68)Ga]-MAA. International journal of molecular imaging. 2014;2014:269365. doi:10.1 155 / 2014 / 269365.

[0206] 30. Canziani L, Marenco M, Cavenaghi G, Manfrinato G, Taglietti A, Girella A, et al. Chemical and Physical Characterisation of Macroaggregated Human Serum Albumin: Strength and Specificity of Bonds with (99m)Tc and (68)Ga. Molecules. 2022;27. doi:10.3390 / molecules27020404.

[0207] 31. Dickson J, Ross J, Voo S. Quantitative SPECT: the time is now. EJNMMI physics. 2019;6:4. doi : 10.1 186 / S40658-019-0241 -3.

[0208] 32. Hardcastle N, Hofman MS, Hicks RJ, Callahan J, Kron T, MacManus MP, et al. Accuracy and Utility of Deformable Image Registration in (68)Ga 4D PET / CT Assessment of Pulmonary Perfusion Changes During and After Lung Radiation Therapy. International journal of radiation oncology, biology, physics. 2015;93:196-204. doi : 10.1016 / j . ijrobp.2015.05.01 1 .

[0209] 33. Le Roux PY, Le Pennec R, Bourbonne V, Blanc-Beguin F, Pavoine M, Kerleguer K, Mauguen M, Pradier O, Salaun PY, Lucia F, Bourhis D. Pulmonary uptake value (PUV): a new quantification method for lung PET / CT imaging. EJNMMI Res. 2025 Jul 1 ;15(1 ):79. doi: 10.1 186 / s13550-025-01274-y. PMID: 40591 1 15; PMCID: PMC12214062.

Claims

CLAIMS1 . A computer-implemented method for determining a PUV value for a region of interest on a lung scan, the lung scan being a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan image, said method comprising:(a) aligning an anatomical image and a functional image such that both images have corresponding coordinates;(b) determining a lung anatomical contour on the anatomical image and determining the Anatomical Lung Volume (ALV) of the region defined by the said lung anatomical contour;(c) transferring the lung anatomical contour on the functional image;(d) obtaining the total radioactivity (Ao) inside the region defined by the lung anatomical volume (ALV) determined in (b);(e) determining a region of interest (ROI) within the ALV and obtaining the concentration of radioactivity inside the ROI (CROI);(f) determining a PUV value for the region of interest, according to the formula:PUVROI = (CROI x ALV) / Ao2. The method of claim 1 , which is for determining an indicia termed PUVRO for a region of interest on a lung scan, the indicia enabling quantifying the intensity of uptake of a tracer in the region of interest, when emission tomography coupled with computed tomography (PET / CT) or scintigraphy coupled with computed tomography (SPECT / CT) is carried out on a patient.

3. The method according to claim 1 or claim 2, wherein the method is reiterated for at least one further region of interest.

4. The method according to any one of claims 1 to 3, which further comprises a step of:- evaluating perfusional and / or ventilatory lung function on the basis of a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan image, and / or- quantifying the intensity of uptake in pulmonary ventilation perfusion scintigraphy coupled with computed tomography (SPECT / CT) or lung positron emission tomography coupled with computed tomography (PET / CT) on the basis of a lung scanthat is a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan image.

5. The method according to any one of claims 1 to 4, wherein the method is carried out at a first point in time and then reiterated at least a second point in time for a same patient that is imaged over time, the at least second point in time being separate from the first point in time and at a later time with respect to the first point in time.

6. The method according to any one of claims 1 to 5, wherein the functional contour of a region of interest (ROI) is obtained by an automatic contouring method.

7. A method for monitoring the condition of a patient over time, comprising: a. Carrying out a method according to any one of claims 1 to 6 for determining a PUV value for a region of interest on a lung scan that is a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan image of a patient at a first point in time, and b. Carrying out a method according to any one of claims 1 to 6 for determining a PUV value for a region of interest on a lung scan that is a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan image of the same patient at a subsequent point in time, and c. Optionally, reiterating step b. at another subsequent point in time.

8. The method according to claim 7, wherein the patient is treated with thoracic radiotherapy, or diagnosed or treated for Chronic Obstructive Pulmonary Disease (COPD), or diagnosed or treated for emphysema, in particular treated with endo-bronchial valve or rheoplasty, or is a lung cancer patient, in particular which can be treated by surgery.

9. A data processing apparatus configured for carrying out the method of any one of claims 1 to 8, in particular when provided with a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan.

10. A computer program product comprising software code adapted to cause a computer to perform a method according to any one of claims 1 to 8, in particular when provided witha lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan.11 . A non-transitory computer readable medium having stored thereon the computer program product of claim 10.

12. A device for determining a PUV value for a region of interest on a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan, said device comprising module(s) adapted to carry out steps (a) to (f) of method claim 1 or comprising the data processing apparatus of claim 9.

13. A medical image apparatus comprising an anatomical image device, a functional image device and the device of claim 12 for determining a PUV value for a region of interest on a lung positron emission tomography coupled with computed tomography (PET / CT) or a scintigraphy coupled with computed tomography (SPECT / CT) lung scan.

Images