Design of a walk-in lab test for lung morphometry characterization

Abstract

The present invention relates to a method for estimating lung morphometry based on aerosol deposition characteristics using an imaging means such as a gamma camera to scan the lungs. An adaptive image threshold technique is used to determine the ratio of deposition in central to peripheral region of the lung (C/P ratio). The morphometric parameters such as length and diameter of distal lung airways (P.sub.8 and P.sub.9 respectively) and mean alveolar diameter (d.sub.alv) are determined from aerosol retention data and clearance data.

Claims

1. A method for estimating lung morphometry comprising of: delivering radio-aerosol in a subject; generating a plurality of images of the lung using an imaging means; determining ratio of aerosol deposition in the Central to Peripheral region of the lung (C/P ratio) by drawing Region of Interest (ROI) using adaptive image threshold technique; and determining the morphometric parameters through an optimization procedure comprising minimizing an error function.

2. The method for estimating lung morphometry according to claim 1, wherein said radio-aerosol is delivered to the lungs using any oral or nasal compliance.

3. The method for estimating lung morphometry according to claim 1, wherein the imaging means is a camera selected from a gamma imaging camera, CT scan and SPECT or PET scan.

4. The method for estimating lung morphometry according to claim 1, wherein the aerosol deposition characteristics include retention data and clearance data of aerosol in lungs.

5. The method for estimating lung morphometry according to claim 4, wherein the retention data and clearance data of aerosol in lungs include time, volume, or percentage of aerosol in lungs.

6. The method for estimating lung morphometry according to claim 1, wherein the adaptive image threshold technique is used to draw Region of Interest (ROI) to calculate C/P ratio, and to identify the lung boundary.

7. The method for estimating lung morphometry according to claim 1, wherein the morphometric parameters include length and diameter of the distal lung airways as well as associated parameter (P.sub.8 and P.sub.9 respectively), and mean alveolar diameter (d.sub.alv).

8. The method for estimating lung morphometry according to claim 1, wherein the method for drawing ROI using adaptive threshold technique comprises of: i) converting the initial image obtained from imaging at t=0 hours and after upto twenty four hours post-inhalation of the aerosol, to a gray scale image; ii) cropping the image to make the right lung as the only focus and cropping dimensions which are to be the same for all individuals; iii) calculating the lung boundary of the static image by extracting the boundaries of the connected areas at a given threshold; iv) increasing the image threshold value to obtain the maximum image boundary; v) obtaining the total right lung area by superimposing said lung boundary over the first image obtained immediately after the aerosol inhalation; vi) calculating the C/P ratio from the ratios of the area between the central and peripheral regions; and vii) shrinking said total right lung area to create an area ratio obtained in step (vi) using an iterative procedure.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The summary of the present invention, as well as the detailed description, are better understood when read in conjunction with the accompanying drawings that illustrate one or more possible embodiments of the present invention, of which:

[0016] FIG. 1 illustrates the images obtained using gamma camera at time intervals of 0, 1, 2, and 4 hours;

[0017] FIG. 2 illustrates the plot of gray scale level and pixel area for the subjects under study;

[0018] FIG. 3A and FIG. 3B illustrate the morphology features of the human lung;

[0019] FIG. 3A illustrates the diameter of the airway predicted at every generation for the sample subjects using the proposed method;

[0020] FIG. 3B illustrates the length of the airway predicted at every generation for the sample subjects using the proposed method;

[0021] FIG. 4 illustrates the TLC predicted for generations 0<z<17 and 17<z<23 for sample subjects; and

[0022] FIG. 5 illustrates the Flow chart/protocol for the Walk-in lab test.

DETAILED DESCRIPTION OF THE INVENTION

[0023] The present invention is related to a method for estimating lung morphometry including the length and diameter of distal lung airways, mean alveolar diameter and lung boundary based on the aerosol deposition characteristics obtained by an imaging means. According to the embodiments of the present invention, the imaging means may be for example, a gamma camera.

[0024] According to an embodiment of the present invention, the proposed method can create a morphology map as well as any required regional deposition map for any given subject.

[0025] According to an embodiment of the present invention, the aerosol used for measuring deposition characteristics includes but is not limited to 99mTechnetium phytate radio-aerosol generated using a Biodex® Venti-scan® radio-aerosol delivery system.

[0026] According to an embodiment of the present invention, the camera used for aquiring lung images includes but is not limited to gamma camera.

[0027] The radio-aerosol is delivered to the lungs using any oral or nasal compliance. As an exemplary case in the present invention, a radio-aerosol such as .sup.99mTechnetium phytate generated using a Biodex® Venti-scan® radio-aerosol delivery system is given to the subject through a mouth piece. The nose is clamped and the subject is instructed to breathe through their mouth. The radio-aerosol solution contains 20 mCi of radioactivity. The subject inhales the aerosol for about three minutes so that about 2 mCi, i.e., one-tenth of the total preparation is deposited inside the thorax.

[0028] An imaging means such as gamma scintigraphy is used to acquire lung images at time intervals (t) of 0, 4 and 22 hours. In the subsequent analysis, only the first image is used for extracting the C/P ratio in which case the original test dosage of the radio-aerosol could be further reduced.

[0029] The C/P ratio is defined as the ratio of deposition in the central to peripheral region of the lung. It responds to changes in important factors that influence the deposition characteristics in the lung such as particle size, inhaled airflow rate, and airway patency. The peripheral zone comprises mostly of small airways and alveoli. The central zone consists of larger central airways.

[0030] The conventional method of getting this ratio is to draw a region of interest (ROI) on the right lung to demarcate the central (C) and peripheral (P) regions of the lung. The time activity curves are obtained for each of the two regions. After correcting the counts for decay and background, the central to periphery ratio of counts C/P is obtained from the initial image, post-inhalation of the radio-aerosol. There are various methods available in the literature to draw these regions of interest. One of the methods is intended to find the C/P ratio of which the central region is one-third of the lung area and the remaining two-third region is considered as the peripheral lung region. The present invention implements this method for the calculation of C/P ratio.

[0031] In order to find the lung area, the first image is recorded at t=0 hour. It is important to minimize user induced variability in drawing the region of interest (ROI) based on which the central to peripheral ratio is calculated. It is proposed to use an adaptive image threshold technique in MATLAB® for finding the lung outline and thereby drawing the region of interest (ROI). Thus, the C/P ratio found using this technique is free of inter-subject variability in drawing ROI manually. The gray scale level or the image threshold in MATLAB® is automated for the subjects based on the intensity of the pixels in the image to provide the lung outline. In each subject, there is a particular threshold value which gives the maximum lung area. Increasing the image threshold above this value decreases the lung area drastically. All the other intensity threshold values below this particular threshold yield a lower lung area. This particular threshold value helps detect all the connected components of the image giving rise to the maximum lung area. The adaptive technique is used to effectively identify the lung boundary.

[0032] According to the embodiments of the present invention, the morphometric parameters for an individual subject are determined from the aerosol deposition and clearance data. The morphometric parameters considered in accordance with the present invention include length and diameter of the last generations P.sub.8 and P.sub.9 respectively, and the mean alveolar diameter (d.sub.alv). These three parameters determine the airway morphometry in the distal lung.

[0033] In order to identify the morphometric parameters, an error function (ε) is postulated as a normed difference between the numerical predictions of (C/P).sub.m ratio and experimentally determined values C/P ratio, specifically, ε=|C/P−(C/P).sub.m|. The numerical predictions of (C/P).sub.m are derived using the model proposed by Devi et al (Designing aerosol size distribution to minimize inter-subject variability of alveolar deposition, Journal of Aerosol Science, 2016). The parameters P.sub.8, P.sub.9, and d.sub.alv are determined by minimizing this error. The proposed method which uses C/P ratio is robust, convergent and predicts physically relevant values for the parameters. The total lung capacity of each subject is used as constraint in the minimization process. The corresponding lengths and diameters of the bronchioles at various generations are calculated using these parameters and Weibel's morphometry.

Examples

[0034] Study Design and Participants:

[0035] In this study, six healthy, non-smoking individuals with pulmonary function tests within the normal range and no respiratory illness were selected. In all the cases, a detailed history was considered and a thorough physical examination was carried out especially focusing on the respiratory system. In all the subjects, a chest radiograph (postero-anterior view) was also taken to ascertain normal lung structure. Lung volumes were determined using spirometry. The anthropometric and spirometry data of the subjects under study are listed in Table 1.

TABLE-US-00001 TABLE 1 Anthropometric and spirometry data FEV1 FEV1/ % FVC Sub- Age Height Weight FEV1 Pre- FEV1/ % ject (yr) (cm) (kg) (L) dicted FVC Predicted 1 26 175 75 4.43 119 0.89 106 2 38 175 86 3.45 102 0.91 111 3 25 172 81 3.78 102 0.87 95 4 39 173 66 3.32 101 0.81 99 5 24 182 87 4.47 108 0.88 98 6 41 159 65 2.68 110 0.91 108 FEV1-Forced expiratory volume in one second; FVC-Forced Vital Capacity

[0036] In the gamma scintigraphy study, the 99mTechnetium phytate radio-aerosol was generated using a Biodex® Venti-scan® radio-aerosol delivery system. The mass median aerodynamic diameter (MMAD) of aerosol produced by the nebulizer is 0.5 μm.

[0037] Aerosol Administration and Image Acquisition:

[0038] The subjects were made to sit comfortably in a chair and relax for 10 minutes. Then 99mTechnetium phytate radio-aerosol generated using a Biodex® Venti-scan® radio-aerosol delivery system was given through a mouth piece. The nose was clamped and the subjects were instructed to breathe through their mouth. The radio-aerosol solution used on all subjects contained 20 mCi of radioactivity. The subjects inhaled the aerosol for about three minutes so that about 2 mCi, i.e. one-tenth of total preparation was deposited inside the thorax. As an exemplary case, the procedure followed by Guleria et al, was followed in acquiring the images at t=0 hour, 4 hours and 22 hours. In the subsequent analysis, only the first image for extracting the C/P ratio was considered. Hence, the original test dosage of the radio-aerosol could be further reduced.

[0039] FIG. 1 of the present invention illustrates the images obtained using a gamma camera at intervals (t) of 0, 1, 2, 4 hours. The red and yellow-coloured bright spots indicate that more aerosol has been retained in those regions. The bright spots in the stomach are due to the assimilation of the aerosol at the mouth walls. The trend of clearance kinetics can be observed from the reducing intensity at increasing time intervals. However, each image is unique for every subject and identifies the importance of inter-subject variability even among a small cohort of healthy subjects. Comparing the initial image of subject 5 with subject 3 or subject 6, it appears as though subject 5 had inhaled about ¼.sup.th of aerosol compared to subject 3 and subject 6. Thus, this study has attempted to bring out the morphometric cause of this variation. The huge variation in brightness is a striking feature which emphasizes the need for personalized treatment.

[0040] Determination of C/P Ratio:

[0041] According to the embodiments of the present invention, in order to find the lung area, the first image was recorded initially at t=0 hour, i.e., immediately after inhalation. It is important to minimize user induced variability in drawing the ROI based on which the central to peripheral ratio is calculated.

[0042] The procedure to draw ROI using adaptive threshold technique comprises of:

i. converting a static image obtained from the imaging means, 2 hours post-inhalation to a gray scale image;
ii. cropping the image to make the right lung as the only focus and cropping dimensions which are to be the same for all individuals;
iii. calculating the lung boundary of the static image by extracting the boundaries of the connected areas at a given threshold;
iv. increasing the image threshold value from a low value to a high value as a particular value of threshold gives the maximum image boundary as at all the other values of threshold, the lung boundaries are not well connected;
v. obtaining the total right lung area by superimposing the lung boundary over the first image obtained immediately after the aerosol inhalation;
vi. calculating the C/P ratio from the area ratios between the central and peripheral regions; and
vii. shrinking the total right lung area to create an area ratio obtained in step (vi) using an iterative procedure.

[0043] In accordance with some of the embodiments of the present invention, the area ratio between the central and peripheral regions is 0.33. The ratio of the total pixels in both the regions corresponds to the C/P ratio.

[0044] Using the above method, the total number of pixels versus the gray scale level for each of the subjects is shown in FIG. 2. This figure is part of the intelligent threshold technique proposed for finding the C/P ratio. The gray scale level indicates the value of the pixel intensity in an image. The number of pixels in the y-axis of the figure corresponds to the right lung area. The line styles are varied to indicate subjects 1 to 6. The value marked with a triangle marker in the figure shows

the maximum right lung area and its corresponding threshold. The lung outline determined from the first image is superimposed over the same image post-inhalation and the total lung area is determined using image processing in MATLAB. This area is also divided into two regions—one-third representing the central region and two-thirds representing the peripheral region.

[0045] Determination of Subject-Specific Morphometric Markers:

[0046] In accordance with the embodiments of the present invention, the morphometric parameters for an individual subject are identified from the aerosol deposition and clearance data. From the present invention, three parameters, P.sub.8, P.sub.9 and d.sub.alv have been identified as important lung morphometry markers.

[0047] In order to identify the morphometric parameters, the error function (ε) was employed. The parameters obtained using the said approach are tabulated in Table 2. Table 2 shows the predicted parameters as well as their deviation from the Weibel values (in percentage terms). The corresponding lengths and diameters of the bronchioles at various generations calculated using these parameters and Weibel's morphometry model are also graphically depicted in FIG. 3A and FIG. 3B.

TABLE-US-00002 TABLE 2 Parameters predicted C/P ratios in accordance with present invention Deviation from Predicted Parameters Weibel Values Subject P.sub.8 P.sub.9 d.sub.alv P.sub.8 P.sub.9 d.sub.alv 1 2.59 1.33 0.0281 4 2 4 2 2.80 1.33 0.0289 12 2 6 3 2.69 1.28 0.0279 8 −2 2 4 1.56 1.79 0.0286 −38 38 5 5 1.70 1.69 0.0290 −30 30 6 6 1.84 1.50 0.0269 −6 15 −1

[0048] FIG. 3A and FIG. 3B depict the variation of bronchiole diameters and lengths at each generation in a sample size of six subjects. The average diameters and lengths of a model individual are also plotted in the same figure for reference. The numerical approach predicts that subjects 4 and 5 have larger diameter bronchioles but are shorter in length. In contrast, subjects 2 and 3 have smaller diameter bronchioles but are longer. This information is corroborated with corresponding images presented in FIG. 1. It is observed that subjects 2 and 3 have greater aerosol deposition, while subjects 4 and 5 have lower aerosol deposition. The increased wall surface area due to longer bronchioles appears to be a self-consistent explanation of this observation. The insets in FIG. 3A and FIG. 3B show the lengths and diameters of the alveolated airways only. Even though the parameters

P.sub.8 and P.sub.9 vary about ±30% from the Weibel values for subjects 4 and 5 (Table 2), the lengths and diameters are physically consistent for all the six subjects. Table 2 also presents the deviation of other functional aspects of the lung.

TABLE-US-00003 TABLE 3 Change in surface area and lung volumes in relation to Weibel values % Change in % Change in % Change in predicted predicted airway predicted alveolar Subject surface area volume volume 1  6  8 11 2 −3 17 20 3  4  4  5 4  6 18 16 5 −8 17 21 6 −6 −2 −3

[0049] Total Lung Capacity (TLC) is another physiologically relevant parameter. FIG. 4 shows the predicted TLC for the six subjects using the proposed method. The generations 0<z<17 represent the non-alveolated airways and the generations 17<z<23 represent the alveolated airways. It can be observed that the non-alveolated airway volume (0<z<17) remains almost the same for all subjects, while the alveolated airway volume (17<z<23) shows significant subject-to-subject variability in contributing to the total lung capacity. Therefore, variation in alveolar volume is a prime cause of inter-subject variability in TLC, and therefore aerosol deposition. The parameter d.sub.alv which is a measure of the mean alveolar diameter, shows only about 1% to 6% variation for all the subjects considered for the study (Table 2) and the corresponding alveolar volume varies by up to 21% (Table 3). This suggests that aerosol deposition is sufficiently sensitive to alveolar volume. The morphometric parameters associated with the later generations (17<z<23), which are immeasurable by current imaging modalities are well-predicted by the said approach of the present invention.

[0050] Table 3 shows the deviation of lung surface areas and total lung volumes for each subject from an average individual to indicate inter-subject variability. Lung surface area is an important factor influencing drug deposition. According to the embodiments of the present invention, the change in surface area for subject 5 amounts to about 8%. As the surface area for subject 5 is reduced by 8%, the drug absorption would also be reduced in this subject compared to an average individual. The percentage change in alveolar volumes for each subject is considerably greater than the volume changes of the airway passages.

[0051] According to the embodiments of the present invention, the alveolar volume, while being only about one-third of the total lung capacity, is a primary cause of inter-subject variation in morphometry. As the lung surface area and alveolar volume are fundamentally important parameters in drug delivery and absorption, it is important to estimate these parameters for an individual subject. Therefore, predicting the parameters would help the physician in personalizing the dosage and therapy of the patient. FIG. 5 represents the flow chart to indicate the protocol for the Walk-in lab test of the present invention. The protocol comprises prescription of a lung morphology test by the physician to an individual or subject, gamma scintigraphy test getting performed by the subject, obtaining P.sub.8, P.sub.9, and d.sub.alv in accordance with the embodiments of the present invention, obtaining a deposition map for the subject for any given size and distribution of the drug which is in the form of an aerosol, and prescription of appropriate customized drug dosage to the subject by the physician.

[0052] According to a preferred embodiment, the method for estimating lung morphometry comprises delivering radio-aerosol in a subject, generating a plurality of images of the lung using an imaging means, determining the ratio of aerosol deposition in the Central to Peripheral region of the lung (C/P ratio) by drawing Region of Interest (ROI) using adaptive image threshold technique, and determining morphometric parameters by postulating an error function as normed difference between the numerical predictions of C/P ratio and the experimentally measured values of the C/P ratios. The adaptive threshold technique comprises of converting the initial image at t=0 hour post-inhalation and after upto twenty four hours post-inhalation of aerosol, to a gray scale image; cropping the image to make the right lung as the only focus and cropping dimensions which are to be the same for all individuals; calculating the lung boundary of the static image by extracting the boundaries of the connected areas at a given threshold;

increasing the image threshold value; obtaining the total right lung area by superimposing the lung boundary over the first image obtained immediately after the aerosol inhalation;
calculating the C/P ratio from the area ratios between the central and peripheral regions; and shrinking the total right lung area to create above obtained area ratio using an iterative procedure.

[0053] It is to be understood, however, that the present invention would not be limited by any means to the techniques, and approaches that are not specifically described, and any change and modifications to the techniques and approaches can be made without departing from the spirit and scope described in the present invention.