"IN VITRO" DIAGNOSTIC METHOD FOR DISEASES AFFECTING HUMAN OR ANIMAL TISSUES DESCRIPTION
The present invention relates to an "in vitro" diagnostic method for diaseases affecting human or animal tissues, in particular for the diagnosis of diseases by means of the metric quantification of a pathologic or pathogenic material spot or aggregate in an examined animal or human tissue. The following description exemplifies the diagnosis of abnormal morphological conditions in human or animal beings by examination of a bioptic sample of liver containing collagen, wherein the collagen spots are quantified by the method of the invention. However, it is clear that the same method can be applied to the diagnosis of several conditions wherein the quantification of an abnormal material spot or aggregate is critical for making the diagnosis and evaluating the morbidity or for determining the advance or regress of the abnormal condition. With the term "abnormal condition" it is intended a pathological condition or a condition which gives rise to a pre- or post- pathological situation and for which an abnormal morphometry can be recognised. Such abnormal conditions may be, for example, oncological diseases,
atherosclerotic plaques, edemas, hematomes, acute or chronic inflammatory lesions, scars and collagen diseases .
The method of the invention can thus be applied to tissues different from hepatic tissues and starting from tissue samples obtained in any manner.
Coming to the specific example hereinafter described, it is known that many hepatic diseases are currently evaluated bioptically using typically sophisticated qualitative methods. However, there are still uncertainties in estimating the evolution of the chronic hepatic process . A key element has proven to be the evaluation of fibrosis, i.e. the lesion commonly observable in the histological pattern of chronic hepatitis. The noticeable presence of collagen, in the shape of highly irregular scars spread in the portal spaces, is a product of the portal inflammation due to necrotic inflammation centers . The newly formed collagen grows into either thin fibers or variously thick sets that dry out the parenchyma. In more advanced stages of the disease, the fiber sets form portal-portal and central-portal links.
The main source of the above named uncertainties in the evaluation of the hepatic disease stage arises from the methods used to assess the fibrosis present in the
hepatic tissue. Semi-quantitative methods, which are the most widely used, indicate categories of disease severity, but do not provide metrical measurements.
In addition, close to the larger collagen' s fragments, we were able to detect, by means of a computer-assisted optical microscope recognition of specifically stained connectival tissues (Sirius Red or other specific stains) , a pletora of extremely small and highly indented fragments, which are otherwise invisible to routine observation. Evaluation of such minor fragments, which has not been made by the diagnostic methods used up to now, is of pivotal importance since the presence of such fragments is an index of a dynamic evolution of the pathology. In fact, the initial three-dimensional configuration of the collagen structure is a dispersed set of small collagen islets that evolve with the disease into a spongy mass (fibrosis) due to the splicing of distal ends of the growing collagen fibers fuelled by the chronic inflammatory process. In two- dimensional histological liver slices, this spongy mass appears as very wrinkled collagen areas irregularly distributed in the tissue.
Morphometrical methods have proved to be unsuitable for measuring the irregular shapes of fibrosis because
of the fact that Euclidean geometry can not be applied to such shapes . Euclidean geometry is conversely apt for measuring points, regular lines, planes and volumetric bodies whose dimensions are respectively expressed by means of integers 0, 1, 2 and 3 and whose shape does not change upon optical magnification.
The microscopic observation of either a normal or abnormal, component of tissue samples taken from any organ, particularly liver, is amazing because of the new irregularities that appear at any magnification (scale of observation) . As the extension form of the image of the samples changes, the new irregular details are given measures and dimensions that are independent at each magnification and can not be arranged in a single linear system. Because of this characteristic, which is due to the scabrousness of the external surface of the object to be observed, the visible details, as well as those that can not be visually identified, make hepatic tissue samples (like all objects with an irregular surface) hardly measurable by means of traditional computer-aided morphometry.
The difficulties encountered in metrically measuring the shapes of the collagen present in the bioptic sample of an hepatic tissue depend on: - the irregularities of the outlines that do not
allow collimations with the smooth shape of the linear method;
- the modification of the shape of these objects at every scale of observation, because of newly appeared details that can not be observed at the previous magnification;
- the dimensional change of the space occupied by the sample at every magnification, as a function of the change in shape; - the multiplicity of the perimetric lenghts and surface areas, whose dimensions scale with the resolution of the measure (the smaller the meter, the higher the measure) .
The classical morphometry tackles the problem of measuring natural objects by approximating their irregular outlines and rough surfaces to rectilinear outlines and plane surfaces. In addition, there is the well known non-representative nature of a bioptic sample as its small volume makes the so-called disease staging hardly indicative because of the unevenness of the distribution of lesions in the organ as a whole. It is known that only a slight difference in the site from which a bioptic fragment is taken is often sufficient to obtain a sample that indicates a different stage from the one of the adjacent tissue.
The purpose of the present invention is therefore to provide a diagnostic method that, although starting from a tissue sample taken in accordance with the state of the art, allows a complete and precise diagnosis of the patient's current pathological status and its evolution.
According to the present invention, this object is achieved by means of an in vitro diagnostic method whose characteristics are specified in the main claim. Further characteristics of the method of the present invention are specified in the subsequent claims.
Irregular objects were defined "fractal" by Benoit Mandelbrot since, in spite of the fact that their shape changes as a function of magnification, they retain the features of their irregularity at all spatial scales . Although the pieces (not fractions) into which they can be divided are not equal, they preserve the similitude of their irregularity. This property of the parts into which irregular objects can be divided is called "self- similarity" . Since the shape of such objects depends on the magnification at which their image is observed, any quantitative metering of the dimensions of the object is a function of the magnification scale. The fractal dimension indicates therefore the "self-similarity" of the fractal pieces of an irregular body and, at each
scale, defines the characteristics of the reference means used to measure the physical and geometrical parameters of the observed irregular object.
The present invention is based on the intuition of the inventors concerning the metric quantification of the tissue spot, as well as any irregular object, itself.
The inventors have also surprisingly found that the "rugosity" , a specific characteristic of the surface of the collagenic structures present in the hepatic tissue, can be metrically quantified. As a matter of fact the inventors have developed an algorithm that makes it possible to evaluate the said "rugosity" by means of measurements of the true perimeter and area of the collagenic structures present in the hepatic tissue. The diagnostic method according to the present invention provides for the use of two apparatuses that are well known in the diagnostic technology, namely a microscope with a motorized stage and a computer. The microscope is used to examine visually the bioptic sample, while a specific software is employed to capture and convert the optical image to a digital image that allows the further measurements and subsequent calculations to be performed. The bioptic sample is taken by means of
conventional bioptic methods and is used to prepare, in a known manner, a slide in which the collagen present in the hepatic tissue is hystochemically stained or immunohystochemically labelled. An operational stratagem of the method according to the present invention is to stain the hepatic collagen present in the sample on the slide by using a standardised stain for which the intervals of the three primary colours (red, green and blue) are well known. The preferred stain is Sirius Red, also known as Direct Red 80, which has a threshold for each primary colour that varies between a minimum value of 0 and a maximum value of 255 intensity units (24 bit depth, 16 million colours BITMAP , image analysis) . Each of appropriately stained biological structures is characterised by a specific staining interval . In the case of Sirius Red-stained hepatic collagen, the thresholds are 0-255 intensity units for red (R) , 0-130 intensity units for green (G) and 0-255 intensity units for blue (B) . The selection of collagen present in the tissue on the slide can be automatically performed by the computer once the operator has set the three specific thresholds for the three primary colours.
After histochemical staining or immunohistochemical labelling, the slide with the bioptic sample or one of
its parts is placed on the motorized stage of a microscope connected to a computer through a tele/photo camera. The apparatus that can be used in the method of the present invention is the one described in the International application entitled METHOD AND APPARATUS FOR ANALYIZING HUMAN OR ANIMAL TISSUE SPECIMENS filed by the same Applicant and having the same filing date of the present application, whose description is herewith incorporated by reference . The movement of the motorized stage along the two main orthogonal axes x-y is automatically controlled by a specific software program. The whole image of the histological preparation is automatically reconstructed by the computer and recorded in the memory thereof as an image file.
Focusing of the image is also automatically performed.
The image file is then processed by the computer, that selects the parts of the tissue on the slide under examination that fall within the predetermined intervals for Sirius Red and therefore correspond to the collagen. By this operation, ' the collagenic structures present on the slide are selected from the image file and their perimeter and areas are exactly reproduced. The next step is the identification and calculation
of the area occupied by the histological preparation as a whole and the area A occupied only by the collagenic structure under examination. The unit of measurement may be μm2 or pixel, taking into account that 1,9 pixel side = 1 μm (i.e. 1 pixel side = 0.526 μm) at a 200x magnification and a videocamera resolution of 1.3 Megapixels. The area A of the structure under examination can be expressed in absolute terms or as a percentage of the total area of the sample under investigation. All measurements of the collagenic I structure can be automatically made by the computer.
The perimeter P of the selected collagenic structure is likewise, and almost simultaneously, ■ identified and calculated, according to a known computer-aided algorithm, and can also be measured in pixel or μm.
Given the considerable irregularity of the perimeter of the selected collagenic structure and in order to be able to meter it in concrete terms, an
I
I evaluation of its fractal dimension DP is made. Similarly, the estimate of the fractal dimension of the area of the selected collagenic structure is indicated! by the symbol DA. Both of these fractal dimensions can be automatically determined using the known "box- ' counting" algorithm.
According to the "box-counting" method, the image is divided into a grid and the fractal dimension D is given by the mathematical formula D=lim(ε->0) [logN(ε)/log(l/ε)l wherein ε is the length of the side of the boxes of the grid and N(ε) is the number of boxes necessary to completely cover the outline (DP) or the area (DA) , respectively, .of the measured object. The length ε is expressed in pixel or μm and, in the present calculation method, ε tends to 1 pixel. I
In order to avoid difficulties in such a calculation, the fractal dimensions DP and DA are approximated as the slope of the straight line obtained by putting in a Cartesian axis system the parameters logN(ε) versus log(l/ε) .
In practice, the method used to determine DP comprises : a) dividing the image of the object into a plurality of grids of boxes having a side length ε, in
! which ε varies from a first value substantially corresponding to the side of the box in which said object is inscribed and a predefined value which is a fraction of said first value, b) calculating a value of a logarithmic function of N(ε), in which N(ε) is the number of boxes
necessary to completely cover the perimeter (P) of the object and of a logarithmic function of l/ε for each ε value of step a) , thus obtaining a first set of values for said logarithmic function of N(ε) and a second set of values for said logarithmic function of l/ε, c) calculating the fractal dimension D
P as the slope of the straight line interpolating said first set of values versus said second set of values of step b) . The same method is applied for calculating the fractal dimension D
A, with the only difference that, in this case, N(ε) is the number of boxes of side ε that completely cover the area of the object to be quantified. From these calculations and applying the fractal geometry's principles, it derives that
wherein P
F is the fractal-corrected perimeter, P is the Euclidean perimeter, D
P is the fractal dimension, D
I is the Euclidean dimension (1) and λP is the dilation coefficient. The value of λP is empirically determined using a histological section acquired at different magnifications (5x, lOx, 20x, 40x objective magnification) and then observing new emerging details of the object under evaluation. The λP is found to be
approximately 4.5.
Analogously, Af, i.e. the corrected area of the irregular object to be observed, is given by the formula
Af = A+[λA(DA-D)] . (Ap - A) (la) wherein A is the Euclidean area, D is the Euclidean dimension (2) , λA is the dilation coefficient which was found to be approximately 0.1, Ap the area of the region including the objects to be quantified and DA is the fractal dimension of the area which is calculated by means of the box-counting method. I
With the term "region including the objects to be quantified" it is intended the region of the image in which it is possible to detect objects, even of small magnitude, belonging to the same morphological item. It is in fact known that in some cases the object to be quantified is composed of a plurality of objects (spots or aggregates) of different magnitude, some of them being non detectable under visible analysis. This algorithm allows to take into consideration the overall
I area of the item under observation and in particular, if applied in the collagen analysis, it is possible to determine not only the area of the larger collagen spot, but also the area of the smallest islets .
It is clear that evaluation of the perimeter or, more particularly, of the area of the observed object
can give a first diagnostic indication. In the example herein described, the evaluation in a patient of the area of the collagen spot in samples taken at different times is indicative of the progression or regression of the disease. In other cases, such as radiography analysis, the extent of a bone fracture or of a tumor mass can be precisely determined.
After having calculated the above values, the aforesaid fundamental parameter of "rugosity" is determined. The inventors have found in fact that the parameter w indicating the degree of "rugosity" of the selected collagenic structure can be calculated by means of the following algorithm: w = P /fX 2 j πAf? — .π R (III) wherein Pf is the rectified perimeter (fractal perimeter) , Af is the fractal corrected area of the collagenic structure and R is the "roundness coefficient" of the collagen islets. R is on its turn calculated with the following algorithm
wherein Pe is the perimeter of the ellipse in which the measured object is inscribed and Ae its area.
Finally, once calculated the rugosity w of the collagenic structure present on the slide, the status of the structure and therefore the so-called staging of the
hepatic pathology as a further confirmation. It has been found in fact that the value of rugosity w is associated with the stage of the hepatic pathology.
The diagnostic method according to the present invention can be further implemented with the determination of the distribution of collagen in the bioptic sample. This distribution is determined by subdividing the whole sample using a grid with a 200 μm squared mesh and by indicating the number n of the squares in the grid by I the symbols from i to An. The determination of the grid follows the geostatic rules for spatial samplings.
The local collagenic area in each square is measured and a calculation is made of the partial sums of the collagenic areas contained in the sequence Ax, A, A3/ ...An. Each partial result (Ax, Aχ+A2, Aι+A2+A3, A1+A2+A3+A, etc.) is reported in a Cartesian system in which the abscissae express An and the ordinates the quantity an of collagen contained in An. This collagen distribution parameter, ' which takes into account all collagen islets, thus gives a fundamental diagnostic information, since it is correlated with the evolution of the disease.
Another complement of the diagnostic method according to the present invention is the determination
of the internal tectonics of each slide-selected collagenic structure by evaluating the degree of RGB colour scale heterogeneity of the set of pixels making up each area. This value, which indicates a densitometric heterogeneity, can be automatically calculated by the computer and expressed by a dimensioned numerical value indicated by the letter I . The value of I corresponds to the percent of pixels that differ more than 5% from their mean value. All the above mentioned parameters can be calculated automatically by the computer implemented with a software and executed in a short time.
In a particular embodiment of the present invention, the "in vitro" diagnostic method is further implemented by taking into account the fact that the quantitative morphological measures of the components of the biopsy sample may be influenced not only by their irregularity, but also by the artefacts created by the squeezing, tearing and twisting occuring during surgical
I
I excision and histological manipulations. The correction factor CF can thus be calculated by the following formula |
CF=jm/L wherein jm is the square root of the mean squared area of normal hepatocytes , and L is the square root of the mean squared area
of the hepatocytes of the sample under observation.
The parameter jm, which represents the side length of the square having the same area of a standard hepatocyte, was calculated over a large number (about 3,000) of samples belonging to healthy subjects and can be approximated to 16 μm.
The fractal perimeter Pf further corrected by such a correction factor can thus be obtained by the formula
Pcor = Pf.CF (VI) Analogously, the corrected area is given by
Acor = Af.CF2 (VII)
The above calculated coefficient of wrinkledness w can also be corrected by substituting Pf with Pcor and Ac with Acor in the above formula (III) . From what has been said above, it is clear that the diagnostic method of the invention represents an improvement if compared with the known methods. The fractal geometry offers mathematical models derived from the infinitesimal calculus that, when applied to Euclidean geometry, integrate the figures of the morphometrical measurements of natural and irregular objects, thus making them closer to the actual values.
The diagnostic method according to the present invention ' has the advantage to eliminate the inconveniencies of all the methods so far used for
examining bioptic samples, in particular of hepatic tissues .
It is however clear, as said before, that the method of the invention can be applied to all the cases in which the quantification of an abnormal material spot or aggregate, i.e. the determination of its magnitude and related parameters, is required. In such cases, any reference made above to a collagen structure can be similarly applied to a different structure such as a cell i aggregate (like in tumors) or a colesterol plaque, without departing from the method of the invention.
Analogously, any reference made above on a computer-aided, automatic determination of the magnitude and dimension of the observed object should be understood as a particular embodiment of the invention. It is clear in fact that the same operations can be performed manually, for example by reporting the image on a sheet, subdividing the image in a grid and applying the known algorithms in a manual method. The method of the invention can also be applied, as said before, not only to histological images of the human or animal body or parts thereof, but also to radiography images, ecography images, Computerized Axial Tomography (CAT) , magnetic resonance (NMR) or Positron Emission Tomography (PET) . In such cases use of the
microscope will not be necessary, since the image can be directly digitalised by a videocamera and acquired by the computer software. Of course, in this case too a manual performance of the method is possible. - Further advantages and characteristics of the procedure according to the present invention will be evident to those skilled in the art from the following operative examples and the attached figures 1, 2 and 3 that show collagen spatial distribution in three ' patients' samples. EXAMPLE
A standard bioptic sample was taken from three different patients with chronic HCV-related disease. The approximately 10 μm long sample was set in 10% formalin and embedded in Paraplast. After the Paraplast was removed, 5 μm thick sections were cut and stained with Sirius Red. The slides were microscopically observed at a 200x magnification using an image analysis system and all the images were digitalised. ' The portion of liver subject to fibrosis was automatically selected on the basis of the similarities
I of colours of adjacent pixels. The images were then converted into 1-bit (black and white) binary images.
The tolerance thereshold was adjusted in such a way as to select all the fibrotic lesions.
The individual pixel boundaries of the perimeter and surface area of the fibrotic portion were automatically traced using the known "box-counting" algorithm and their fractal dimension was determined. The fractal dimension was automatically measured using the "box-counting" method. The morphometrical values (A, P, Af, Pf) were then determined and, on their basis, the computer calculated the degree of rugosity w. The computer also calculated the spatial distribution of the collagen (see figures 1-3) as well as the values of H and I by using a computer-assisted image analysis system.
The collected data are shown in the following table which lists the various morphometric values:
Table 1
The table shows that the rugosity value for the first patient was 38.56, which corresponds to an initial stage of hepatic pathology.
In the case of the second patient, the quantified rugosity of collagen was 1082.58, which indicates an intermediate stage of hepatic fibrosis .
The calculated rugosity for the third patient was 62421.72, thus indicating the presence of cirrhosis.
The graph in figure 1, which relates to the first patient, shows a stepped state that confirms the initial stage, of the disease.
The graph in figure 2 (second patient) shows that the spatial distribution of collagen has fewer steps and therefore confirms an intermediate stage of the disease.
The graph in figure 3 (third patient) has an almost linear trend that confirms an advanced disease stage.
The spatial distribution of collagen in a healthy liver is along a curve consisting of rather regular steps.
In general, in the case of collagen determination
in hepatic diseases, the evaluation of the disease stage can also be effected by determining the coefficient of rugosity w of the patient and comparing it with predefined values, wherein w values below the predefined threshold value are indicative of the stage of the pathology.
Said predefined threshold values are determined by statistical analysis of the test results collected from a statistically significant patient population, wherein liver samples from the patients have been subjected to observation and parameter determinations according to the invention method.
The above description of the diagnostic method of the invention, as well as the data of the operative example shown above in the table and figures, refer to the "in vitro" diagnosis of hepatic pathologies by means of the metric quantification of collagen. It is obvious that the same metric quantification of collagen can be used to diagnose other pathologies staged as a function of the bi-univocal presence of collagen in bioptic samples .