Method and device for measurement and quantification of objects
Technical field
The present invention relates to a method for the reproducible measurement and quantification of intracellular objects, for example organelles such as nuclei, mitochondria, etc., or other small objects such as lipid droplets or vesicles, which involves the following stages: staining the cells with an appropriate dye, photographing the cells with enlargement, digitalisation of the photographic images and analysis of the said image with a computer program.
The inven-tion also relates to an arrangement for the reproducible measurement and quantification of intracellular, stainable objects including organelles such as nuclei, mitochondria/ etc., or other small objects such as lipid droplets or vesicles, in conjunction with which the arrangement comprises means for analysing stained, enlarged and photographed cells, images of which have been digitalized.
Prior art
Morphology is the study of the form and structure of living organisms. As biomedical research becomes increasingly morphological with new techniques for the visualisation of small processes, increasingly high requirements are being imposed on the precise analysis of morphological data. A considerable need exists for standardized quantification and image analysis.
The majority of biological and medical publications today contain histological images as a part of the argumentation. These images are of little evidentiary value, because it is difficult to measure the changes that it is wished to demonstrate. It is often necessary to accept representative images instead of images which show the average. The researcher is obliged to conduct the argumentation by other means and only
to reinforce the argumentation with images. When the publication is reviewed, images of histological preparations are often met with scepticism because the researcher himself has selected the images. It is also often impossible to reproduce the images. In research into cardiovascular diseases and cancer, studies of intracellular objects such as lipid droplets and vesicles have been extremely important (Unger RH, Lipotoxic diseases, Annu Rev Med. 2002;53 :319-36. Hammar SP, Metastatic adenocarcinoma of unknown primary origin, Hum Pathol. 1998 Dec;29 (12) : 1393-402), in conjunction with which a major need exists for methods for studying these.
US 5991028 A describes a system for cell classification according to which the cells are stained and photographed, after which the photographed image is digitalized. The digitalized image is then studied by classifying every pixel according to its appearance on its spectra. These are then compared with the appearances of the reference spectra, when the cell in its entirety is classified. The system is used primarily for the purpose of detecting and categorizing different cell types, and the system calls for a large library of stored reference spectra in order to be able to determine the condition of the cell.
WO 03091729 Al describes a graphical method for investigating diseased cells, in which the investigation involves staining, photographing and digitalizing the cells in accordance with the same method described in US 5991028 A above. A resulting colour spectrum is then analysed and compared with previously stored spectra to determine the condition of the cell. This method also calls for a large library of stored reference spectra in order to enable the condition of the cell to be determined.
Indicated in us 3919530 A is a system for analysing cells based on colour spectra in a blood sample. The analysis is performed by scanning a blood sample and filtering the result in
order to obtain a quantified colour spectrum. The result of this colour spectrum is converted into electronic data, which are then processed mathematically.
A method which provides the possibility of measuring and quantifying intracellular objects in an accurate and reproducible fashion is not shown in any of the above documents. These can be organelles such as nuclei, mitochondria, etc., or smaller objects such as lipid droplets and vesicles. The objects must be capable of being distinguished by their colour, which differs from that of the surroundings.
Description of the invention
The object of the present invention is thus to make available a method and an arrangement which solve the above- mentioned problems. The aforementioned object is achieved by means of a method in accordance with the present invention, which is characterized essentially in that the analysis stage comprises the following steps: analysis of the colours on the pixels of the digital image and the arrangement of pixels whose wavelengths have values which lie between certain predetermined upper and lower limit values, with adjacent pixels whose wavelengths also have values which lie between these predetermined upper and lower limit values, forming specific objects from the aforementioned combined pixels, and registering the number of the aforementioned specific objects.
The object is also achieved by means of an arrangement which is characterized essentially in that the arrangement comprises a computer program with means of analysing the pixels of the digital image and means for combining together pixels whose wavelengths have values which lie between certain predetermined upper and lower limit values and adjacent pixels whose wavelengths also have values which lie between these
predetermined upper and lower limit values, and for registering the number of the aforementioned specific objects.
The preferred illustrative embodiments are described in the dependent Claims .
Brief description, of the Figures
The invention is described below in a non-restrictive fashion and for illustrative reasons with reference to the accompanying drawings, in which Fig. 1 shows a digital photograph of a cell;
Fig. 2 shows the photograph in Figure 1 with circles around the cell's objects.
Fig. 3 shows a digital representation of the objects cut from the photograph in Figure 1. Fig. 4 shows the number and size of the intracellular objects.
Fig. 5 shows the size of each group of objects in relation to the combined size of the objects.
Fig. 6 shows the number and the volume of the intracellular objects.
Detailed description of the invention
A procedure for the quantification of lipid droplets in fat cells in accordance with a preferred embodiment of the present invention is described initially with reference to the Figures. The information that it is wished to obtain is: the number of lipid droplets in cells, the size of the lipid droplets that are found and the combined area of the lipid droplets that are found. A fat cell with lipid droplets is cultured on a cover glass and is stained with haematoxylin and Oil Red O in order to visualize the cell nucleus and lipid droplets. The cover glass is then mounted on a microscope slide, and the cells are
photographed in a microscope at a resolution which clearly represents the intracellular objects to be quantified. The images are digitalized and saved in JPEG format, for example (Figure 1) . The image in accordance with the invention is then analysed; all points/pixels on the image are scanned, and those which are judged to be present inside a lipid droplet 1 are identified on the original image with the help of global and local colour contrasts. The position of these points (where they are present in the image) is then investigated in order to determine which of the selected points are situated inside the same lipid droplet 1. These points are assembled, and this results in the identification of coherent structures with points that are judged to lie inside a lipid droplet 1 - that is to say the lipid droplet. Each of the lipid droplets is then given its own circle 2 (shown in Figure 2) in order to judge the specificity / selectivity. The lipid droplets are also "cut" out (Figure 3) in order further to demonstrate that something other than a lipid droplet has not been found.
In conjunction with photographing, a microscopic ruler is also photographed for comparison with the actual image. This gives a value for the resolution (pixels/μm) .
The lipid droplets that are found are then categorized according to their size (how many points they contain) and are compared with the resolution in order to obtain a value for the actual size of the lipid droplet, measured in terms of its diameter in ym.
Data relating to the number of lipid droplets of different sizes that are present in the image are then plotted (Figure 4) . A further graph in respect of the number of points present in lipid droplets of a certain size in relation to the total number of points is shown in Figure 5. The latter gives an impression of what lipid drops are most characteristic of the cell.
Figure 4 shows that the greatest quantity of lipid droplets lies in the size category of 0.2-0.8 μm in diameter. It can be appreciated from Figure 5, however, that the most characteristic category of droplets (which exhibits the largest total area of droplets) is in the size range 1.9-2.1 μm in diameter.
Which points/pixels are judged to belong to a lipid droplet is influenced by the fact that the threshold value that must be exceeded in order for a point to be regarded as belonging to an object can be influenced manually and is dependent on the dye that is used. Oil Red 0 was used in the example, which stains neutral lipid red, in conjunction with which the threshold value that was set required red/ (green+blue) to be greater than 0.7 in order for a point to be judged to be part of a lipid droplet.
By treating a sample to be investigated with different dyes, the objects contained in the sample that it is wished to measure and quantify can assume different colours. By determining a number of different threshold values, it is thus possible to measure and quantify a number of substances that are present in the sample. All the objects that lie between the upper and the lower threshold values of 0.5 and 0.6 may belong to substance A, for example, the objects that lie between 0.6 and 0.7 may belong to substance B, for example, and the objects that lie between 0.7 and 0.8 may belong to substance C, for example. The investigations can be made more efficient in this way, and a holistic image of the various intracellular objects can be provided.
The reason for using computerized analysis of images in digital format is to make an exact and reproducible qualitative and quantitative analysis. It is possible in this way to obtain scientific evidence solely from digital images of histological preparations.
The example above relates only to 2-dimensional analysis, although 3-dimensional analysis can also be carried out, in conjunction with which the number of objects and also their volume can be determined. The procedure in this case is similar to that referred to above, although the photography takes place in a confocal microscope. A 3-dimensional image is obtained here, which is then digitalized, and all the elements of the image are scanned and those which are judged to be situated inside an object are identified on the original image with the help of global and local colour contrasts. The position of these elements (their location in the image) is then studied in order to determine which of the selected elements are situated inside the same object. These elements are paired up, and the result is that coherent structures with elements that are judged to be within a specific object are identified.
By studying a sample for a given time, it is also possible to measure and analyse the growth, aggregation and movements of the objects - so-called 4-dimensional analysis. In this case, too, photographing also takes place in a "spinning disc confocal microscope" but is updated at regular intervals, for example 30 seconds, between the photographs. The result of such a study is reported in Figure 6. In conjunction with the study, the mean number of drops per unit of time was measured at 419, the number of aggregations at 91, and the mean mobility of droplets of less than 50 pixels at 1 μm/s.
A user of a computer program based on the invention is able to select for himself/herself the threshold values and the digitalisation accuracy of the image - that is to say the size of the pixels, the time interval between the photographs and the magnification by the microscope.
The invention is particularly interesting when one or more of the following points are affected:
1. Analysis of images of preparations that have heen treated with colour-characteristic markers.
2. Identification, quantification and measurement of coherent intercellular structures according to their colour and expression in the image.
3. Identification of structures as ixnimmohistochemical markers, cell organelles, lipid droplets and vesicles.
The invention is naturally not restricted to the embodiments described above and illustrated in the accompanying drawings. Modifications are possible, in particular with regard to the nature of the various parts, or by the use of equivalent technology, but without for that reason departing from the area of protection afforded to the invention, as defined in the Patent Claims.