COMPUTERIZED MORPHOMETRIC ASSESSMENT ... - Annals of RSCB

0 downloads 0 Views 1MB Size Report
Amanatiadis et al., 2011). In biological and medical applications, contour-based descriptors are more popular than region-based descriptors. The shape ...

Annals of RSCB

Vol. XVII, Issue 2/2012

COMPUTERIZED MORPHOMETRIC ASSESSMENT OF THE HUMAN RED BLOOD CELLS TREATED WITH CISPLATIN Cristina Bischin1, Ş. Ţălu2, R. Silaghi-Dumitrescu1, M. Ţălu3, S. Giovanzana4, Carmen Alina Lupaşcu5 1”BABES-BOLYAI” UNIVERSITY, DEPARTMENT OF CHEMISTRY AND CHEMICAL ENGINEERING, CLUJ-NAPOCA, ROMANIA; 2TECHNICAL UNIVERSITY OF CLUJNAPOCA, FACULTY OF MECHANICS, DEPARTMENT OF AET, DISCIPLINE OF DESCRIPTIVE GEOMETRY AND ENGINEERING GRAPHICS, CLUJ-NAPOCA, ROMANIA; 3UNIVERSITY OF CRAIOVA, FACULTY OF MECHANICS, DEPARTMENT OF APPLIED MECHANICS, CRAIOVA, ROMANIA; 4UNIVERSITY OF MILANO-BICOCCA, MILANO, ITALY; 5UNIVERSITÀ DEGLI STUDI DI PALERMO, DIPARTIMENTO DI MATEMATICA E INFORMATICA, PALERMO, ITALY

Summary The objective of this study is to perform an evaluation of representative size and shape parameters that characterise the human red blood cells either exposed to cisplatin or exposed to control solutions containing no cisplatin. A set of fourteen digital images corresponding for the human red blood cells were evaluated. Image processing and analysis of digital images were performed with ImageJ and MRI Cell Image Analyzer (MRI-CIA) softwares. We found for the human red blood cells either exposed to cisplatin or exposed to control solutions containing no cisplatin, a set of representative size and shape parameters for quantification. The central tendency and dispersion measure of the parameters were expressed by the mean value and standard deviation. The computerized geometric morphometric analysis of the human red blood cells is an efficient noninvasive prediction tool that provides important insights into cell states. Keywords: image analysis, human red blood cells, cisplatin, morphometry, shape, shape descriptor

[email protected]

Introduction Cell shape is a large-scale expression of many subtle biological processes, controlled by interactions between the cytoskeleton, the membrane and membrane-bound proteins and the extracellular environment (Pincus and Theriot, 2007). Over the last few decades different methods to analyze the structure of red blood cells (or erythrocytes), and their structure-function relationships were performed (Canham, 1970; Evans and Celle, 1975; Evans, 1983; Ruberto et al., 2002; Dao et al, 2003; Suresh, 2006). The red blood cell has a relatively simple structure and possesses a discoid form. It does not contain a nucleus. The red

The modern computerized geometric morphometric methods have been established as efficient tools to quantify differences in the cell shape or in particular morphological structures and can provide a better characterisation in describing the complexity of anatomical structures (Grizzi and Chiriva-Internati, 2005; Russ, 2007; Rosioru et al. 2012). Computational tools for the analysis of cell shape allow to quantify the similarity or difference between images of homologous anatomical structures containing multifactorial information, as well as the relationship between cell shape and experimental conditions (Pincus and Theriot, 2007; Talu, 2012). 105

Annals of RSCB

Vol. XVII, Issue 2/2012

cell membrane is considered uniform over its surface. The normal biconcave shape of the red blood cell corresponds to the shape of minimum electrostatic energy. The existence of a shape memory implies that the elastic energy has a minimum when the cell membrane is in static equilibrium (Canham, 1970; Adams, 1973; Fischer, 2004; Gov and Safran, 2005). These methods and measurements prove sufficient for some studies, however they are less well suited for quantifying the red blood cells structure concerning to the cell shape and size. Cisplatin (cis-DDP), cisdiamminedichloroplatinum(II) is one of the most effective and widely used chemotherapeutic agents, into general oncology practice. Cisplatin’s mode of action involves the binding of the drug to DNA and non-DNA targets and the subsequent induction of cell death through apoptosis, necrosis, or both (Fuertes et al. 2003). The binding of this drug to red bloods cells induce structural and functional changes, most probably due to the interaction with phospholipids located in the inner monolayer of the red blood cell membrane (Suwalsky et al., 2000; Mahmud et al., 2008; Florea and Büsselberg, 2011). To describe complex biological structures many descriptions are used. Shape representation and description is a difficult task (Grizzi and Chiriva-Internati, 2005; Zhang and Lu, 2004). Shape descriptors are one of the key computational tools used for biological and medical image processing applications. In the literature, several shape descriptors have been proposed for 2D and 3D objects. Shape descriptors are mathematical functions which are applied to an image and produce numerical values which are representative of a specific characteristic of the image. These numerical values can then be processed in order to provide some additional information about the image. The

shape descriptors can be classified into two groups: contour-based shape descriptors and region-based shape descriptors. Contour based shape descriptors only exploit shape boundary information, ignoring the shape interior information. Therefore, these descriptors cannot represent shapes for which the complete boundary information is not sufficient or not available. On the other side, regionbased descriptors exploit both boundary and internal pixels within patterns, and therefore are applicable to generic shapes. Regionbased descriptors are more computationally intensive and most methods need normalization steps. For generic purposes, both types of shape descriptors are necessary (Zhang and Lu, 2004; Martinez-Ortiz, 2010; Amanatiadis et al., 2011). In biological and medical applications, contour-based descriptors are more popular than region-based descriptors. The shape descriptors depend on the methodological and experimental parameters involved as: diversity of subjects, image acquisition, type of image, image quality, its processing, analysis methods, including the algorithm and specific calculation used (Russ, 2007; Talu, 2012). In our study we have investigated the red blood cells either exposed to cisplatin or exposed to control solutions containing no cisplatin, using computerized geometric morphometrics.

Material and methods The human blood was extracted on citrate from a health donor without any known chronic medical conditions. After incubation with cisplatin (400 µM) at 37 ºC for 20 hours, the red blood cells were fixed on the glass slides according to the protocol May-Grünwald-Giemsa, described by Alteras (Alteras and Cajal, 1994). Cisplatin powder (obtained from Sigma-Aldrich, Germany) was dissolved in saline buffer (1% NaCl). The control solution also contains saline buffer (1% NaCl). 106

Annals of RSCB

Vol. XVII, Issue 2/2012

The SEM images were obtained by the use of a scanning electron microscope (FEI Quanta 3D FEG dual beam) in high vacuum work mode using EDT (Everhart Thornley detector). Figure 1 shows the morphology of the red blood cells either exposed to control solutions containing no cisplatin or exposed to cisplatin.

(http://www.mri.cnrs.fr/index.php?m=38) (Bäcker and Travo, 2006). Let us consider a set of fourteen digital images corresponding for the human red blood cells. After correction of the digital images, by contrast adjustment and spatial filtering, shape outlines of cell membrane were extracted from binary images with a classical contour-extraction method. With area selections, the following geometrical parameters and numerical descriptors were determined: Area, Center of Mass, Perimeter, Bounding Rectangle, Fitted Ellipse, Feret’s Diameter, Skewness and Kurtosis. Details of the representative size and shape parameters, as well as the intensity statistics, used to obtain information on the red blood cells characteristics complexity are given in the Appendix.

(a) Statistical analysis After image processing and analysis with ImageJ and MRI Cell Image Analyzer, all the raw data were statistical analyzed. Descriptive statistics were calculated for the controls and cisplatintreated cells in each group and for the two different groups (controls and cisplatintreated cells). It was found that the average values of the size and shape parameters followed a normal distribution. Statistical comparison between groups was made by one-way analysis of variance (ANOVA) (p < 0.01 statistical significance).

(b) Fig. 1. SEM images of red blood cells: (a) control and (b) cisplatin-treated.

Results and discussions We evaluated the representative size and shape parameters by the three criteria: fidelity, capture of biologically relevant details and human interpretability. The obtained average results were expressed as (average ± standard deviation). A summary of the obtained results are presented in the tables given below.

Geometric morphometric analysis The computerized geometric analysis of binary images was made using the Image J software (Wayne Rasband, National Institutes of Health, in Bethesda, Maryland, USA) (http://imagej.nih.gov/ij) together with MRI Cell Image Analyzer (MRI-CIA) software, developed by the Montpellier RIO Imaging facility (CNRS) 107

Annals of RSCB

Vol. XVII, Issue 2/2012

Table 1. Results of the evaluation for the representative size and shape parameters of the red blood cells (average ± standard deviation).

Some remarks can be obtained concerning the results from Tables 1 and 2: - the red blood cells are morphologically changed from control group when are treated with cisplatin; - the average values for circularity, roundness and solidity parameters of the control group are lower than that of the treated group. - the average values for aspect ratio parameter of the control group are higher than that of the treated group. - the geometrical parameters are interrelated and provide equivalent information about the size and shape geometry. - the geometric morphometric analyses were in agreement with the histological observations. These analyses were performed using 2D representations of 3D biological shapes.

Red blood cell shape and values Parameters Area [µm2] Perimeter [µm] Circularity [-] Aspect ratio (AR) [-] Roundness [-] Solidity [-] Major [µm] Minor [µm] Angle [°] Feret [µm] FeretAngle [°] MinFeret [µm] Width [µm] Height [µm]

Control cells 10.846± 1.150 16.292 ± 1.603 0.518±0.056 1.241±0.073

Treated cells 11.554 ± 1.477 15.843 ± 1.492 0.585±0.087 1.157±0.104

0.808±0.045 0.822±0.037 4.129± 0.164 3.339± 0.247 76.162± 61.628 4.726± 0.239 102.832±54.658 3.747±0.304 4.339± 0.206 4.221± 0.623

0.870±0.075 0.846±0.040 4.109± 0.236 3.573± 0.326 83.106± 57.667 4.759± 0.321 71.912±52.344 3.860±0.296 4.380± 0.372 4.269± 0.501

In intensity statistics, each pixel has a brightness value that ranges between 0 (black) and 255 (white). Higher values usually mean lighter pixels and lower values mean darker pixels.

Conclusions Computerized geometric morphometric method for shape analysis is a useful tool for investigation of the red blood cells in a more realistic and integrative way and allows a numeric evidence of cell shape complexity. The relationship between the shape of the red blood cells and cisplatin is complex. Our results suggest that the red blood cell shape complexity can be performed using by a set of size and shape parameters in an accurate and statistically powerful way. These size and shape parameters allows a more sensitive characterization of very subtle variations in red blood cells form that could remain undetected when using traditional particle sizing techniques.

Table 2. Results of the intensity statistics evaluation for red blood cells corresponding of the digital images from Table 1 (average ± standard deviation). Red blood cell shape and values Parameters Mean Gray Value Standard Dev. 1 Modal Gray Value Min Gray Level Max Gray Level XM YM Integrated Density Median Skewness Kurtosis

Control cells 80.598 ± 3.614

Treated cells 78.742 ± 5.625

16.293 ± 1.836

15.311 ± 2.516

86.833 ± 6.494

84.417 ± 6.748

30.333 ± 3.141

34.750 ± 5.643

140.833 ± 4.167 11.963 ± 5.027 13.737 ± 5.359 874.113± 99.322 82.667±3.615 -0.399±0.126 -0.310±0.244

140.583 ± 12.638 13.149 ± 4.237 13.411 ±4.587 906.591± 103.740 80.417±5.744 -0.299±0.350 -0.266±0.498

Acknowledgements This research has been financially supported by the Romanian Ministry for Education and Research (grants ID 565/2007 and PCCE 140/2008). I would like to thank Prof. Corina Rosioru (“BabesBolyai” University, Department of Biology, Cluj-Napoca) for helpful discussions. I would also like to thank Ph.D. Adriana 108

Annals of RSCB

Vol. XVII, Issue 2/2012

Vulpoi and Prof. Simion Simon (“BabesBolyai” University, Department of Physics, Cluj-Napoca) for their assistance in SEM technique. A Ph.D. scholarship from Contract POSDRU/88/ 1.5/S/60185 – “Innovative doctoral studies in a knowledge based society” is gratefully acknowledged by C.B.

known as maximum caliper. The length of the object’s projection in the X (FeretX) and Y (FeretY) direction is also displayed. 10. FeretAngle: The angle of the Feret diameter with the x-axis of the image (0°180°). 11. MinFeret: This is a measure of the particle's width. It is called the minimum caliper diameter as well. It is defined as the shortes distance between two parallel planes touching the particle on opposite sites, for any orientation of the particle. 12. AR: The aspect ratio of the particle. This is the length of the major axis divided by the length of the minor axis of the fitted ellipse. 13. Round: The roundness of the particle, defined as: 4·area/[π·(major axis)²]. The roundness is 1 for a circle and approaches 0 for very alongated objects. 14. Solidity: The area of the particle divided by the area of the convex hull of the particle. The convex hull is a boundary enclosing the foreground pixels of an image using straight line segments to each outermost point. The intensity statistics parameters are defined as following: 1. Mean Gray Value: The average gray value within the ROI. This is the sum of the gray values of all the pixels in the selection divided by the number of pixels. 2. Standard Dev. 1: The standard deviation of the mean gray value within the ROI. 3. Modal Gray Value: Most frequently occurring gray value within the ROI. This corresponds to the highest peak in the histogram of the ROI. 4. Min & Max Gray Level: Minimum and maximum gray values within the ROI. 5. XM: The x-coordinate of the center of mass, that is the brightness weighted average of the x-coordinates of the pixels in the ROI. The coordinates (XM and YM) are the first order spatial moments. 6. YM: The y-coordinate of the center of mass, that is the brightness weighted average of the y-coordinates of the pixels in the ROI. The coordinates (XM and YM) are the first order spatial moments.

Appendix In our study, MRI Cell Image Analyzer (MRI-CIA) software, developed by the Montpellier RIO Imaging facility (CNRS)(http://www.mri.cnrs.fr/index.php? m=38) was used to determine the size and shape parameters of the given shape data. With area selections, the following geometrical parameters and numerical descriptors were determined: Area, Center of Mass, Perimeter, Bounding Rectangle, Shape Descriptors, Fitted Ellipse, Feret’s Diameter, Skewness and Kurtosis. The size and shape parameters are defined as following: 1. Area: The surface of the region of interest (ROI), measured in [µm2]. 2. Perimeter: The length of the outside boundary of the ROI, measured in [µm]. 3. Width: The width of the bounding box of the ROI (the smallest rectangle enclosing the selection). 4. Height: The height of the bounding box of the ROI. 5. Major: The length of the major axis of the best fitting ellipse. The ellipse has the same area, orientation and centroid as the original selection. 6. Minor: The length of the minor axis of the best fitting ellipse. 7. Angle: The angle of the major axis of the best fitting ellipse against the x-axis of the image. 8. Circularity: the ratio is given by 4πA/p2, where A is the area of the shape and p is the perimeter. A value of 1 indicates a perfect circle. As the value approaches 0, it indicates an increasingly elongated polygon. 9. Feret: The longest distance between two points on the boundary of the ROI, also 109

Annals of RSCB

Vol. XVII, Issue 2/2012

Fuertes M.A., Alonso C., Perez J.M., Biochemical modulation of cisplatin mechanisms of action: enhancement of antitumor activity and circumvention of drug resistance, Chem. Rev., 103(3), 645-662, 2003. Gov N., Safran S.A., Red blood cell shape and fluctuations: cytoskeleton confinement and ATP activity, J. Biol. Phys., 31(3-4), 453464, 2005. Grizzi F., Chiriva-Internati M., The complexity of anatomical systems, Theor. Biol. Med. Model., 2, 26, 2005. Mahmud H., Föller M., Lang F., Suicidal erythrocyte death triggered by cisplatin, Toxicology, 249(1), 40-44, 2008. Martinez-Ortiz C., PhD. Thesis: 2D and 3D Shape Descriptors, University of Exeter, UK, 2010. Pincus Z., Theriot J.A., Comparison of quantitative methods for cell-shape analysis, Journal of Microscopy, 227 (Pt. 2), 140-156, 2007. Rosioru C., Talu S., Talu M., Giovanzana S., Craciun C., Morphometric assessments for the healthy rat hepatocytes, Annals of RSCB, vol. XVII, issue 1, 74-79, 2012. Ruberto C.Di., Dempster A., Khan S., Jarra B., Analysis of infected blood cell images using morphological operators, Image Vision Comput., 20(2), 133-146, 2002. Russ J.C., The image processing Handbook, 5th edition, CRC Press, Taylor & Francis Group, USA, 2007. Suresh S., Mechanical response of human red blood cells in health and disease: some structure-property-function relationships, J. Mater. Res., 21(8), 1871-1877, 2006. Suwalsky M., Hernandez P., Villena F., Sotomayor C.P., The anticancer drug cisplatin interacts with the human erythrocyte membrane, Z. Naturforsch. C., 55(5-6), 461466, 2000. Talu S., Texture analysis methods for the characterisation of biological and medical images, ELBA Bioflux, 4(1), 8-12, 2012. Talu S., Mathematical methods used in monofractal and multifractal analysis for the processing of biological and medical data and images, ABAH Bioflux, 4(1), 1-4, 2012. Zhang D., Lu G., Review of shape representation and description techniques, Pattern Recognition, 37(1), 1-19, 2004. ***http://imagej.nih.gov/ij ***http://www.mri.cnrs.fr/index.php?m=38

7. Integrated Density: The integrated density is the sum of the gray-values of all pixels within the ROI. 8. Median: The median gray value of the pixels within the ROI. 9. Skewness: A measure of the asymmetry of the distribution of the gray values around the mean within the ROI. The third order moment about the mean. 10. Kurtosis: A measure of the "peakedness" of the distribution of the gray values around the mean within the ROI. The fourth order moment about the mean.

References Adams K.H., A theory for the shape of the red blood cell, Biophys J., 13(10), 1049-53, 1973. Alteras I., Cajal N., Methods in clinical laboratory, Ed. Medicală, Bucharest, 1994. Amanatiadis A., Kaburlasos V.G., Gasteratos A., Papadakis S.E., Evaluation of shape descriptors for shape-based image retrieval, IET Image Process., 5(5), 493-499, 2011. Bäcker V., Travo P., Cell Image Analyzer - A visual scripting interface for ImageJ and its usage at the microscopy facility Montpellier RIO Imaging, in Proceedings of the ImageJ User and Developer Conference, Edition 1, p. 105-110, ISBN: 2-919941-01-1, EAN: 9782919941018, 2006. Canham P.B., The minimum energy of bending as a possible explanation of the biconcave shape of the human red blood cell, J. Theor. Biol., 26(1), 61-76, 1970. Dao M., Lim C.T., Suresh S., Mechanics of the human red blood cell deformed by optical tweezers, J. Mech. Phys. Solids, 51, 22592280, 2003. Evans E.A., Celle P.L., Intrinsec material properties of the erytrocyte membrane indicated by mechanical analysis of deformation, Blood, 45(1), 29-43, 1975. Evans E.A., Bending elastic modulus of red blood cell membrane derived from buckling instability in micropipet aspiration tests, Biophys. J., 43(1), 27-30, 1983. Fischer T.M., Shape memory of human red blood cells, Biophys. J., 86(5), 3304-13, 2004. Florea A.M., Büsselberg D., Cisplatin as an anti-tumor drug: cellular mechanisms of activity, drug resistance and induced side effects, Cancers, 3(1), 1351-1371, 2011.

110

Suggest Documents