This paper presents a new method to separate cells on microscopic surfaces joined together in cell clusters into individual cells. Important features of this method are that the remaining object geometry is preserved and few contour points are required for finding joints between neighboring cells. There are alternative methods such as morphological operations or the watershed transformation based on the inverse distance transformation but they have certain disadvantages compared to the method presented in this paper. The discussed method contains knowledge-based components in form of a decision function and exchangeable rules to avoid unwanted separations.
In the process of testing implant materials for biocompatibility, it is important to evaluate whether a material is suitable for use in human bodies. An important aspect of biocompatibility is the determination of the exact number of cells which are in contact with the surface of the material being tested. For the specimen preparation process, a suspension with a defined cell concentration reacts for a certain time with the substrate under test and allows the cells to settle on the contact surface. Afterwards, the cells are stained using the May-Grünwald suspension [
L929 cells on the substrate steel. Right: a biological cell division process, two joint cells which have a sand glass appearance.
Several papers deal with different image processing methods for cell segmentation [
The separation of connected cells is still a great challenge. Several papers provide different approaches to separate cells of a specific type [
An iterative erosion method may create a separation of cells or objects at joining points between cells. After each iteration, it has to be checked whether separated objects have been created. The algorithm is not able to separate the cells without altering the cell contours. Therefore, a reconstruction of the cell area is required. A big disadvantage of this method is that it also removes or separates cell extensions, which is not acceptable in our case. An improved method for separating cells is described in [
Another common procedure to separate connected cells is the watershed transformation based on the inverse distance transformation [
The method for the separation of connected cells described in [
The images were created by the Olympus XC10 camera, which is installed on an Olympus Bx51 M microscope with 100x magnification. The size of the images is
The resulting three binary images
Cell segmentation process. The steps within the brackets are carried out for the cell region segmentation (Section
Figure
The new method presented in this paper is called the context-based separation (CBS). This method performs splitting operations on cells or other objects. This splitting is restricted to narrow joints between cells or objects. In an image
The contour of the object
For each
Average relative deviation of the automatic cell count (2376 cells) to the reference cell count compiled by an expert is determined.
|
|
|
|
|
|
---|---|---|---|---|---|
Relative deviation | 8.6% | 4.6% | 2.7% | 3.1% | 5.4% |
The background
With help of
With the help of
Average relative deviation of the automatic cell count (2376 cells) to the reference cell count compiled by an expert.
|
|
|
|
|
|
---|---|---|---|---|---|
Relative deviation | 5.8% | 2.7% | 2.9% | 3.8% | 7.4% |
Average relative deviation of the automatic cell count (2376 cells) to the reference cell count compiled by an expert.
|
|
|
|
|
|
---|---|---|---|---|---|
Relative deviation | 26.7% | 5.3% | 2.7% | 4.2% | 8.1% |
Schematic illustration of possible object shapes at a cell contact point.
Schema | Description | |
---|---|---|
Case 1 |
|
The circular ring |
|
||
Case 2 |
|
There are two separated cell regions within |
|
||
Case 3 |
|
There is a separated cell area within |
|
||
Case 4 |
|
|
|
||
Case 5 |
|
Compared to case 4, there are one or more separate cell areas located within |
If
Applying the region labeling after the separation process of the cluster shown in Figure
The result of the cell separation process. Three new cells
There may be situations in which object separations do not make sense, for example, at contact points between cell extensions and cell body, even though
Incorrect separation of a cell extension from its cell.
The extended decision function is
Incorrect splitting of a cell extension.
All object regions in
The compactness of all object regions in
Generally in image processing the compactness is defined as
Figure
Simple flow-diagram of separating objects at the narrow joints.
The agglomerate
Activity diagram of the CBS.
Figure
After calculation of
If
(a) 2 cells; (b) two connected cells; (c) two connected cells where a left cells region extends into
(a) Cells of type L929. (b) Cell segmentation result before the separation process. Some cells are connected to each other. (c) Cell separation with the method described in [
The number of cell areas within
In the context of biocompatibility testing of implant materials, the determination of the proliferation rate (cell count) and the evaluation of the cell morphology are important features. Therefore, an accurate determination of the cell borders within the clusters is necessary. The cytotoxicity of an implant material is classified in 4 levels depending on the proliferation rate (Table
Assignment of the cell proliferation rate to the four levels of cytotoxicity.
Cytotoxicity level | Proliferation rate (%) | Interpretation |
---|---|---|
0 | 100–81 | Not cytotoxic |
1 | 80–71 | Low cytotoxicity |
2 | 70–61 | Moderate cytotoxicity |
3 | 60–0 | Strong cytotoxicity |
Table
To evaluate the efficiency of the presented method, first the number of cells for 10 samples (2409 cells) of the cell type L929 on the substrates steel as well as titanium is determined and compared to the number of cells specified by an expert (reference cell count). The part of L929 cells which are connected to clusters is in average 23.3%. Table
Comparison of cell counting using the CBS and the methods presented in [
Sample | ACC | FP | FN | PC | REF | RFP (%) | RFP (%) | ERR (%) |
---|---|---|---|---|---|---|---|---|
CBS Algorithm | ||||||||
1 | 234 | 4 | 2 | 230 | 232 | 1.7 | 0.9 | 0.9 |
2 | 350 | 3 | 22 | 347 | 369 | 0.8 | 6.0 | 5.1 |
3 | 240 | 2 | 12 | 238 | 250 | 0.8 | 4.8 | 4.0 |
4 | 193 | 3 | 9 | 190 | 199 | 1.5 | 4.5 | 3.0 |
5 | 232 | 4 | 12 | 228 | 240 | 1.7 | 5.0 | 3.3 |
6 | 205 | 9 | 3 | 196 | 199 | 4.5 | 1.5 | 3.0 |
7 | 217 | 3 | 1 | 214 | 215 | 1.4 | 0.5 | 0.9 |
8 | 208 | 5 | 7 | 203 | 210 | 2.4 | 3.3 | 1.0 |
9 | 220 | 6 | 1 | 214 | 215 | 2.8 | 0.5 | 2.3 |
10 | 270 | 6 | 16 | 264 | 280 | 2.1 | 5.7 | 3.6 |
|
|
|
|
|||||
|
||||||||
Method described in [ | ||||||||
1 | 255 | 24 | 1 | 231 | 232 | 10.3 | 0.4 | 9.9 |
2 | 380 | 16 | 5 | 364 | 369 | 4.3 | 1.4 | 3.0 |
3 | 269 | 21 | 2 | 248 | 250 | 8.4 | 0.8 | 7.6 |
4 | 223 | 25 | 1 | 198 | 199 | 12.6 | 0.5 | 12.1 |
5 | 260 | 23 | 3 | 237 | 240 | 9.6 | 1.3 | 8.3 |
6 | 222 | 25 | 2 | 197 | 199 | 12.6 | 1.0 | 11.6 |
7 | 247 | 36 | 4 | 211 | 215 | 16.7 | 1.9 | 14.9 |
8 | 231 | 23 | 2 | 208 | 210 | 11.0 | 1.0 | 10.0 |
9 | 236 | 22 | 1 | 214 | 215 | 10.2 | 0.5 | 9.8 |
10 | 296 | 20 | 4 | 276 | 280 | 7.1 | 1.4 | 5.7 |
|
|
|
|
|||||
|
||||||||
Method described in [ | ||||||||
1 | 245 | 14 | 1 | 231 | 232 | 6.0 | 0.4 | 5.6 |
2 | 391 | 25 | 3 | 366 | 369 | 6.8 | 0.8 | 6.0 |
3 | 258 | 8 | 0 | 250 | 250 | 3.2 | 0.0 | 3.2 |
4 | 211 | 14 | 2 | 197 | 199 | 7.0 | 1.0 | 6.0 |
5 | 252 | 13 | 1 | 239 | 240 | 5.4 | 0.4 | 5.0 |
6 | 208 | 10 | 1 | 198 | 199 | 5.0 | 0.5 | 4.5 |
7 | 226 | 13 | 2 | 213 | 215 | 6.0 | 0.9 | 5.1 |
8 | 219 | 10 | 1 | 209 | 210 | 4.8 | 0.5 | 4.3 |
9 | 230 | 16 | 1 | 214 | 215 | 7.4 | 0.5 | 7.0 |
10 | 292 | 14 | 2 | 278 | 280 | 5.0 | 0.7 | 4.3 |
|
|
|
|
ACC: automatic cell count; FP: false positive detection; FN: false negative (missed) detection; PC: positive correct detection (ACC − FP); REF: reference cell count; RFP: relative false positive detection (FP/REF * 100); RFN: relative false negative detection (FN/REF * 100); ERR: relative error (((REF − ACC)/REF) * 100).
Separating the cells with the CBS method results in a mean cell count error of 2.7%. In comparison, using the methods described in [
The used algorithms described in [
As mentioned the CBS can be used independently of other procedures or it can be combined with other methods, for example, the algorithm described in [
Comparison of calculation time for the separation algorithm described in [
Sample | Calculation time Algorithm excluding CBS (s) | Calculation time Algorithm including CBS (s) | Speed-Up |
---|---|---|---|
1 |
|
|
6.53 |
2 |
|
|
2.76 |
3 |
|
|
4.89 |
4 |
|
|
6.21 |
5 |
|
|
3.50 |
6 |
|
|
3.73 |
7 |
|
|
2.75 |
8 |
|
|
4.00 |
9 |
|
|
12.40 |
10 |
|
|
37.75 |
|
|
By using the CBS a speed-up factor of about 8 may be achieved, compared to the cell count algorithm without the use of CBS (Table
In comparison, the method described in [
The evaluation of the cell morphology is an important aspect in the context of biocompatibility testing. Therefore, an accurate segmentation of the cell boundaries within the clusters is important in order to obtain a reliable result. To evaluate the methods precision concerning the cell area segmentation, 100 automatically segmented cell regions within clusters were compared with the cell regions evaluated manually by an expert. The Jaccard coefficient is a suitable method to determine the quality of the separation. If two objects are equal in shape and area, the Jaccard coefficient is 1. Table
Calculated Jaccard coefficient based on 100 reference L929 cells within clusters for different cell separation procedures.
CBS | Method described in [ |
Method described in [ |
Method described in [ |
---|---|---|---|
0.84 | 0.80 | 0.75 | 0.76 |
Table
The methods [
The required parameters for the decision function
In this paper, a method for segmenting and separating cells in clusters is presented. The algorithm first segments histological stained cell regions in microscopic images with a standard threshold method applied to each color channel. The separation of connected cells at narrow joints is carried out by sampling the cluster contour with a circular structural element. Within the circular element, the cell geometry is analyzed and the result of a decision function indicates whether a local narrowing exists or not. An extension of the decision function with two exchangeable conditions avoids unwanted separation processes and improves the cell area segmentation. The method can be used to separate any segmented objects which have narrow joints at their contact areas. It has a very fast execution speed since not all contour points have to be processed. The procedure can be combined very well with other separation methods. This leads to better results and reduces the overall calculation time.
The authors declare that there is no conflict of interests regarding the publication of this paper.
This work was supported by the “Bundesministerium für Forschung und Entwicklung,” Germany (no. 17033X10).