NeuroCounter

NeuroCounter is an image analysis program for semi-automated particles detection and quantification. The user can select several detection methods with different level of interaction (Threshold, Watershed, Fill, Polygon and Circle drawing).
Image is split into a grid of counting frames, one of them is selected as active (displayed bold). Program detects particles inside the active frame, outlines their contours and calculates the features like area, diameter, circularity etc. User collects results frame-by-frame sorting them by individual parameters. Final results can be saved in a text format readable by Excel.

DEMO 1

1. In File menu, open 'Demo.jpg' sample image file).
2. In Detection menu:
   • Select a counting frame in first row and second column by Shift+Left mouse button click. Keep default values of all controls unchanged and press button Detect. Program paints detected cells green.
   • Press Select all and then Lock buttons.
   • Select counting frame in first row and third column and repeat the same steps.
3. Switch to Results menu mode which displays the total number of counted cells in all (two) already processed frames (Figure 1).
4. Switch to File menu again and Save results in text format readable by e.g. Excel.

Figure 1. Simple detection of cells in first row of counting frames. The Counts on the left side displays the number of cells in selected frame and in all frames (equal to estimated because the counting frames cover 100% of image area)

DEMO 2

1. In File menu, open 'Demo.jpg' sample image file). Try zoom (CTRL+mouse wheel) and scroll (CTRL+left mouse button)
2. In Image menu:
   
   • Define Area of interest (AOI) by pressing Outline area button and clicking points (polygon vertices) around the larger structure on the image. If finished, close polygon by the same button and adjust its shape by the slider under Outline area button.
   • Adjusting of counting frames is possible by Distance and Size sliders and Random button. In the default configuration (Size=Distance) the counting frames cover 100% of AOI area.
   • Calibrate system if you want to display values in real units. Use Draw line to find the correspondence between the drawn line length in pixels and real units in microns.
3. In Detection menu:
   • Select a counting frame in first row and second column by Shift+Left mouse button click. Keep default values of all controls unchanged and press button Detect. Program paints detected cells green. Press Zoom frame to see details of detected contours.
   • Change available sliders Threshold, Circularity and Size and watch their influence to the detection results. While Threshold slider is related to the Watershed detection method, the other two sliders filter cells having circularity and size outside the given limits.
   • Select a group of cells by left mouse click or by drawing rectangle bounding them. Selected particles (drawn orange) can be removed by Delete or corrected by Adjust>, Inflate or Deflate buttons. An alternative way to delete wrongly detected particles is to use right mouse click. Check button reveals the overlapped contours and allows their deleting
   • Change Mode of operation from Watershed to Fill. Clicking the interior of a not detected cell, algorithm detects the particle inside the red circle (see Figure 2).
   • If Watershed or Fill modes fail, switch to Polygon or Circle and manually enter polygon vertices or circle position (its diameter can be changed by mouse wheel).
   • Press Select all and then Lock buttons. Locked contours can be saved ( Save or Append buttons in File menu ) and subsequently released from memory (by Reset button) which speeds-up program execution. Warning: Selection of another counting frame during processing deletes all unlocked cells from the current counting frame.
   • Repeat described steps for all relevant counting frames.
4. Switch to Results menu item, select sorting by Area. Use the slider to set a value for the Split operation. Figure 3 demonstrates splitting all cells into three categories according to their area. Reset button clears the selection for a new settings.
5. Switch to File menu again and Save results in text format readable by e.g. Excel.

Figure 2, Left: Zoomed counting frame. Particles not found by automated Watershed method are detected by semi-interactive Fill mode (search for the maximal gradient contour inside red cursor) or by manual enter in Polygon or Circle mode. Right: Counting frames where Distance = 2 X Size.

Figure 3. Sorting of cells into three categories by their area

DEMO 3

1. In File menu, open 'Demo.jpg' sample image file). Try zoom (CTRL+mouse wheel) and scroll (CTRL+left mouse button)
2. In Image menu:
   
   • Define Area of interest by pressing Outline area button and clicking points (polygon vertices) around the larger structure on the image. If finished, close polygon by the same button and adjust its shape by the slider under Outline area button.
   • Adjusting of counting frames is possible by Distance and Size sliders and Random button. Keep Distance unchanged but change Size to one half (cca. 256) as shown on Figure 2 right. It means that the estimated total area is 4 x the area of counting frames.
   • Calibrate system if you want to display values in real units. Use Draw line to find the correspondence between the drawn line length in pixels and real units in microns.
3. In Tools menu:
   • Repeatedly Run the stereological counting frames test and watch how the estimated value converges to the generated one. In every cycle, program generates 1000 random positions and counts these lying inside counting frames. This value multiplied by 4 estimates the total count and should converge to 1000 on the average.
   • External program exploiting Deep Neural Network U-net architecture requires training set consisting of annotated images. Both can be exported by Images and Labels buttons. This feature is currently under testing.
4. You can continue with the Detect and Results menu by the same manner as in Demo 2.

License agreement

Copyright (C) 2020, Institute of Experimental Physics, Slovak Academy of Sciences, all rights reserved.
This software is provided by the copyright holders and contributors "as is" and any express or implied warranties, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose are disclaimed. In no event shall the copyright holders and contributors be liable for any direct, indirect, incidental, special, exemplary, or consequential damages (including, but not limited to, procurement of substitute goods or services; loss of use, data, or profits; or business interruption) however caused and on any theory of liability, whether in contract, strict liability, or tort (including negligence or otherwise) arising in any way out of the use of this software, even if advised of the possibility of such damage.