Application Name

IKOSA AI Multichannel Instance Segmentation

for Multichannel Images

App
Version	1.0.0
Documentation Version	13.09.2022 - 1
Input Image(s)	Multichannel Images (standard and/or WSI; 8/16 bit); same channel number as in training images (max. 10)
Input Parameter(s)	Max. outgrowth: For each class (trained label) a maximum outgrowth in the range of 0 - 99 pixels. Channel properties channel name channel color Regions of interest (optional)
Keywords	fluorescence, multichannel, immunofluorescence, multiplex, phenotype, microscopy
Short Description	Instance Segmentation of objects that resemble ones annotated and labeled in the training data of this application. Extraction of morphometric parameters and intensity/distance/density measures of objects and corresponding outgrowth areas.
References / Literature	-

IKOSA AI Instance Segmentation App for Multichannel Images

App

Your IKOSA AI application was created using our training for instance segmentation algorithms models for multichannel images. All information presented in this document is applicable to your trained application.

Application description

This application automatically segments instances of objects that resemble ones previously annotated and labeled in the training data of this application. The application supports multiple classes, for each class an outgrowth region can be defined separately. Object count, morphometric feature measurements, densities and distances between instances are calculated. Channel intensities are measured for the object-, outgrowth-, and combined area.

In the following, the prerequisites for an accurate analysis are outlined and the output of the application is described.

Input data requirements

Input image(s)

Input for this application is the following image data:

Image type

Color channels

Color depth (per channel)

Size (px)

Resolution (μm/px)

Multichannel (standard and/or WSI)

Check image format

File formats

same as training images (max. 10)

8 Bit or 16 Bit

WSI (CZI) formats: arbitrary

Standard images: max.

6

7,

144

168 x

6

7,

144

168

arbitrary

Image content

Arbitrary

Additional requirements

None

Note

Important:

For all images, the following requirements apply:

The illumination must be constant throughout the image(s).
The sample must be in focus, i.e. no blurry regions in image(s).

Input parameter(s)

Required input parameters:

Max. outgrowth: For each class (trained label) a maximum outgrowth in the range of 0 - 99 pixels can be defined.
channel name (e.g. “channel-DAPI”)
channel color (e.g. “00FF00”)

As an optional parameter, a single or multiple regions of interest (ROIs) can be defined in which the analysis should be performed (‘inclusion ROIs’).

Description of output files and their content

Files

File format

Description

1

csv

results.csv

A csv file containing the overall analysis results for the input image or all inclusion ROIs.

2

csv

results_<xx>_<class-name>.csv

A csv file containing the analysis results for all detected objects of class number <xx> with class name <class-name> (in training data) in the input image or inclusion ROIs.

3

jpg

results_vis/<xx>_<class-name>_vis.jpg (2D image, no ROI), or

results_vis/<xx>_<class-name>_t<time-step>_z<z-layer>_c<channel>.jpg (for time series, z-stack, or multichannel image, no ROI), or

results_vis/<xx>_<class-name>_<roi-id>.jpg (2D image, ROI <roi-id>), or results_vis/<xx>_<class-name>_t<time-step>_z<z-layer>_c<channel>_<roi-id>.jpg for time series, z-stack, or multichannel image, ROI <roi-id>):

A visualization of the analysis result for a specific time step (of a time series), z-layer (of a z-stack), or channel (of a multichannel image) for either the whole image (if no inclusion ROIs selected for analysis) or each individual inclusion ROI, for each class number <xx> with class name <class-name> (in training data).

Each visualization includes two parts:

Input
- input image/ROI, channels blended with set channel colors to RGB image
Objects
- blended RGB image (same as first part)
- object instances are shown as an overlay in random color.
- object instance index number corresponds to the object_id as listed in the results_<xx>_<class-name>.csv file.
Objects and outgrowth
- blended RGB image (same as first part)
- instance contours of detected objects and outgrowth contours from instance are marked with black/white solid lines

4

json

annotation_results.json

A json file that includes geometries of the objects detected in the input image or inclusion ROIs.

5

json

roiMeta.json

A json file containing all information regarding the ROIs defined for the analysis job to ensure reproducibility. The file is empty if no ROIs were defined for analysis.

6

jpg

rois_visualization.jpg or

t<time-step>_z<z-layer>_rois_visualization.jpg

An overview visualization to show locations of all analyzed ROIs for the 2D image or time step <time-step> of a time series, z-layer <z-layer> of a z-stack.

Info
Please note: This file is only created if inclusion ROIs were defined for analysis.

7

json

jobResultBundleMeta.json

A json file containing all information regarding the analysis job (application name and version, project, etc.) to ensure reproducibility.

Info
Please note: This file is only included if bundled or merged analysis jobs are downloaded.

Info

Please note:

In the case of inclusion ROIs that are partially outside of the image, the ROIs are cropped to the areas that lie inside the image.
In the case of inclusion ROIs that are completely outside of the image, no analysis is performed. However, they are still listed in corresponding results files.
A <roi-id> is generated automatically by the application corresponding to the creation date of a ROI. The location of a ROI within an image with its specific <roi-id> can be seen in the file “rois_visualization.jpg.” ROIs that are completely outside of the image are not shown in this file.
All visualizations are downscaled to 25 megapixels (MP), if the original image or inclusion ROI is larger than 25 MP.

Content

results.csv

Single csv-file

If one or more time steps (of a Time Series), or z-layers (of a z-Stack) were specified, the results in a specific row refer to the time step/z-layerspecified layer specified in the corresponding column.

If one or more ROIs were specified, the results in a specific row refer to the ROI specified in the corresponding columns, otherwise (empty ROI columns) the results refer to the whole image.

Column NO.	Column name	Examples	Value range	Description
1	t	3	1 -	Time step, i.e. the position of the image in the time series.
2	z	5	1 -	z-layer, i.e. the position of the layer in the z-stack.
3	roi_id	ROI-03	ROI-01 -	<roi-id> starting from “ROI1”. Empty, if no inclusion ROI is specified and the whole image was analyzed.
4	roi_name	“central”	text	Custom text to identify the ROI. Empty, if no inclusion ROI is specified and the whole image was analyzed.
5	roi_size [Px^2]	1212212	1 -	Size of the ROI that was analyzed in pixels^2. The size of the whole image is given if no inclusion ROI is specified and the whole image was analyzed.
6	bit depth [Bit]	8	8, 16	Bit/color depth of each channel of the image
7	class	“cell nuclei”	text	Name of the trained label (<class-name>).
8	object_max_outgrowth [Px]	2	0 - 99	Maximum outgrowth in pixels.
9	total_num_objects	788	1 -	Total number of objects which are detected in the image/ROI.
10	density_num_objects [1/Px^2]	0.0007514953	0 - 1	Total number of objects divided by area of image/ROI.
11	total_area_objects [Px^2]	101863	8 - #pixels	Total area of objects which are detected in the image/ROI.
12	density_area_objects	0.0971441268	0 - 1	Total area of objects divided by area of image/ROI.
13	total_area_outgrowth [Px^2]	125326	0 - #pixels	Total area of object’s outgrowth which are detected in the image/ROI.
14	density_area_outgrowth	0.11952018737	0 - 1	Total area of outgrowth divided by area of image/ROI.
15	total_area_objects_incl_outgrowth [Px^2]	227189	8 - #pixels	Total area of objects including outgrowth which are detected in the image/ROI.
16	density_area_objects_incl_outgrowth	0.2166643142	0 - 1	Total area of objects including outgrowth divided by area of image/ROI
17	mean_area_objects [Px^2]	14.324523	8 - #pixels	Mean area of detected objects in Pixels^2.
18	stddev_area_objects [Px^2]	1.213133	0 - #pixels	Standard deviation of area of detected objects in Pixels^2.
19	median_area_objects [Px^2]	13.983252	8 - # pixels	Median area of detected objects in Pixels^2.
20	mean_area_outgrowth [Px^2]	21.363824	0 - #pixels	Mean area of outgrowth in Pixels^2.
21	stddev_area_outgrowth [Px^2]	2.573912	0 - #pixels	Standard deviation of areas of outgrowth in Pixels^2.
22	median_area_outgrowth [Px^2]	20.874623	0 - #pixels	Median area of outgrowth in Pixels^2.
23	mean_area_objects_incl_outgrowth [Px^2]	35.836274	8 - #pixels	Mean area of detected objects including outgrowth in Pixels^2.
24	stddev_area_objects_incl_outgrowth [Px^2]	2.309003	0 - #pixels	Standard deviation of area of detected objects including outgrowth in Pixels^2.
25	median_area_objects_incl_outgrowth [Px^2]	35.473222	8 - #pixels	Median area of detected objects including outgrowth in Pixels^2.
26	mean_distance [Px]	13.5016475693	0 -	Mean distance (center-to-center) between detected objects in Pixels. The distance is calculated by the nearest-neighbor distance to other objects of the same class.
27	stddev_distance [Px]	5.5485792042	0 -	Standard deviation of distances between detected objects in Pixels. The distance is calculated by the nearest-neighbor distance to other objects of the same class.
28	median_distance [Px]	12.529964086	0 -	Median distance between detected objects in Pixels. The distance is calculated by the nearest-neighbor distance to other objects of the same class.
29	mean_bbox_area [Px^2]	202.03426395	8 -	Mean area of bounding box of detected objects in Pixels^2.
30	mean_area_ratio [%]

*

	0.0124492385	0 - 100	Mean area of detected object as percentage of overall image area or ROI area inside the image.
31*	mean_perimeter [Px]

*

	40.434010152	8 -	Mean perimeter of detected objects in Pixels.
32*	mean_circularity

*

0.92114213197

0 - ~1

Mean circularity factor of detected objects.

Circularity = 4*pi*area/(perimeter^2).

The circularity of a circle is 1.

Note: Detected objects with a circularity of infinity are ignored when calculating mean_circularity.

33*

mean_solidity

*

0.9310659898

0 - 1

Mean solidity value of all detected objects.

Solidity is calculated out of the ratio of an object’s area to the area of its convex hull.

34*

mean_eccentricity

*

0.7109644670

0 - 1

Mean eccentricity of all detected objects.

Eccentricity of the ellipse that has the same second-moments as the object area. The eccentricity is the ratio of the focal distance (distance between focal points) over the major axis length. When it is 0, the ellipse becomes a circle.

35*

mean_equivalent_diameter [Px]

*

11.737309644

1 -

Mean equivalent diameter of all detected objects.

The equivalent diameter is the diameter of a circle having the same area as the detected object.

36*

mean_extent

*

0.709022842

0 - 1

Mean extent of all detected objects.

The extent value is the ratio of the object area to the total bounding box area.

37*

mean_minor_axis_length [Px]

*

9.4835025380

1 -

Mean minor axis length of all detected objects.

The minor axis length is the length of the minor axis of the ellipse that has the same normalized second central moments as the object.

38*

mean_major_axis_length [Px]

*

	15.843908629	1 -	Mean major axis length of all detected objects. The major axis length is the length of the major axis of the ellipse that has the same normalized second central moments as the object.
For the different channels [CC] in the multichannel image:
39/ 51/ …	channel_#[CC]_index	1	1, 2, 3	Index of channel
40/ 52/ …	channel_#[CC]_name	“DAPI”	text	Name of channel
41/ 53/ …	channel_#[CC]_total_intensity	170486667	0 - 65535 x #pixels	Total intensity (sum) of this channel CC in the image/ROI.
42/ 54/ …	channel_#[CC]_objects_mean_intensity	272.93074030	0 - 65535	Mean intensity of this channel CC over the areas of all detected objects.
43/ 55/ …	channel_#[CC]_objects_stddev_intensity	130.84258715	0 - 65535	Standard deviation of intensity of this channel CC over the areas of all detected objects.
44/ 56/ …	channel_#[CC]_objects_intensity_dominant	265	0 - total_num_objects	Number of objects for which mean intensity is highest (=dominant) in this channel CC, compared to all other channels.
45/ 57/ …	channel_#[CC]_outgrowth_mean_intensity	273.26551765	0 - 65535, or ”” (empty)	Mean intensity of this channel CC over the areas of the outgrowths of all detected objects. Is empty ““ if no outgrowth was set (max. outgrowth was set to 0 pixels).
46/ 58/ …	channel_#[CC]_outgrowth_stddev_intensity	203.15327249	0 - 65535, or ”” (empty)	Standard deviation of intensity of this channel CC over the areas of the outgrowths of all detected objects. Is empty ““ if no outgrowth was set (max. outgrowth was set to 0 pixels).
47/ 59/ …	channel_#[CC]_outgrowth_intensity_dominant	724	0 - total_num_objects	Number of objects for which mean intensity of outgrowth areas is highest (=dominant) in this channel CC, compared to all other channels.
48/ 60/ …	channel_#[CC]_objects_incl_outgrowth_mean_intensity	273.115521066	0 - 65535	Mean intensity of this channel CC over the areas of the objects including outgrowths of all detected objects.
49/ 61/ …	channel_#[CC]_objects_incl_outgrowth_stddev_intensity	174.47822028	0 - 65535	Standard deviation of intensity of this channel CC over the areas of the objects including outgrowths of all detected objects.
50/ 62/ …	channel_#[CC]_objects_incl_outgrowth_intensity_dominant	530	0 - total_num_objects	Number of objects for which mean intensity of areas of detected objects incl. outgrowth is highest (=dominant) in this channel CC, compared to all other channels.

results_<xx>_<class-name>.csv

Single or multiple csv-file(s)

If one or more time steps (of a Time Series), or z-layers (of a z-Stack) were specified, the results in a specific row refer to the time step/z-layer specified in the corresponding column.

If one or more ROIs were specified, the results in a specific row refer to the ROI specified in the corresponding columns, otherwise (empty ROI columns) the results refer to the whole image.

Column NO.	Column name	Examples	Value range	Description
1	t	3	1 -	Time step, i.e. the position of the image in the time series.
2	z	5	1 -	z-layer, i.e. the position of the layer in the z-stack.
3	roi_id	ROI-03	ROI-01 -	<roi-id> starting from “ROI1”. Empty, if no inclusion ROI is specified and the whole image was analyzed.
4	roi_name	“central”	text	Custom text to identify the ROI. Empty if no inclusion ROI is specified and the whole image was analyzed.
5	roi_size [Px^2]	1212212	1 -	Size of the ROI that was analyzed in pixels^2. The size of the whole image is given if no inclusion ROI is specified and the whole image was analyzed.
6	bit depth [Bit]	8	8, 16	Bit/color depth of each channel of the image
7	max_outgrowth [Px]	2	0 - 99	Maximum outgrowth in pixels.
8	object_id	5	1 -	ID of detected object.
9	object_center_x	741	1 - image/ROI width	x-coordinate of object center (starting from upper left corner, along image/ROI width).
10	object_center_y	26	1 - image/ROI height	y-coordinate of object center (starting from upper left corner, along image/ROI height).
11	object_area [Px^2]	174	8 -	Area of object in pixels^2.
12	outgrowth_area [Px^2]	310	0 - or ““ (empty)	Area of outgrowth in pixels^2. Is empty ““ if no outgrowth was set (max. outgrowth was set to 0 pixels).
13	object_incl_outgrowth_area [Px^2]	484	8 -	Sum of areas of object and outgrowth in pixels^2.
14	distance_nearest_neighbor [Px]	22.090722034	0 -	Distance (center-to-center) to the nearest other object of the same class in pixels.
15	bbox_area [Px^2]	304	8 -	Area of bounding box of detected object in pixels^2.
16	area_ratio [%]	0.02	0 - 100	Area of detected object as percentage of overall image area or ROI area inside the image.
17*	perimeter [Px]	53	8 -	Perimeter of detected object in Pixels.
18*	circularity	0.78	0 - ~1	Circularity of detected object. Circularity = 4piarea/(perimeter^2). The circularity of a circle is 1.
19*	solidity	0.97	0 - 1	Solidity of detected object. Solidity is calculated out of the ratio of an object’s area to the area of its convex hull.
20*	eccentricity	0.92	0 - 1	Eccentricity of detected object. Eccentricity of the ellipse that has the same second-moments as the object area. The eccentricity is the ratio of the focal distance (distance between focal points) over the major axis length. When it is 0, the ellipse becomes a circle.
21*	equivalent_diameter [Px]	14	1 -	Equivalent diameter of detected object. The equivalent diameter is the diameter of a circle having the same area as the detected object.
22*	extent	0.57	0 - 1	Extent of detected object. The extent value is the ratio of the object area to the total bounding box area.
23*	minor_axis_length [Px]	9	1 -	Minor axis length of detected object. The minor axis length is the length of the minor axis of the ellipse that has the same normalized second central moments as the object.
24*	major_axis_length [Px]	23	1 -	Major axis length of detected object. The major axis length is the length of the major axis of the ellipse that has the same normalized second central moments as the object.
For the different channels [CC] in the multichannel image:
25/ 36/ …	channel_#[CC]_index	1	1, 2, 3	Index of channel
26/ 37/ …	channel_#[CC]_name	“DAPI”	text	Name of channel
27/ 38/ …	channel_#[CC]_object_mean_intensity	272.93074030	0 - 65535	Mean intensity of this channel CC over the area of the detected object.
28/ 39/ …	channel_#[CC]_object_stddev_intensity	130.84258715	0 - 65535	Standard deviation of intensity of this channel CC over the area of the detected object.
29/ 40/ …	channel_#[CC]_object_intensity_dominant	“True”	“True”, or ”False”, or ”nan”	Boolean indicator to show if the mean intensity of this channel CC over the area of the object is the highest (=dominant), compared to all other channels).
30/ 41/ …	channel_#[CC]_outgrowth_mean_intensity	273.26551765	0 - 65535, or ”” (empty)	Mean intensity of this channel CC over the area of the outgrowth for the detected object. Is empty ““ if no outgrowth was set (max. outgrowth was set to 0 pixels).
31/ 42/ …	channel_#[CC]_outgrowth_stddev_intensity	203.15327249	0 - 65535, or ”” (empty)	Standard deviation of the intensity of this channel CC over the area of the outgrowth for the detected object. Is empty ““ if no outgrowth was set (max. outgrowth was set to 0 pixels).
32/ 43/ …	channel_#[CC]_outgrowth_intensity_dominant	“True”	“True”, or ”False”, or ”nan”	Boolean indicator to show if the mean intensity of this channel CC over the area of the outgrowth for the detected object is the highest (=dominant), compared to all other channels. Is “False“ if no outgrowth is set (max. outgrowth is 0 pixels).
33/ 44/ …	channel_#[CC]_object_incl_outgrowth_mean_intensity	273.115521066	0 - 65535	Mean intensity of this channel CC over the combined area of the detected object and outgrowth.
34/ 45/ …	channel_#[CC]_object_incl_outgrowth_stddev_intensity	174.47822028	0 - 65535	Standard deviation of intensity of this channel CC over the combined area of the detected object and outgrowth
35/ 46/ …	channel_#[CC]_object_incl_outgrowth_intensity_dominant	“True”	“True”, or ”False”, or ”nan”	Boolean indicator to show if the mean intensity of this channel CC over the combined area of the detected object and outgrowth is the highest (=dominant), compared to all other channels.

Info
Please note: The parameters marked with an asterisk (*) are calculated using https://scikit-image.org/.

Error information

More information about errors can be found in the Application Error Documentation.

Contact

If you have any questions about IKOSA AI and the applications you can create with it, please refer to Algorithm Application Training with IKOSA AI section in the Knowledge Base.

You can also always contact our team at support@kmlvisionsupport@ikosa.comai for any further clarifications.

Feel free to book a 30-minute meeting to speak with us about IKOSA and the apps!

https://calendly.com/kmlvisionkolaido/book-the-ikosa-platform-demo

Versions Compared

Old Version 1

New Version Current

Key

Table of contents

IKOSA AI Instance Segmentation App for Multichannel Images

Your IKOSA AI application was created using our training for instance segmentation algorithms models for multichannel images. All information presented in this document is applicable to your trained application.

Application description

Input data requirements

Input image(s)

Input parameter(s)

Description of output files and their content

Files

Content

results.csv

results_<xx>_<class-name>.csv

Error information

Contact

Page Comparison

Versions Compared

Old Version 1

New Version Current

Key

Table of contents

IKOSA AI Instance Segmentation App for Multichannel Images

Your IKOSA AI application was created using our training for instance segmentation algorithms models for multichannel images. All information presented in this document is applicable to your trained application.

Application description

Input data requirements

Input image(s)

Input parameter(s)

Description of output files and their content

Files

Content

results.csv

results_<xx>_<class-name>.csv

Error information

Contact