The Daily Scan

chalk tip #40: Pixels and Resolution

For this edition of Chalk Tips, we’ll discuss the pixel. Small in size, mighty in stature, the pixel provides and enormous amount of information about your image.

Like the atom to the molecule, the pixel (abbreviated from “picture element”) is the smallest component of an image. You’ll often hear of pixel counts in digital cameras, like “12 million pixels”. Each one of those tiny elements is a physical location on a silicon substrate capable of storing intensity information. And since each pixel is of a known and fixed size, it is possible to obtain distance, size and intensity measurements for objects and features in your images. Which leads us to an important question- what is meant by resolution?

Resolution is described two ways- what our eyes can distinguish in terms of intensity and what they can distinguish in terms of distance. Each form bears its own resolution, so here goes:

First off, intensity resolution- our eyes can detect approximately 30 shades of a given color’s intensity in an image. Some eyes are naturally more sensitive, but on average we see thirty. On the other hand, the most basic digital camera can detect and resolve at least 256 levels of intensity! So, it is possible to enhance detail we can’t detect on our own, and make judgments about images by merely enhancing the intensity resolution of the image in question. It’s a great way to see detail that may be close to background (like what you can often expect to see in fluorescent images) and a great way to highlight and explore detail across a stained slide.

Alright, now we can talk about what the pixel can provide in terms of area and distance measurements- what we traditionally talk about when describing the features we see in an image. Since each pixel has a known size, typically measured in microns, it is really easy to calibrate a known magnification to a stored spatial measurement. If you know how many pixels make up a known distance, it’s a simple matter to create distance and area measurements.  What’s best is that these measurements are highly repeatable. Once calibrated, imaging software makes measurements using the same repeatable steps and calibrations, time after time. Which in turn improves the reliability and accuracy of your measurements.

In a future Chalk Tip, we’ll explore intensity resolution a little more, through the introduction of a new concept- bit depth. Stay tuned!

(Note: Will Casavan is Aperio's Product Manager for Image Analysis, and will be joining me in posting to the Daily Scan...  Ole Eichhorn)

chalk tip #39: Aperio FDA clearances

Hi everyone...  it's been a while since my last chalk tip and I'm getting back into the habit :)

I get asked a lot about Aperio's FDA clearances...  people want to understand exactly which applications are FDA cleared, and all the details.  The detailed information on each filing is posted on our website, including links to the corresponding FDA premarket notification database entries.  There are also links to the press releases announcing each clearance.

Here is a list of our current FDA clearances, linked to the details:

• HER2 Image Analysis - K071128
• Manual Read of HER2 Digital Slides - K071671
• ER/PR Image Analysis - K073677
• IHC PR Breast Tissue Manual Read of Digital Slides - K080254
• HER2 Image Analysis for Breast Tissue – K080564

It is important to note that Aperio products are FDA-cleared for specific clinical applications, and are intended for research use for other applications.

chalk tip #38: linking images and annotation regions

Greetings!  Today we will demystify one of ImageScope's coolest and yet least well-known features. That's right, we're going to talk about annotation links.  (dum dum dum)

Did you know?  With annotation links you can create an ordered sequence of digital slide regions for later display.  Perfect for tumor boards, conferences, many education applications, and even pathologist-to-clinician communications.  The linked regions can all be within one slide, or they can span any number of slides.

Pull up a chair, I'm going to give you a couple of examples.  In the first example we'll annotate several regions within an H&E slide and link them into an ordered sequence.

Step 1: Launch ImageScope, and open the digital slide (or simply select the slide from within Spectrum).  Here's our example slide:

 

Step 2: Create a few annotation regions.  First a rectangle labelled "interesting area", then a free-form region (with the Pen tool) labelled "free-form region of interest", and finally an arrow labelled "wow, this is amazing".  This is what the image looks like with the three annotation regions:

 

Step 3: Link the regions in a sequence.  Open the Link Manager with View | Annotation Link Manager or by clicking the icon on the toolbar, it looks like this:

 

The three annotations are listed at left.  Next create links by dragging and dropping an annotation region on top of another one.  First click on "interesting area" and drag it over "free-form region of interest", then drop it.  This creates a link from the rectange to the free-form region.  Next click on "free-form region of interest" and drag it over "wow, this is amazing", then drop it.  This creates a second link from the free-form region to the arrow. 

Clicking on each region in the Link Manager displays the links from (blue) and to (red) the region:

 

 

 

That's it!  We've created three regions and linked them in a sequence.

To use these links, close the Link Manager and use the Next Link and Previous Link icons on the toolbar to move from one annotation region to another.

Okay, that was cool, but do you have time for another example?  Now we're going to link four separate digital slides together in a sequence.  Ready...set...

Step1: Launch ImageScope, and open the digital slides (or simply select the slides from within Spectrum). Here's our example slides:

 

Step 2: Link the slides in a sequence.  Open the Link Manager with View | Annotation Link Manager or by clicking the  icon on the toolbar, it looks like this:

 

The four slides are listed at left.  Next create links by dragging and dropping a slide on top of another one.  Link the first slide to the second, the second to the third, and the third to the fourth.

Clicking on each slide in the Link Manager displays the links from (blue) and to (red) the slide:

 

 

 

 

That's it!  We've linked four slides in a sequence.

To use these links, close the Link Manager and use the Next Link  and Previous Link  icons on the toolbar to move from one slide to another.  You will notice this works even if all the slides have not been opened.  For example, close all the slides except the first one:

 

Now click Next Link on the toolbar and the second slide will be opened and displayed!

 

Similarly, click Next Link again and the third slide will be opened and displayed:

 

And finally click Next Link once more and the fourth slide will be opened:

 

Isn't that cool?  I thought you'd think so.

In addition to linking regions in one slide together, you can link a region in one slide to another region in a second slide.  This is great for preparing a display sequence of regions for presentation in a tumor board.  And you can also link regions to entire slides!  This is useful for annotating a gross image with links to slides made from tissue blocks cut from it.

Please try this yourself; ImageScope's annotation links are easier to create and use than they are to explain!

chalk tip #37: making stand-alone DVDs and thumb drives

Greetings y'all!

Today's tip is incredibly useful: did you know?  With Aperio software you can make stand-alone DVDs and thumb drives!  A stand-alone DVD or thumb drive contains one or more digital slides which are viewable on any computer without installing any software.

To do this, you need a software program called DVDGen (earlier versions were called WebGen, and work the same way).  DVDGen is part of Spectrum and also the KickStart product.

How about an example?  I thought you'd never ask...

Let's say I have four digital slides, and I want to make a stand-alone DVD from them:

  

Step 1: create a folder to hold your slides.  I'm calling mine DVD_100.

Step 2: copy the images into this folder.

Step 3: launch DVDGen.  The program displays a dialog which looks like this:

 
Navigate the folder tree on the left to select folder you created with the images.  Make sure "Make CWS from SVS" is checked.  Then click Generate Website.  You will see the following warning message:

This is a true statement!  It can take several minutes to convert each slide.  Next click OK.

DVDGen will display a running dialog as it converts each slide, like this:

  

When all slides have been converted, you will see the message "generation complete".  And a web browser window will be opened with the contents of the DVD, like this:

That's it, we're done!  Clicking any of the thumbnails launches an interactive web browser view of the slide, without installing any software!

My DVD_100 folder is a complete stand-alone way to view my digital slides.  The folder can be burned onto a DVD, or copied to a thumb drive, or posted to a website, or whatever you want!

* Bonus tip: this is also a great way to prepare for tumor boards...

chalk tip #36: using SecondSlide to share any images with anyone

Greetings, you're going to like today's tip, because with SecondSlide.com you can share any images with anyone.  And it's FREE. 

If you'd like click on the SecondSlide.com link to open a new window, and we can do this together.

1. On the SecondSlide home page, click "Sign Up" in the top right corner.  You'll see this:
   
    
    
   
   
2. Fill in the form with your name and email address, and click Sign Up.  You will see this form:
   
    
    
3. You will receive an email from SecondSlide welcoming you to the service.  This email includes a temporary password.  Enter your email address and the temporary password, and click User Login.  You will see this form:
   
    
    
4. Enter a new password (in both fields) and click Save.  You'll see this page:
   
    
   
5. Okay, you're signed up!  Now let's share some images.  Click on Upload a New Case.  You'll see this page:
    
   
    
6. Enter the email address of the person with whom you want to share images.  (You can send the images to more than one person, just separate their email addresses with a comma.)  Note that recipients don't have to have joined SecondSlide, you can share with anyone. 
   
   Use the check boxes to specify whether you allow recipients to modify the case information, and whether you allow them to re-share the case.  Then click Share Case.  You will see this page:
   
    
    
   
7. In the top of the page, you can specify Case Information.  You don't have to enter anything at all, but if you'd like enter as much information as necessary to describe the image(s) you are sending.  You can also enter Specimen Information if you'd like.
8. Next scroll the page down to the Digital Slide Information, which looks like this:
   
    
    
   
9. Here you identify the slides and images you wish to share.  There can be multiple images for each slide, as well as multiple slides in the case.  To specify the image(s) for the a slide, click the Images of Slide button.  You will see the following dialog:
   
    
    
   
10. Next select one or more images for the slide on your computer.  SecondSlide supports many different kinds of images, including SVS, JPEG, TIFF, JP2, etc.  The images can be large whole-slide images created by a scanner, or smaller photomicrographs made with a camera on a microscope.  After you have selected the image(s), click Open.  The Case Information page will look like this:
    
     
      
11. You will see one line for each image you selected (I selected three in the example above). 
12. You could add images for another slide to the case being shared by clicking the Add Digital Slide button, but in this example let's say we have only one slide.  Next click Add Case.  
13. The images you specified will each be uploaded to SecondSlide.  Depending on the size of the images, this could happen almost instantly, or could take a few minutes.  You will see progress indications as the slides are copied:
    
     
    
    When all the images have been uploaded, you'll see a "Case Successfully Shared" message.
14. That's it, you're done!  

Your recipient(s) will receive an email notification that a SecondSlide case has been shared with them.  If they have not already joined SecondSlide, they will be walked through the process, and then they'll be able to view the images you've shared with them.  

That's it, you can use SecondSlide.com to share any images with anyone, for FREE.  And all without installing any software!

chalk tip #35: setting image resolution

Greetings all...

Did you know?  In ImageScope, it is easy to set image resolution.  I'm going to show you *how* easy, and I'll talk a little bit about what image resolution is, too...

First, what *is* image resolution?  Every image represents something with little dots called pixels.  Image resolution is simply the size which each pixel represents.  For example, suppose you had a picture which was 5" x 7" in dimension.  Maybe it consists of 1,500 x 2,100 pixels, so that each pixel represents 1/300th of an inch.  For this case the resolution of the image is 1/300th of an inch per pixel.  (This is sometimes called 300dpi, "dots per inch", and is a typical resolution for photographs.)  As another example, consider a digital slide with a sample area of 20mm x15mm, which was scanned with a 20X objective.  This would result in an image with 40,000 x 30,000 pixels, where each pixel represents 1/2 micron.  The resolution of this image would be .5 mpp (microns per pixel).

Okay, so how is the image resolution for an image known?  For many images including most digital slides the resolution is stored somewhere in the image, as an inherent property.  For example Aperio ScanScope instruments create SVS (TIFF) files which always contain the resolution.  When the resolution is known, it can be used by viewing tools such as ImageScope to display rulers, grids, annotations etc. using the correct dimensions.

Here's an example, a digital slide where the resolution is known:

 

Note that since the resolution is known, the ruler can be displayed with micron units and the magnification can be displayed as 40X.  Selecting Image | Resolution displays the following dialog box:

 

This shows the current image resolution is .23 microns per pixel, corresponding to an apparent magnification of 40X.

Okay, but what if the resolution is not known?  Consider the following JPEG image:

  

Note that now since the resolution is unknown, the ruler is displayed simply in "pixels" and magnification can only be displayed as a percentage.  Selecting Image | Resolution displays the following dialog box:

 

At this point, if the resolution is known, it can simply be entered!  For example if you knew this was a 40X image, you could enter 40 as the apparent magnification, or if you knew the actual resolution in microns per pixel you could enter it.  After clicking OK, ImageScope will use the resolution you have specified for rulers, magnification, etc.

It often happens that an image for which the resolution is unknown contains one or more objects of known size.  In that case, you can draw a ruler to set the resolution.  For example, consider the following gross image:

  

The resolution of this image is not known, but the image contains an image of a ruler measured in centimeters.  We can use that to set the resolution.

First, draw an annotation ruler the length of the known object in the image:

  

You will note the length of the ruler is given simply in pixels, since the image resolution is not yet known.  Now select Image | Resolution to show this dialog box:

  

Note that "Compute from Currently Selected Annotation Ruler" is selected.  Enter the length of the object measured by the ruler, in this case 5cm = 50,000 microns, and then click OK.  That's it, the image resolution will be calculated automatically from the known object size!  Here's how that image will be displayed now:

  

Note that the annotation ruler's length is now known to be 50mm.

Please try it yourself, it is easier to use than to explain!

chalk tip #34: useful ImageScope keyboard shortcuts

Greetings all...  in today's tip I want to tell you about a few useful ImageScope keyboard shortcuts.

shortcut	action	toolbar icon
Control-O	Open Image
Control-F4	Close Image
Control-K	Keep Image Open	(none)
Control-A	Image Adjustments
Control-R	Rotate Image	(none)
Control-N	ANnotations Window
Control-G	Analysis AlGorithms Window	(none)
Control-D	Digital Slide Conferencing	(none)

That's it - you're welcome!

Oh, and if you want a complete list of all shortcuts, please see the ImageScope Quick Reference Guide (PDF).  It is just eight pages and tells you everything you need to know: opening images, the main window, toolbar icons, and *all* keyboard shortcuts!

chalk tip #33: bookmarking digital slide views

I'm back with another chalk tip!  (pauses to let applause subside :)

Today's tip is simple and useful: did you know, you can easily bookmark digital slide views?  Yes, indeed, any view of a digital slide can easily be saved as a URL bookmark.  Here's a digital slide we can use as an example:

By default the slide is displayed "zoomed out" to fit the entire slide in the window.  But let's say there's a particular view we want to bookmark, like this one:

  

Position the slide to the view you want, and then click the checkmark on the toolbar.  This opens a new window which has exactly the URL needed to bookmark this view, like this:

The following link can be used as a "bookmark" for WebScope: http://images2.aperio.com/H_E 40x.svs/view.apml?X=-0.0182238266781054&Y=-0.228077872826575&zoom=25 Options for display of the bookmarked slide view: Display navigation window (thumbnail) Display toolbar Display menubar Display annotations The following link can be used as a "bookmark" for ImageScope: http://images2.aperio.com/H_E 40x.svs/is.sis?X=-0.0182238266781054&Y=-0.228077872826575&zoom=25&chost=images2.aperio.com

Go ahead and click the link, it works!

Bonus tip: this window also has a link which bookmarks the slide view for the Aperio ImageScope viewer.  Yes that's right, you can bookmark any view of any digital slide with a single URL.  How cool is that?

chalk tip #32: fun with the magnifier

For today's tip, we're going to have some fun with the magnifier.

The magnifier is a cool feature of Aperio's ImageScope viewer, a small window drawn over the main window which "magnifies" a region of the image being viewed by 2X:

 
The magnifier always shows the region of the image under the mouse cursor.  As you move the mouse around over the image the magnifier image is updated to match.

The magnifier is toggled on/off with the View | Magnifier command, or by clicking the toolbar icon shown at right.

By default the magnifier is positioned at the lower right of an image, as shown above.  However, the magnifier window may be dragged anywhere you like, like this:

 
If you try this, you'll notice that as you're dragging the magnifier, it magnifies the part of the image it is floating over.  So you can move the magnifier around over the image, just like a magnifying glass. 

Finally, the magnifier window can be resized.  Just position the cursor over the lower right corner, and it will turn into a crosshair; then you can drag the corner to resize the magnifier:

 
After resizing, you can again move the cursor around over the image to magnify different parts of the image.  Or you can drag the resized window over the image to magnify the part of the image it is floating over.

The magnifier is a useful and efficient way to inspect digital slides...  please try it!

chalk tip #31: combining multiple images together into one big one

Greetings!  Did you know?  With Aperio's software you can easily combine multiple images together into one big one.

This capability makes use of a file format supported by all Aperio software called an Aperio Compound Image, or ACI.  An ACI is simply a little text file which contains instructions for combining other images together.  ACI files are sometimes made programmatically, but they can also be made "by hand".

Each line in an ACI file specifies the name of an image file, and gives the X and Y coordinates where the image is to be located in a new "compound image".  (By convention, X and Y are measured from the upper left corner, with X going across the image to the right, and Y going down the image to the bottom.)  Each line of an ACI file looks like this:

<image_filename> <Y-offset> <X-offset>

Let's take a simple example.  Suppose we have two images which we want to combine side-by side; here they are shown in the Aperio ImageScope viewer:

madison.svs 58,984 x 35,659

katherine.svs 69,608 x 37,675

(The numbers below each image give their dimensions in pixels.)

To combine these images, let's create a file called paul.aci, which has the following contents:

madison.svs     0     0
katherine.svs   0     58984

Notice that madison.svs is positioned at (0,0), the upper left corner, and katherine.svs is positioned at (58984,0), so that the images will be side-by-side in the compound image.

Now we can open the file paul.aci in the ImageScope viewer, and it looks like this:

  
paul.aci
128,592 x 35,659

Pretty cool!  We can pan and zoom and otherwise view the compound image just as if it was a single file. 

You might notice that katherine.svs is actually a little taller than madison.svs; the default logic is to "trim" the edges.  However this might not be what we want.  Let's change paul.aci just a little, adding a line which says "edgetrim=no", like this:

madison.svs     0     0
katherine.svs   0     58984
edgetrim=no

Now when we open paul.aci in ImageScope, it looks like this:

   
  paul.aci
128,592 x 37,675

No edges were trimmed, and there is now a "gap" at the lower left which is filled with gray.

Let's try a more complicated example, suppose we have two more images (again shown in ImageScope): 

hannah.svs 72,503  x 43,361

samantha.svs 55,848 x 35,307

Let's combine all four of the images above.  We create a new text file named daniel.aci, with the following contents:

madison.svs     0     0
katherine.svs   0     58984
hannah.svs      37675 0
samantha.svs    37675 72503
edgetrim=no

Now if we open the file daniel.aci in the ImageScope viewer, it looks like this: 

daniel.aci
128,592 x 81.036

Magic, huh?  We can pan and zoom through all four images as if they were a single image.  There is no limit to the number of images you can combine in this fashion.  In fact, we used this capability a little while ago to combine 225 digital slides together into an image which was more than one terapixel in size!

There is one more thing I want to mention: you might want to turn a compound image into a single image file.  You can do this with ImageScope!  Just use the Image Extract tool select the entire image, then extract it into a new SVS file.  That might be a good subject for another tip in the future :)

Aperio Compound Images are a powerful capability - please try it!