Start Protocol

1. switch on the computer
2. on the NanoZoomer main switch
3. launch the NDP.scan software

Shutdown Protocol

1. exit NDP.scan
2. switch off the Nanozoomer
3. switch off the computer
4. if the lamp is off (no green LED above the on/off button), let the ventilator cool down the lamp unit at least 10 minutes before switch of the main switch of the lamp unit

Documentation

The following schema sketches out the light path of the NanoZoomer. This is a normal fluorescent setup with the particularity of the Camera. This RGB camera always acquires 3x8bit -> 24bit RGB images, even if we scan one single fluorescence channel.

 

NDP.Scan (Acquisition Software)

The following paragraphs represent a collection of tips, tricks and additional explanations you might not find in the official manuals.

Illumination profile types and focus-channel

When configuring a profile the top panels select a subset of parameters (the buttons at the top: Region, Scanning, Output). When configuring the Region, the illumination panel allows to between the following options:

	Brightfield	Brightfield+Fluorescence	Fluorescence
illumination icon
channels available for scanning	brightfield	brightfield, DAPI, FITC, Cy3, Cy5
pre-scan image type	transmitted light	scattered light
focus channel choices	brightfield		DAPI, FITC, Cy3 or Cy5

Things to know about the automatic focus

The following facts should be taken into account when defining the focus points:

• A flat, linear plane is fitted to the focus points. This does not compensate for big local variations (as compared to a polynomial surface).
• To define such a plane at least 3 points are needed.
• If there are more than 4 focus points, the algorithm tries to remove positional outliers among the focus points and uses the remaining points to fit the focus plane. If after the outlier removal, less than 3 points remain, the region will not be scanned.
• Acquiring focus points is time consuming.

Scenarios:

• Thin and opaque samples are generally easy and reliable to focus -> few sample points needed
• Thick and transparent samples can cause larger variance in the focus-position (z-position) of the focus points and result into point rejections -> more sample points needed (try not to put too many for time considerations)

Channel acquisition order

In certain cases the order of acquisition of several channels in a row might be important. For instance if one fluoroscope bleaches easily, you might want to choose this one first in order to minimize the exposure time until the actual acquisition of the channel.

When editing the acquisition Profile, the Filter Set panel (see the screenshot bellow) lets you select the channels for acquisition from the drop-down menu.

Acquisition order:

1. Brightfield (if selected with the “Also brightfield” check box or via illumination profile).
2. Fluorescent channels in the order the ticks were set!
   You can see the selection order in the ‘Settings For … Filter Set‘ panel in the drop-down menu; top to bottom – first to last selection

A note on the color balance

Here is some additional information about the color-balance. This can be thought of like a liner scaling or gain. The figure bellow illustrates the input-output relation for different color balance values, translating the camera output in image gray-scale values.

If the signal is rather weak, and a high value for the color-balance is chosen (e.g. 4x or 8x), then the result might look like the example bellow when using the NDP.viewer. The  low-intensity signal then looks “patchy” and shows abrupt changes, that do not correspond the the known morphology of the scanned sample.

While choosing a high color-balance-value increases the contrast of the images, keep in mind that noise (photon noise, camera, etc.) will also become more visible. We therefore suggest to use values between 1x and 4x, and not higher. If the signal is weak, try increasing the exposure.

Note: In case you tested your staining on our epi-fluo microscope or even on the confocal, you have to take into account, that color camera of the NanoZoomer does not have the same sensitivity and that a mercury lamp does not have the same excitation power as a laser. Both of these technological factors, will translate in a weaker signal when acquiring fluorescence with the NanoZoomer.

 

ImageJ: quality control of your scans

Unfortunately the NDP.view doe not have a histogram feature. But we can use Fiji (is just ImageJ) instead. Open your ndpi-file:
Menu > Plugins > Bio-Formats > Bio-Formats Importer.
Then choose a low magnification image (the intensity distribution is roughly the same, and the loading time is much shorter), series-1 is usually the highest resolution (see pixels x-y), so choose something close to the macro image (see the screenshot bellow on the left).

Once the image is open,  you can plot the histogram:
Menu > Analyze > Histogram
In the above histogram (on the right) you can see that the maximum intensity value in your scan is 52. This means that you only use one fifth of the dynamic range of your detector (camera). In this case you might want to increase the exposure, to collect more photons from your specimen and get a stronger output signal.

ImageJ is also installed on the NanoZoomer computer, so can use it to check a couple of scans and adjust the acquisition settings accordingly.

 

NDP.View (visualization software)

One trick that is not in the manual:
ctrl+shift+F on a PC or cmd+shift+F on Mac renders the focus points visible.

 

NDP.Toolkit (file format utilities)

The NDP.Toolkit is a software that allows editing the ndpi-files and doing batch conversions on a folder full of files. It is installed on the NanoZoomer computer and on the cosmos workstation.

 

Manuals

NDP-Scan quick start (FR)

NPD-Scan Instruction Manual (ENG)

 

Virtual Slidebox

Virtual Slidebox Manual on Bioimage