Quantcast

Digital Pathology – why you need it and how to choose the best camera for it

Yup, we really are in the digital age…even the pathology is digital. Facebook, Instagram, Tumblr. It has never been easier to take pictures and share them. The digital revolution is upon us and nothing is safe, even your pathology samples. Digitizing your pathology samples can help you better organize and manage pathology for later. And is essential for telepathology (pathology over long distances). But if you want to digitize your pathology you need to consider your camera equipment carefully.

 

The Case for Digitizing Your Pathology

Here is the deal: research, good research, and most cutting edge research involves collaborations between specialties. And the crux of good collaborations is good communication. However, for a long time our ability to communicate about pathological samples has lagged other disciplines. When you perform just about any other experiment – gel electrophoresis, mass spectrometry, PCR, etc. – you record the resulting data by taking a picture or printing a graph. You do this so you can easily share our data with others and yourself later. However, good data recording and tracking is rarely done in pathology. Instead, if someone wants to share a pathology slide with a collaborator you are likely to see them do this: 1) Re-sort through a zillion boxes of glass slides to find “the one” again; 2) spend hours trying to get the microscope setup the same; 3) dragging said collaborator to said microscope; 4) if lucky, use a teaching microscope to show the collaborator what they are talking about. If they are unlucky, frantically pantomime or sketch wildly on a piece of paper what they saw, as their collaborator stares befuddled into the eyepiece; and 5) kill the better part of an afternoon trying to share a pathology slide. (Sadly, I know this is how it is done because I have done it.) For a long time pathology sharing was done this way because cameras were not up to the job of capturing the subtleties of tissue staining. However, this is no longer the case! So take advantage of the digital age you live in, and improve how you manage and share your pathology. Invest in a good pathology camera and digitize your slides. It is important that your camera accurately and reliably captures your pathology slides, so be sure to consider the following:

Consideration 1: Color

True color and image representation in pathology is critically important. After all, someone’s diagnoses and life just might depend on it, so make sure your camera can capture the myriad of pathology stains. In general a camera filters color wavelengths one of two ways before they reach the camera’s sensor: mosaicing filters or three different pixel sensors for each RGB color. Whatever your choice though, the camera should be calibrated to human perception using color charts. Also when optimizing for color be mindful of your display monitor(s) quality and consistency.

Consideration 2: Noise

Digital pathology requires high standards and that means low noise. All cameras have noise. Camera noise comes from two main sources: fixed pattern noise or temporal noise. Fixed-pattern noise is due to pixel-to-pixel variability, which can be minimized by using high-quality components and calibration. Temporal noise happens during image capture and includes 1) ‘shot noise’, which can be overcome with software post-processing algorithms; and 2) electronic noise, which is influenced by the camera’s component quality and the nature of the analog-to-digital conversion.

Consideration 3: Dynamic Range/Sensitivity

Ever take a picture at sunset only to be disappointed at your picture’s quality? Well that is because while thehuman eye is really good a lowlight detection, cameras struggle to keep up. But it is crucial for digital pathology, especially fluorescence, that your camera works well in low-light AND high-intensity images – i.e. that it has a good dynamic range. Dynamic range is affected by two main things: the camera’s circuitry and pixel size.Logarithmic response cameras have wider dynamic ranges than linear response cameras. However linear response cameras have less noise. Leaving camera makers (and you!) with the hard choice: wide range or low noise? Smaller pixels increase spatial resolution but they also lower the number of photons that can hit the image sensor, which limits dynamic range. Therefore larger pixels can increase sensitivity. In addition, sensitivity is dictated by the camera’s physical components, including the size of the aperture and the magnifying optics – as the higher the magnification the darker and smaller your view.

Consideration 4: High Spatial Resolution

The camera’s resolution dictates how well (or if) small details can be captured. In certain pathology stains high resolution is essential, say if you are trying to identify cell types or make statements about cellular organization. Resolution is component of pixel pitch or individual pixel sensors’ size.

Consideration: Large Optical Sensor

How much of the specimen that can be seen is all down to the size of the camera’s sensor. Often for digital pathology it is desirable that the camera captures the same field of vision as would normally be seen through the microscope’s eyepiece. To do this your camera usually needs to have a medium- or large-format sensor.

Consideration: Frame rate

A fast frame rate improves your scanning process time. However fast frame rates don’t just happen. All of the above conditions must be able to act quickly, including the image sensor. Additionally, the memory storage and data transfer must be fast. USB 3.0 is ideal for this, is available on almost every computer and allows transfer of almost 4 Gbps.

Consideration: Software

If you are planning to do telepathology you need to seriously think about your software. You need to be able to do all the normal stuff such as import, stitch, compress, annotate and measure. But it also needs to be easy to use for all uses. Has going digital revolutionized your pathology? Tell us your experiences of digital pathology in the comments below.

1 Comment

  1. giangluong on October 30, 2014 at 8:40 am

    great article

Leave a Comment





This site uses Akismet to reduce spam. Learn how your comment data is processed.