Guide to Selecting an In Vivo Optical Imaging System

What are the key questions you should be asking to prioritize capabilities based upon your needs?

By

In vivo optical imaging is a powerful technology that allows researchers to monitor biological processes taking place in living mammals — non-invasively and in real time. By capturing light emitted from within living organisms, an optical imaging system provides a “window” into the organism and makes possible the real-time tracking of biological activity at the molecular level.

There are two major types of optical imaging reporters: bioluminescent and fluorescent. In vivo bioluminescent imaging utilizes luciferase — the enzyme that makes certain insects, jellyfish, and bacteria glow. The luciferase gene is incorporated into cells, microorganisms, and animals and, when active, causes a reaction that emits light. In vivo fluorescent imaging utilizes a fluorescent reporter — a molecule that emits a photon when excited by a particular wavelength of light. Fluorescent reporters can be fluorescent proteins, dyes, or nanoparticles.
 
By measuring and analyzing the light emission, in vivo optical imaging systems can help researchers monitor cellular or genetic activity and use the results to track gene expression, the spread of a disease, or the effect of a new drug candidate.

General imaging conditions

There are multiple optical imaging systems available on the market today, making it difficult to determine which optical imaging system is best suited to your needs. With all of the different technologies available, examining specification sheets can be confusing and may not be the best way to compare systems. There are some key questions you should be asking and help you determine how to prioritize capabilities based upon your needs.
 

Key instrument considerations

CCD Camera
The CCD camera is the heart of an optical imaging system. Performance variables to evaluate are: camera sensor, CCD chip size, pixel count, cooling temperature, spectral sensitivity, quantum efficiency, sensor readout noise, dark current, dynamic range, signal to noise ratio, spatial resolution, and frame rate. All of these are key considerations when selecting an imaging system since these specifications will have a dramatic effect on sensitivity, especially for bioluminescent applications. A back thinned, back-illuminated CCD is designed to increase the absorption of photons by the detector which results in a much higher quantum efficiency of greater than 90% vs. less than 50%. The CCD chip size determines the field-of-view and the amount of photons the CCD can collect (larger chip = larger field of view and more photons). The number of pixels will affect resolution and dynamic range; smaller pixels allow for a higher resolution but lower dynamic range. Binning combines pixel charges mathematically, thus increasing sensitivity but decreasing resolution. The dynamic range is a measurement of the minimum and maximum intensities that can be simultaneously detected, correlates to the numbers of bits required to digitize the signal, and determines the charge capacity (the number of counts) a pixel can hold before saturation. Sixteen-bit digitizers are recommended. Dark current is thermal noise generated by the camera in absence of light. Cooling the operating temperature will significantly reduce this background noise. To guarantee proper cooling of the sensor, a sealed vacuum is essential. Readout noise generated by the camera electronics such as the output amplifier is the limiting factor for low light images. CCDs only have one amplifier; cooling the CCD will slow the transfer rate and increase charge transfer efficiency, reducing readout noise.


Figure 1: Spectral Unmixing of multiple reporters and autofluorescence.
Raw multi-spectral fluorescence images in top panel, unmixed
data in lower panel.

Optics
In conventional photography, the lens often separates a good camera from a great camera. This also holds true for optical imaging. Lenses are important for maximum light collection, but it is also important to consider the flexibility of the lens. A system that has a lens capable of multiple fields of view will allow the researcher to balance sensitivity and throughput, depending on the application.

Filters
For fluorescent imaging, the excitation light source can be a filtered broadband Tungsten halogen lamp, a tunable laser, or several laser diodes of specific wavelengths. Fiber optics can be used to illuminate the entire subject or to focus the excitation light to a point source. Typically the entire subject is illuminated in the reflectance mode(known as Epi illumination) where the excitation light and the detector are on the same side of the subject. In the transillumination imaging mode, a point source is located on the opposite side of the subject from the detector. Upon integration of data taken at several different transillumination points, transillumination fluorescence imaging can provide sensitive detection, accurate quantification, and signal depth location of deep tissue sources.

There are numerous fluorescent markers available at a variety of excitation and emission wavelengths, and it is important to be able to select appropriate filters for different markers. Check to see what types of filters are available for the imaging system and how easy it is to switch between them. Narrow band excitation and emission filters can be used to detect and separate multiple fluorescent reporters and/or to minimize the detection of tissue autofluorescence using a technique know as spectral unmixing (see Figure 1).

Other hardware

Other system features related to subject handling can make an instrument easier to use and less stressful for the subject.

  • A fully motorized system secures reproducible data — motor-controlled stage, filter wheels, lens position and f-stop
  • Integrated gas anesthesia system — providing a delivery system, knockdown box, and a manifold in the chamber simplifies use of gas anesthesia for animal sedation
  • Heated sample stage — prevents hypothermia during anesthesia
  • Laser alignment grid — allows subjects to be conveniently and consistently aligned in the field of view
  • Anesthesia/cardio monitoring system — non-invasively monitors animal well being
  • Isolation box — to isolate infectious diseases
  • Catheter port — for catheters, ECG leads, etc.
  • In vitro multi-well plate imaging capabilities — allows for convenient correlation of in vitro cell based assays with in vivo live animal studies 

Key software considerations

Software is the link between the user and the optical imaging system. It is imperative that the software is easy to use while providing a comprehensive set of analysis tools. Be sure to ask about training, how long it takes to get up to speed using the system, and if a dedicated operator is required.


Figure 2: 3D Fluorescence Imaging Tomography Reconstruction
A series of images is taken at different transillumination points to determine
signal location and intensity. The signal source is then reconstructed, illustrating depth and intensity
information. The example above shows co-registration of a CT-scan from the SkyScan imaging system.

Quantification

Obtaining quantitative data is essential for in vivo imaging. Some imaging systems use a relative light unit (RLU) for their quantification. If this is the case, it is difficult to compare data from experiments performed on the same instrument, and impossible to compare data from different instruments. The only way to achieve high-quality data is to use absolute quantification in physical units. One way to perform absolute quantification is to use the unit photons/cm2/steradinew12an. This unit accounts for all camera settings such as the location of the source within imaging chamber, imaging time, and other parameters. Only by using imaging standards and calibration tools is absolute quantification possible. Understanding how calibration and quantification is performed, and whether it is “absolute quantification” is essential to obtain high-quality, reproducible data.

Software features

There are some key features that make software easy to use. Most imaging system software packages have a similar set of functions; however, the following features set some systems apart:

  • Complete instrument control: all functions of the imaging system should be under control of the software (e.g., camera settings, stage height, and filter selection should all be user-selectable from a computer control panel)
  • Data analysis: automatic creation of a “region of interest” (ROI) delivers reproducible data
  • Absolute quantification
  • Automatic quantification: automatically analyze and quantify a series of images
  • Programmed imaging sessions: investigators can run preset imaging sequences; this helps in kinetic and other time course experiments
  • Data management: multiple users compiling large amounts of image data can generate a database of enormous proportions in little time. The task of managing the data is an important one. Choose software that provides a basic data management system for organizing and referencing image data, and allows users fast and easy access to their data.
  • Software compatible for co-registration with other imaging modalities such as MRI, CT and PET, or SPECT
  • GLP compatibility

Throughput considerations

Imaging systems can image one or more subjects at a time, depending on the system and field of view. Researchers doing large-scale studies would want to ensure that more than one subject could be placed in the imaging chamber.

Along with the physical size of the imaging chamber and selectable fields of view, it is important to consider the image acquisition time. Some imaging technologies result in long imaging times (up to 30–60 minutes for one subject) and may not yield accurate bioluminescence data, where luciferin kinetics must be taken into account. Imaging systems that can capture a view of the whole subject at one time will yield lower acquisition times and reduce the impact of luciferin kinetics.

Advanced instrumentation considerations

3D Imaging Tomography
All of the imaging systems available on the market today can perform 2D optical imaging. However, there have been advances in the technology in the past few years, allowing researchers to reveal 3D tomographic information. The following describes two approaches currently possible using a 3D system.

Single-view Tomography
One way to perform 3D optical imaging is to generate single-view tomographic reconstructions of mouse images. With such as system, a laser scans the surface of the mouse and creates a surface “map” of the animal. The system then performs a “spectral analysis.” Due to the light transmission properties of tissue, the imaging system can determine the depth of a signal by taking multiple images of the same subject using different filters. This allows investigators to reconstruct a whole mouse and determine location and depth of a source signal. This feature is the result of both hardware and software-based technologies. For an example, see Figure 2.

Multi-view Tomography
Another approach to 3D imaging is to reconstruct a full 3D tomographic image of a single mouse by taking images at different angles. In the imaging chamber of this type of system, the subject rotates around a movable mirror and up to eight images of the animal are taken. These images allow for a more precise 3D reconstruction and in-depth information pertaining to source signal and intensity — 3D diffuse tomography reconstruction.

Fluorescence Imaging
Fluorescence is a powerful tool with a broad set of applications. It is one of the best ways to detect a target of interest when doing microscopy. When using fluorescence for in vivo optical imaging, an auto-fluorescent background can make data collection difficult. Fluorescence requires the use of a light source to excite the reporter. Unfortunately, the skin of the subject will auto-fluoresce, making it difficult to detect the target signal above the background noise. Be sure to examine how each imaging system removes and/or subtracts auto-fluorescent background.

Two possible methods to subtract background noise are: 1) to use background subtraction filters and image math in order to improve signal-to-background ratio and 2) separating reporter signal from tissue autofluorescence by applying spectral unmixing algorithms (software) to multi-spectral fluorescence images (see Figure 1).

References and publications

Be sure to ask for a bibliography list from your optical imaging vendor. Ensure that the imaging system has been used by other researchers in your field and that there are a number of publications in peer-reviewed journals.

Conclusion

Selecting an optical imaging system can be challenging; reviewing the long list of specifications can be a daunting task. Knowing which questions are most important — CCD chip size, cooling temperature, lens quality, quantification, software features, throughput, 3D capability, and references — will help you make the best decision for your needs.

Alexandra De Lille, DVM, Ph.D., is director of technical applications, Caliper Life Sciences, alexandra.delille@caliperls.com. Caliper Life Sciences, 68 Elm Street, Hopkinton, MA 01748; www.caliperls.com.

Categories: Laboratory Technology

Published In

Evolution of the Laboratory Magazine Issue Cover
Evolution of the Laboratory

Published: November 1, 2007

Cover Story