The digital revolution of the past few years has generated a multitude of consumer digital cameras with capabilities once the province only of professional photographers. The neurosurgeon with one of these cameras and a working knowledge of computers can obtain spectacular intraoperative images far beyond the typically frustrating results obtained with a 35mm camera and film. The following basic guide to intraoperative digital imaging using today’s digital cameras and the operating microscope attempts to illustrate how excellent images of interesting cases can be achieved.
Anatomy of the Digital Camera
The heart of the digital camera is the CCD, or charged-coupled device. This electronic device is capable of transforming a light pattern into an electrical charge. That is, as photons of light strike a pixel (picture element), an electrical charge is created. This charge is proportional in magnitude to the intensity of the sensed light. The electrical charge then is transferred in sequence to an analog-to-digital converter; this outputs an electronic rendition of the image that can then be stored or displayed on a computer screen.
Important Camera Features
There are a number of excellent cameras available on the market, and the list grows longer each day with the addition of more sophisticated and higher resolution capabilities. Important features to consider when selecting a camera for intraoperative imaging include the CCD resolution; lens capabilities comprising optical zoom, internal zoom, macro capability, and a threaded lens; an LCD (liquid crystal display); and image storage and transfer features.
CCD resolution is critical for sharp, crisp output of the image to a monitor or for printing to slide or film. The more pixels present on a CCD chip, the higher the resolution. The multi-megapixel cameras with greater than two million pixels (non-interpolated) are preferable. Cameras of up to five megapixels now are available. In practical applications, however, this amount of resolution may not be needed. Depending on the final output (computer presentation or print), three megapixel cameras will usually suffice. However, if printing large images (greater than eight by 10 inches) is desired, then more megapixels are desirable.
Several lens features also are critical. Optical zoom-the ability of the lens to change focal lengths, thereby enlarging the subject image-should be at least 3X. Care must be taken in discerning whether a camera has optical zoom or digital zoom. Digital zoom refers to a camera’s ability to further enlarge the image by interpolating new pixels. The result is significantly inferior and has little usefulness for intraoperative imaging. Also, a camera with internal zoom is preferable so that the lens assembly will not move during focusing.
The ability to take macro images-extreme close-up shots, usually within one foot or less-is important for through-the-microscope shots as well as for framing individual pictures of CT scans or MRI films. The presence of threads on the camera’s lens allows significant flexibility in using alternative lens converters, filters, and coupling the camera to the microscope. A variety of accessories that screw onto these threads are available.
Most digital cameras offer a preview screen apart from the optical viewfinder. These small, color LCDs are essential for framing and composing intraoperative images because they not only provide a preview of what the camera will see before the picture is taken, they allow for fine adjustments during coupling of the camera to the microscope. In addition, they allow review of taken images to ensure the expected result. The optical viewfinder, although useful for most general photography, has no value during intraoperative imaging.
The method of image storage and transfer refers to how the camera stores and downloads its pictures to the computer or printer; different camera manufacturers offer different features that facilitaate storage. These include fixed storage within the camera, 3.5 inch floppy disks, removable memory cards and, in some newer cameras, recordable CDs or miniature hard drives such as the IBM Microdrive. For ease of use and maximum flexibility as well as affordability, removable storage media is preferred over cameras with internal memory and those with standard 3.5 inch floppy disks. The preferred types of storage media include SmartMedia, CompactFlash, and Memory Stick memory cards. Obviously, with greater memory capability, more images can be stored.
Downloading images to a computer can be accomplished via parallel, SCSI (small computer system interface), or serial cable, and in newer cameras via USB (universal serial bus) cable or FireWire (also known as IEEE1394). Although FireWire is the fastest method by far, the expense associated with this technology has allowed USB to become the industry standard. USB is preferable to older technologies such as serial connections due to the much faster communication with the computer, but it requires a USB equipped PC or Mac. Most cameras that offer USB also offer serial and/or parallel connection options. An alternative, if one uses an older serial-type camera, involves using removable CompactFlash media and a USB card reader to download to the computer.
Camera Recommendations
The ideal camera for intraoperative imaging should contain a high resolution megapixel CCD, optical zoom, internal zoom, macro capability, removable storage media, color LCD, and remote control.
There are a number of excellent consumer digital cameras on the market for less than $1,000 that meet these specifications. Just one year ago the selection included the Olympus C-3000Z, Epson PhotoPC 3100Z, Canon PowerShot G1, and the Nikon family of cameras, but the explosion of digital cameras on the market now renders discussion of individual cameras beyond the scope of this article. However, the Nikon family of cameras continues to be preferred by the author because these cameras have some of the best color purity and tonal balance, in addition to superb optics, resulting in superior image quality. In addition, the small threaded internal zoom lens allows for efficient coupling to a microscope eyepiece. The Nikon family includes the older 950 and 990 models as well as the newer 995. The 880 remains an excellent camera but requires an adapter in place of standard lens threads.
The author uses the Nikon 990 camera. This 3.34 megapixel camera can take images with a maximum size of 2038 by 1536 pixels. It has both full automatic and full manual settings. Its lens is a nine-element 3X optical zoom lens that contains 28mm threads and boasts the closest macro capability on the market. The camera uses CompactFlash cards and has a 1.8-inch LCD. It has USB and serial capabilities as well as a TV video output. Lastly, it can be controlled remotely with an optional wire remote control.
Affixing the Camera
Unlike 35mm SLR (single-lens-reflex) cameras, commercial digital cameras are equipped with lenses that are not removable. In order to photograph through the microscope, one must employ a configuration termed afocal coupling. Afocal coupling is a method of photography frequently used by amateur astronomers for imaging the planets; the camera’s lens is lined up with the eyepiece of the telescope or, as in our case, the microscope.
Photographing afocally can by done by simply holding the camera up to the eyepiece. The camera must be held as close as possible to the eyepiece with the image centered on the LCD view screen. Zoom in on the image to enlarge the field of view and minimize vignetting (when the image from the eyepiece does not fill the cameras field of view, resulting in a circular image that does not reach the corners of the picture frame).
Although this method generally works, it is a less than ideal setup. A more stable configuration involves threading an appropriate adapter to the camera lens, allowinng it to be mounted on the microscope eyepiece throughout the case. The specific thread size must be known for the type of camera used, and the adapters can be purchased from photography or astronomy vendors.
There has been significant interest in this type of configuration, and a plethora of adapters specific to a variety of digital cameras, including those with threadless lenses, currently are on the market. For the Nikon 990 the thread size is 28mm, and the adapters required are a 28-to-T thread step-up ring and an eyepiece projection adapter. The step-up ring is connected to the camera lens followed by the eyepiece projection adapter. The entire 
assembly is then secured to an unused eyepiece on the microscope via three thumb-screws on the eyepiece projection adapter. A remote control cable can be hooked up to the camera and secured near the microscope’s control arm, where the surgeon or assistant can reach it during the case. The entire setup is depicted in the photo shown right.
Storing Your Images
There are several standard file formats for storing digital images. On many cameras the user can set different levels of quality and resolution. The degree of resolution needed depends upon the final output desired. When taking images, using the highest resolution is recommended. Doing so will allow flexibility if one wishes to output the images for print, 35mm slide, or reproduction for publication at a later date.
However, the greater the degree of resolution, the larger the file that must be stored on the computer. In order to store images more compactly, various file compression formats are in use. The two most common are the TIFF (tagged image file format), which allows the original image to be reconstructed from the compressed image with no loss of information, and JPEG (joint photographic experts group), which selectively discards low contrast fine detail that is superimposed on higher contrast features. For onscreen viewing, such as PowerPoint presentations or Web pages, the smaller JPEG files are recommended, but the original files should be stored as TIFF images.
Storage becomes an issue as images accumulate. It is quite easy in the course of a few months to fill a 1GB hard drive with stored digital images. Storage solutions include purchasing separate hard drives, using storage media such as superdiscs, Zip disks and drives, and “burning” CDs. All of these solutions work; the challenging issue then becomes finding your images later.
One option is to catalog images so that they can be accessed by subject, keywords, and other search criteria. The author created a searchable database using FileMaker Pro that tracks operative cases and allows images to be accessed by patient last name on previously archived CDs. Recently, third-party image cataloging products have become available, and some manufacturers have begun to include them with a camera’s software. The author recently has begun exploring Apple iPhoto, a product that allows albums of images to be created and quickly and efficiently sorted by keywords and various search algorithms.
Improving Your Images with Photoshop
There are a number of ways to improve photos digitally. These include adjusting color balance, contrast, brightness, and sharpness, as well as removing and touching up elements in the image. Most cameras include some basic image editing software that performs at least some of these functions, while powerful image editing applications like Adobe Photoshop and Corel Photo-Paint available on the market integrate all of them.
Using Photoshop, one of the more powerful image editing tools, involves adjusting the histogram and levels of an image. The histogram is a graph of the image’s brightness. The X-axis represents the scale of brightness vallues (0 being black and 255 being white, with all the shades of gray in between), while the Y axis represents the number of pixels with that brightness in the image. The utility of the histogram is that it allows more precise alterations than brightness and contrast controls. By adjusting points on the histogram, the shadows, highlights, and mid-tones can be adjusted independently. As a start, adjust the black and white points on an image by sliding the triangles on the left to the first groups of pixels, and the triangle on the right to the first group of pixels on that side of the graph. This maneuver basically adjusts the contrast. A more powerful adjustment of the image’s brightness scale is available using the curves function.
Next, adjust the color levels (basically, the luminosity of the red, green and blue channels) by opening the levels tool and setting it so the individual color channels are adjusted independently. Adjust the settings until a better color balance is achieved. At this point, the image has enhanced contrast and better color balance.
Then the sharpness of the image can be adjusted using the sharpness tool, or preferably using the unsharp-mask tool in Photoshop or Photo-Paint. This tool analyzes the image’s contrast variations, in effect looking for edges and changing the degree of contrast between pixels in these edges. Using this command, an amount is set (how strongly the effect is applied), a radius (indicating how far from the edges to apply the change), and a threshold (telling the software what level of contrast variation will be affected). The best way to apply this tool is to backtrack from a high radius until distortions are lost and the appropriate amount of fine detail enhancement is reached. Then adjust the amount between 50-150 (trial and error works best) and change the threshold until only the areas you are targeting are affected. At this point save the image in TIFF file format and store it.
With these minor changes, less than optimal images can be improved dramatically.
Intraoperative imaging through the microscope need not be restricted to 35mm film. The digital revolution has placed in the hands of consumers a variety of equipment with capabilities of a professional level. As this article has briefly outlined, using a relatively simple setup and modest equipment, spectacular images can be recorded and stored for future use with a minimum of effort.
Carlos A. David, MD, is director of neurovascular and skull base surgery at the Lahey Clinic Medical Center, Burlington, Mass.
For this Technology Issue of the Bulletin the author has significantly updated his article from its first appearance in the Bulletin‘s Summer 2001 issue.
first group of pixels on that side of the graph. This maneuver basically adjusts the contrast. A more powerful adjustment of the image’s brightness scale is available using the curves function.