digital cameras

CAMERAS: Are More MEGAPIXELS Better?

Keep a critical eye on features, optics, software power, image presentation and (yes) megapixels when selecting a digital fundus camera.

DAVID G. KIRSCHEN, O.D., Ph.D., Brea, Calif.

If you're shopping for a digital fundus camera, you might quickly conclude that the number of pixels in its images rates extremely high in determining camera quality. After all, many digital fundus camera manufacturers — following the lead of consumer digital cameras — are engaged in megapixel (MP) one-upmanship.

But, I can't think of any reason a general O.D. would need more than five million pixels (five MPs) in an image. And what the most exhaustive retinal subspecialist would do with more than 10 MPs is a mystery to me (especially given a computer screen's optimal resolution is roughly four MPs). And yet fundus cameras exist that offer 15, 20 and 24 MP images.

So, do additional pixels enhance our ability to effectively diagnose and treat retinal disease? To answer this question, I'll discuss the basics of digital imaging.

Picture elements

Images on a computer are composed of tiny dots called pixels (a word derived from the term PICture ELement) arranged horizontally and vertically on the screen. In stored image files, the amount of pixels that run horizontally or vertically determine the resolution.

Pixel count is expressed in MP and is calculated the same way you would calculate area: It is the product of multiplying the number of horizontal pixels by the number of vertical pixels. Therefore, a camera that has 2,048 horizontal pixels and 1,536 vertical pixels has a resolution of 3,145,728 pixels (2,048 × 1,536) or 3 MP.

When we consider MP in these terms, we understand that it takes large differences in MP to translate into significant pixel differences. For example, a 3 MP camera has 2,048 pixels horizontally, and a 14 MP camera has 4,500. In other words, a nearly five-fold increase in MP power only raises horizontal resolution by a little more than half.

Through the consumer digital camera manufacturers' struggle to outdo each other in MP power, image quality has often been a casualty. To increase MPs but still maintain their small size, these cameras have had to squeeze more and more photo sites onto their image chips. These small photo sites receive less light and are subject to more noise and other aberrations, due to their tightly packed arrangement. Then, the camera must artificially boost and de-noise the noisy, dim-photo captures before writing them to its flash card.

"You end up with a photograph that is technically higher-resolution," observes an editorial in Digital Camera magazine, "but may in fact be of inferior subjective quality compared to the same shot taken with a camera offering a more modest pixel count."¹

Medical cameras

Cameras used in a medical setting are held to higher standards than consumer cameras. So, you are by no means doing your patients any harm by using a high MP count digital fundus camera. Therefore, too many pixels won't cause you to miss a diagnosis or otherwise misinterpret retinal health status. However, excessively large MP images occupy a lot of space on your hard drive's memory and can slow down your computer system. Further, you're legally obligated not only to store the image as part of the patient's medical record, but also to back it up. Storing and backing up thousands of large MP images can quickly become a practical concern.

Moreover, if you judge the quality of a digital fundus camera based primarily on MPs, you're not accurately evaluating that device. Consider these other more relevant factors:

► Use. Detachable and fixed digital fundus cameras are available. Prior to choosing between them, you must have an understanding of you and your patients' needs. The basics: If you'd like to use your digital camera for other purposes, such as documenting ptosis, and you want to play the "MP" game (i.e. eventually replacing the camera with a higher-resolution camera), the detachable system might be best for you. The fixed imaging system is ideal for the practitioner who must have his fundus camera attached all the time, due to both a high volume of patients and his need for instantaneous image transfer from the device to his computer system. (I've discovered that most detachable cameras store the images on the camera itself, requiring you or your staff to remove a memory stick and manually upload the images to your patients' records.)

I decided to buy a fixed camera because I have no desire to play the "MP" game; I appreciate instantaneous transfer; and I wanted the camera to remain in place. (I didn't want my staff or I to spend any time detaching it; I didn't want my patients to have to wait for a re-attachment; and I feared the camera could fall and break during handling.) These were my needs for use. What are yours?

► Optics. From an optical standpoint, you want a digital fundus camera to provide a usable image through as small a pupil as possible, so you don't have to dilate unless it's medically necessary. Why? Because doing so provides high-quality eye care fast (which impresses patients and improves practice efficiency) and prevents the patient from experiencing the physical and visual discomfort of drops. It's a win-win.

► Image presentation. When assessing the health of the retina, you want the image to provide as large a view as possible to ensure you don't miss any pathology and can provide each and every patient — regardless of the presence of pathology — with an impressive and understandable image. This reinforces to the patient that you're up to date on the latest technology and are, therefore, providing him with the best care. For example, my camera system offers a "montage" feature that combines a series of images into one large image of the retina. This has been beneficial, as it's facilitated diagnosis, treatment decisions, patient education and compliance (showing the patient his entire retina and related pathology has instilled adherence to my recommendations). Also, it's saved me time, as I don't have to click through several images to get a clear idea of the health status of the retina.

► Software power. The device's software should produce an unalterable first image (time and date stamp) as a means of providing you with legal protection. Many devices have this feature, while also allowing you to alter the first image (i.e. for the purpose of zooming in on pathology, etc.), though save it as separate image from the original.

In addition, you want the software to have tools, such as image manipulation, filters and a measurement feature, so you can accurately monitor one's retinal health.

Finally, the device's software should have an easy-to-use image transfer component. Many practitioners, including myself, appreciate this feature, as it prevents mistakes, such as transcription errors. For instance, my camera instantly transfers the image to the patient's medical file, which a staff member or I can then access from any computer terminal throughout my practice. Further, this image transfer component enables me to efficiently e-mail photos to retinal specialists.

In addition to these four relevant factors, you must test the device under several conditions, such as staff ease of use, and on patients who have varying retinal pathology. You want to make sure the camera fully meets your needs. Many digital fundus camera manufacturers will place a system in your office for a week or so, so you can assess whether the device will meet your needs.

Because the variations in optics, software and image quality among digital fundus cameras can be mind-bogglingly complex, it's not surprising that many of us have used and continue to use MPs as the primary criterion when selecting a camera. But, as much as we would like to boil down all imaging intricacy into one simple feature, the fact is no such yardstick exists. Keep this in mind prior to purchasing that shiny new digital fundus camera. OM

1. David MacNeill. "10 megapixels: Too much of a good thing?" Digital Camera. April 2006:10;39.

Dr. Kirschen is a professor of Vision Science and Optometry at the Southern California College of Optometry in Fullerton, Calif. Also, he's the chief of Binocular Vision and Orthoptic Services at the Jules Stein Eye Institute in Los Angeles. He's in private practice in Brea, Calif., where he avidly performs retinal photography. E-mail him at david1@kirschen.net.