Classic Computer Magazine Archive START VOL. 4 NO. 7 / FEBRUARY 1990

D O C T O R

IS THAT

AN ST I SEE?

MARCO TECHNOLOGIES'

ST-BASED VISUAL FIELD ANALYZER


BY BRUCE D. NOONAN, M.D.


The ST has proven itself to be an extremely versatile computer. Not only can you find STs in homes, offices and schools, but also on stage in concerts--and even, perhaps, in your ophthalmologist's office! Florida-based Marco Technologies has incorporated an Atari 520ST into a custom commercial system to measure, record and analyze a patient's visual response. In this article, ophthalmologist Dr. Bruce Noonan takes a look at this surprising use of ST power.


Although many of you know me as a programmer, my primary occupation is as an ophthalmologist (eye physician and surgeon). While poring over my usual mountain of professional journals recently, I came across an advertisement from a company called Marco Technologies for a visual field analyzer which made use of a mouse, as in computer mouse. On close examination of the picture, I saw that it was an ST mouse! I scanned the ad and read, "The MT-336 is powered by a one megabyte personal computer with a dual density floppy disk. A printer is included along with the mouse and the color monitor."

I then spoke with Mr. Scott Lewis, Vice President of Sales at Marco's home office in Jacksonville, Florida. I discovered that they were indeed using the ST for their perimeter. The software was written in compiled GFA BASIC by Dr. Bill Coan of Massachusetts. Coan has a Ph.D. in biophysics and is an avid Atari enthusiast. He has engineered and built several perimeters.

eyedoctor1.jpg
The Marco Technologies MT-336 Auto-Perimeter
is a sophisticated ophthalmological measuring
system built around an Atari 520ST. The program
that runs the system was written in compiled GFA
BASIC! You've come a long way baby!.

"OK," you may be saying, "so what the heck is a perimeter?" A perimeter is an instrument used by doctors to measure the vision, one eye at a time, for your entire visual field--horizontally, vertically and everywhere in between. Since it would take infinitely long to measure the ability to see an object at every single point, a perimeter takes a sample of points.

Marco's system comes with a "bowl" which serves as the view screen of the patient test equipment, a built-in chin and forehead rest, an automated table to hold the bowl, a 520ST with a double-sided drive and color monitor, a 9-pin Panasonic dot-matrix printer and a table for the computer and printer. Originally, Marco planned to use the 1040ST, but when Atari converted the less-expensive 520ST to a double-sided drive, the 1040ST was dropped.


Oh Say, Can You See...

Perhaps a bit of explanation is in order here. A person's keenest vision is in the very center of the visual field, representing the vision of the macula in the human retina. More specifically, it's the fovea, the center of the macula, where the highest resolution occurs. The fovea's resolution is very high, whereas the peripheral retina can detect only large objects.

The concept of resolution can be illustrated by how we view images created by scanners. Scanners can digitize an image, such as a photograph, from resolutions of 2400 dots per inch (very high resolution) down to 150 dpi or less with low-resolution scanners. They create an image by placing a series of dots in the resulting image to represent the most significant information in the scanned photo. The fovea can detect such dots in all but the very highest dot resolutions.

The macula area of the retina is also where color vision is perceived and reading vision (recorded as 20/20 or higher) is achieved. As images are focused further from the center of the visual field, only very bright or large objects can be detected and vision in these areas can be 20/400 or worse.


Testing a Visual Field

A basic static perimeter consists of a large illuminated bowl with a central fixation target. The patient must keep his attention focused on the fixation target while spots of light (stimuli) are projected or illuminated at random points in the visual field. When the patient perceives the stimulus, he presses a hand-held button and the intensity, spot size and location are recorded by the examiner or computer. The results can be drawn or printed out on a visual field map and then can be compared with normal fields. In this way, diagnoses of glaucoma or neurologic impairment can be made.

One of the most widely used perimeters is the Goldman perimeter. This perimeter has become the standard, although it requires a full-time operator and is not automated. The intensity of the stimulus and its size can be set by the examiner and then projected at a peripheral location and moved radially, like moving along a wagon wheel spoke toward the hub. When the patient presses the button indicating he has perceived the stimulus, a mark is made on the field map and the process is repeated along another 'spoke' until the full 360 degrees have been examined. This is called kinetic perimetry, since a moving stimulus is used.


Enter Marco Technologies

eyedoctor2.jpg
Marco Technologies takes advantage of the ST's
GEM system to produce accurate displays and
printouts of a patient's visual response.

The Marco perimeter uses static perimetry. All of its 336 spots are created by fixed-size LED's in fixed locations, although the brightness of each spot can be varied. The examiner presents a dim target to the patient and gradually increases its illumination until the patient indicates he can see it. The point at which it can be barely perceived is called the threshold level of brightness. The MT-336 has 25 levels of brightness available in 1/2-log unit steps.

The MT-336 can also be used for pseudo-kinetic measurements. Lights along the "spokes" are set at a specified brightness and, starting with the LED furthest out, adjacent stimuli are turned on and off in sequence until the patient signals that it was perceived.

An examiner can select from any of eleven preset types of field patterns or design a custom field. This would depend on whether the examiner wanted to screen a patient for a defect or examine his visual field for a known type of defect.

The operation of the program was nearly flawless. It will occasionally poll the perimeter bowl (connected to the RS-232 port) and the printer to see if they are "on," and notify you if they are not. Although designed for a 9-pin printer, my 24-pin Star NB24-15 worked fine and printed faster than the Star NX-10.

When performing perimetry, always make sure that the patient is looking only at the fixation point. If he or she looks away, the test will not be accurate and the resultant data will be meaningless. But each eye has a physiological blind spot on the retina on which you can't detect any stimulus. The program maps these out initially and then will occasionally display a stimulus in the blind-spot area. If the patient signals that he perceives it, then it's recorded as a fixation failure. The number of fixation failures is printed out along with the field test results, thus giving the doctor an indication of the reliability of the results.

Marco also uses a unique alignment system to make sure that the eye being tested is exactly centered in the bowl. This precise alignment is necessary for reproducible results. Marco's system involves the detection of color in the fixation point on the bowl. It's simple to use and accurate.


Three Flavors of Programs

Software for the MT-336 comes in three levels of sophistication: Basic, Plus 1 and Plus 2. The Basic program will only print out a hard copy of the data. Plus 1 lets the examiner store data in memory so that fields for both eyes can be printed at the same time. It also gives the examiner a greater selection of stimulus intensities for various tests. The Plus 2 program goes further, letting the examiner store field data on a floppy disk. It also provides a much needed utility: the Delta comparison.

The Delta comparison is an exciting concept. It lets examiners make comparisons on a point-by-point basis of threshold changes over a period of time. For example, if point "A" is seen with a threshold intensity of 7 dB during one examination, and has a threshold of 5 dB on a subsequent examination, this might be an indication that the patient is developing a field defect, such as a loss of retinal function, at that point. Remarkably, this sophisticated program allows comparisons of different patterns of fields, and even on fields where the patient did not complete the test! The printout will display graphically where such changes are occurring, alerting the doctor to do a more definitive test of that specific area. He or she then may choose to design a custom test concentrating on that area.

I had the opportunity to test the program on a number of glaucoma patients with known visual field defects documented on Goldman perimetry. Without exception, the program detected the field defects. Unlike other automated perimeters, the printed output is a screen dump in graphics mode which prints out an exact replication of the field as seen in the bowl. This is mostly because of GEM's ability to adapt output to any device without shape distortion.

There are several things I would like to see improved in the program, most of which the company is aware of and is implementing. The perimeter has a higher density of spots than nearly all of its non-projection competitors. Thus, it usually operates in "half-density" mode, where only half of the LEDs in the area are used. Such fields are equivalent to the competition, but if you want to use full-density fields, you must design a custom field. It would be nice to be able to select full density for the eleven preset field patterns rather than having to design a custom test. Designing a custom test is not all that difficult, however; you simply use the mouse to drag a rubber-band box around the area to be tested.

Although the mouse is used for nearly all program functions, if you wish to recall a previous field from disk, you must remember the name of the file and type in the filename--no file-selector box, an important oversight.

Similarly, a dialog box should be used for entry of the patient's data-- name, age and refraction. Also the date and time must be entered every time the program is booted; it would be much better for the program to scan for the date and time and ask if it is correct.

The learning curve for the program is fairly short. I did have trouble switching field types between eyes on one patient because I did not return to an earlier menu. Full-screen menus are used in place of drop-down menus and alert boxes appear in abundance. However, help screens are available at almost every stage of the program. The full-screen menus may be confusing, especially when choices are listed with alphanumeric characters, since it gives one the erroneous impression that you can use the keyboard to make choices.

The system modularity is what makes this perimeter a truly outstanding value. If the printer goes on the blink, you can substitute any Epson-compatible printer and continue working. If the computer goes down, any 520ST will work. For system-specific components, the support is excellent--I had trouble with a defective patient signal button and Marco had a replacement to me in two days. They are anxious to succeed in a very competitive market and I found their service to be excellent.

The system is certainly not inexpensive when compared to other ST programs. The cost of the system varies from $6,500 to more than $9,000, depending upon the software selected. But in comparison to similar automated perimeters--some costing $20,000 or more--the price is very competitive. Marco's system supplies the clinician with one more means of tracking disease, and given the history of many high-tech medical advances, may prove to be the standard by which perimetry is measured in the future. I personally was very impressed by the system and would definitely recommend it to health care professionals looking for an inexpensive perimeter with state-of-the-art computer capabilities.

Oh yes, by the way, Marco is currently including ST Writer Elite 3.0 free of charge along with the package. It seems they know good software when they see it!

Dr. Bruce Noonan is an ophthalmologist who is best known throughout the ST world for his stalwart support of ST Writer. Watch for the latest version of ST Writer Elite on an upcoming START disk.


PRODUCTS MENTIONED

MT-336 Auto-Perimeter, $6,500-$9,000. Marco Technologies, 11825 Central Parkway, Jacksonville, Florida 32216, (800) 874-5274 Nationwide, (800) 227-0084 in Florida.