Infant Vision Assessment with LabVIEW on Linked Macintosh Computers

Author(s):
Scott Steinman - Southern College of Optometry

Industry:
Life Science

Products:
LabVIEW,

The Challenge:
Building a state-of-the-art clinic for visual electrodiagnosis, eye movement recording and psychophysics in young infants, with three design goals: (1) fast simultaneous visual stimulus displays and data acquisition (2) easy operation, and (3) expandability.

The Solution:
Using LabVIEW software and DAQ boards to acquire physiological data while controlling the visual stimulus display.

Design Challenges
When I was asked to create a new, state-of-the-art clinic for testing the visual function of infants with eye or neurological diseases, or those at risk for developing visual disorders, I knew that it could be done only with LabVIEW.

The clinic had to be capable of performing some fairly complex tasks. First, we needed the ability to measure a very wide range of recordings from infants. These included visual electrodiagnostic tests, such as evoked potential measurements of the brain’s neural activity in response to visual input, as well as electroretinograms, a measure of the electrical activity of the neurons within the eye’s retina. Many different types of evoked potentials had to be recorded, so that we could determine the infant’s visual acuity, the presence of eye or neurological disease, or maldevelopment of the visual system that could lead to strabismus (eye turn) and amblyopia (developmental loss of vision). Other tests included eye movement recordings to assess binocular visual functioning, and stereoscopic psychophysical tasks to measure the infant’s developing ability to perceive depth. Second, all of these tests needed to be operated by a single (sometimes computer-shy) clinician who not only had to run the software, but also had to monitor the infant’s direction of gaze and attentiveness, as well as the quality of the recorded data. Third, the tests needed to run fast. Because infants have short attention spans, you need to gather data rapidly. Finally, new developments in infant testing had to be added to the clinic quickly and easily.

I could purchase turnkey systems to perform most of the electrodiagnostic tests on infants, but they could not perform all of the tests that I had in mind. Some of these tests were so new that scientists had only conducted them in infant research laboratories. The turnkey systems also were not expandable, so I could not add new tests in the future.

I had programmed for many years in my laboratory, and recently purchased LabVIEW for data acquisition and analysis. With LabVIEW, I could build simple, easy to operate "instrument panels" that my assistant could use easily. The infant vision tests involved displaying an animated visual display at up to 60 frames/s while simultaneously recording the infant's physiological responses to the visual stimulus at 1,000 Hz sampling rates using the NB-MIO-16X board, then displaying the recorded data.

The Infant Vision Clinic Hardware and Software
The clinic was therefore constructed using two interconnected Macintosh Quadra computers. One computer is dedicated solely to data acquisition and analysis using a National Instruments NB-MIO-16X analog input board and LabVIEW software. The second computer is dedicated solely to displaying rapidly animated visual stimuli by means of small C++ programs that perform either palette animation or frame animation. With two Macintosh computers, we can perform data acquisition and animation at the highest possible rate. AppleEvents, specifically UserLand Frontier, is the glue that ties everything together, while the LabVIEW program coordinates the overall operation of the software, and serves as the "master". The clinician selects which test to run, the recording parameters, and the stimulus parameters for the test to be executed.
Once the infant is ready, the test begins with a mouse click. An AppleEvent is sent by the LabVIEW program across an Ethernet cable (this transmission is accelerated by Frontier), to the stimulus display program on the second Macintosh. The stimulus display computer acts as a "slave" to the recording computer. The AppleEvent it receives tells the display program what to display and when. The precise time locking of the visual display and data acquisition is ensured by sync signals sent across digital I/O lines between NB-DIO-96 boards within each computer.

Each infant visual test is therefore composed of three modules - a LabVIEW subroutine or "virtual instrument" that records data and sends AppleEvents, a Frontier script on each computer to transmit the AppleEvents, and a C++ program that accepts AppleEvents and animates the appropriate visual stimulus. Because of the modular nature of the software, we can quickly develop and use new tests in the clinic. The unified, simple user interface of the LabVIEW main program makes the operation of the software simple and rapid for the clinician.

Conclusion
LabVIEW and the Macintosh computer were both critical to the development of the infant vision clinic and to its continued operation. LabVIEW slashed my development time, opened up a wider choice of recording and analysis techniques, and provided AppleEvents tools for the intercomputer communication that made the simultaneous high-speed visual stimulus display and data acquisition possible. The infant vision software demonstrates that LabVIEW is not just a laboratory research or engineering tool - it can also solve practical real-world problems in such unlikely places as infant vision clinics.

Dr. Scott Steinman is the Chair of the Department of Biomedical Sciences at the Southern College of Optometry and a vision scientist whose main interest is the development of computer-based tests of visual function. He has authored several articles on research programming and a textbook entitled "Visual Programming with Prograph CPX" (Manning Publications/Prentice-Hall).

For more information, contact:

 Dr. Scott Steinman

Southern College of Optometry

1245 Madison Avenue

Memphis, TN 38104

E-mail: steinman@sco.edu

Browse All Case Studies »