We’re offering a system for clinical testing that takes advantage of the qCSF algorithm and delivers it to the marketplace: the Manifold®. The Manifold® is a distance vision clinical testing system that includes several parts: a display, cart, and handheld remote control, with software components that include the qCSF algorithm.
Our development of the Manifold® has refined the test to improve usability and comfort, while maintaining laboratory-grade precision. Our first research version used sine-wave gratings similar to other assessments of the contrast sensitivity function, with a high guessing rate (50%) and lower statistical efficiency. To translate this work to a clinical tool we have tapped previous work on spatial vision. The Manifold® uses the set of Sloan optotypes, which are then bandpass-filtered with a raised cosine window, with peak frequency 4 cycles per letter. The hybrid letter-grating stimulus that results provides the best of both worlds: the pattern is easily identifiable as a letter, but also exhibits the narrow-band frequency information of a grating. Another advantage of the Sloan set is the reduction of the guessing rate (1 of 10 = 10%), which improves test efficiency.
We seeded prototypes of the Manifold® to research and clinical labs in the United States and Germany. Our research partners—prominent clinicians from Harvard Medical School, Johns Hopkins, Hamburg-Eppendorf, Wright-Patterson Air Force Base, Navy Surgical Medical Center, Kellogg Eye Center, Retina Foundation of the Southwest, and Nova Southeastern College of Optometry—helped us gather real-world data to validate the device and method.
-
We search the stimulus space of contrast sensitivity testing and adaptively infer the best test combinations. These stimuli represent a letter-grating hybrid that provides easy recognizability by patients, with precision in stimulus information content.
-
In each screen, it is very likely that at least one of the letters is easily recognizable by the observer, which helps the observer to determine the location and size of the test letters.
-
The quantitative CSF method is adaptive—each screen informs the next. Algorithms learn from patients’ answers to adapt the next question to ensure that only meaningful questions are presented.
-
Accuracy of the results varies at first, but over time, the accuracy begins to converge until the Manifold® arrives at a personalized vision profile. This makes the test rapid, repeatable, and highly precise.
-
Screenshot of remote control during a test session. The trial history (correct answers, triangles; mistakes, crosses; ‘no answer’ responses, slashes) and the current best estimate of the CSF are shown in the top panel. Spatial frequency is on the x-axis and contrast sensitivity on the y-axis.
-
Screenshot of remote control, showing test results after a session. The solid line denotes median sensitivity estimate for the posterior distribution of possible CSFs, with the shaded region denoting the 66% confidence interval.
The commercially available first-generation Manifold Contrast Vision Meter was successfully used in more than 20 completed and ongoing clinical trials. Its successor, the Manifold Vision Meter, additionally offers quantitative Visual Acuity assessments, a wider spatial frequency range for testing from low-vision IRD patients to healthy controls, and enhanced usability with the unique protocol management feature. By now, the Manifold® has been deployed in international Phase II and Phase III and provided secondary and primary endpoints to support drug development across a number of indications including diabetic retinopathy, intermediate AMD, and geographic atrophy.
We believe AST products will become a new standard: a rapid and precise way to assess and monitor visual health over time.