Artificial vision with wirelessly powered subretinal electronic implant alpha-IMS

Abstract

This study aims at substituting the essential functions of photoreceptors in patients who are blind owing to untreatable forms of hereditary retinal degenerations. A microelectronic neuroprosthetic device, powered via transdermal inductive transmission, carrying 1500 independent microphotodiode-amplifier-electrode elements on a 9 mm(2) chip, was subretinally implanted in nine blind patients. Light perception (8/9), light localization (7/9), motion detection (5/9, angular speed up to 35 deg s(-1)), grating acuity measurement (6/9, up to 3.3 cycles per degree) and visual acuity measurement with Landolt C-rings (2/9) up to Snellen visual acuity of 20/546 (corresponding to decimal 0.037° or corresponding to 1.43 logMAR (minimum angle of resolution)) were restored via the subretinal implant. Additionally, the identification, localization and discrimination of objects improved significantly (n = 8; p < 0.05 for each subtest) in repeated tests over a nine-month period. Three subjects were able to read letters spontaneously and one subject was able to read letters after training in an alternative-force choice test. Five subjects reported implant-mediated visual perceptions in daily life within a field of 15° of visual angle. Control tests were performed each time with the implant's power source switched off. These data show that subretinal implants can restore visual functions that are useful for daily life.

Figures

Figure 1.

Human eye. (a) The structures of the eye and (b) the retinal layers in detail. (c) The function of photoreceptors lost because of hereditary degeneration can be partially replaced by a subretinal chip. The chip carries a microphotodiode array with amplifiers and electrodes on a 3 mm × 3 mm area and is surgically placed subretinally in the location corresponding to the layer of degenerated photoreceptors.

Figure 2.

The alpha-IMS subretinal implant. (a,b) The subdermal coil behind the ear provides power and sends control signals via a subdermal cable and a thin intraocular foil to the chip in the eye. (c) The chip is placed surgically beneath the fovea and contains 1500 pixels (independent microphotodiode-amplifier-electrode elements) on a 3 mm × 3 mm area. Via a thin black cable, a small battery pack (not shown) powers the primary external coil, (d) which is magnetically kept in place above the subdermal coil behind the ear and provides power and signals via transdermal electric induction.

Figure 3.

Screen tasks. (a) Using a projector-screen set-up*. (b) Light perception threshold, light source localization, motion detection, grating acuity (spatial resolution of periodic stripe pattern) and visual acuity (standardized Landolt C-shaped optotypes) were assessed. A four-alternative-forced-choice mode (4AFC) was applied, except for light perception where a 2AFC was used. (c) Light perception with the implant was possible in eight subjects, light source localization was possible in seven subjects, motion detection (angular speed up to 35 deg s−1) was possible in five subjects, grating acuity measurement (up to 3.3 cpd) was possible in six subjects, and visual acuity measurement with Landolt C-rings was possible in two subjects (20/2000 and 20/546, corresponding to gap sizes of 1.6° and 0.45° visual angle). At least 75% (in 2AFC) or 62.5% (in 4AFC) correct responses were required to pass the test (‘yes’). *Grating acuity measurement and Landolt C-ring tests were carried out on a table with subject S5 using a set-up similar to that shown in figure 4.

Figure 4.

Table tasks of ADL. First, four geometrical objects (out of six possible objects: circle, ring, crescent, triangle, square, rectangle) were placed on the table. The patient was not aware of the maximal possible number of the shapes put in front of him. The patient was asked to report, how many, where and which shapes he/she could see. For every question, the number of correctly identified, discriminated and localized objects was documented on a scale ranging from 0 to 4. In the second part of the test, a table setting was presented using white tableware. A large white plate in the middle was obligatory and known to the patient. Around the plate, four objects (out of six possible objects: middle-sized plate, small plate, cup, spoon, fork or knife) were arranged. Again, recognition, description and localization were reported and documented. (a,b) Significant differences were found between the ON/OFF (grey bars and black bars, respectively) implant power supply conditions for all tasks performed on a table with geometrical shapes (c) and (d) tableware objects in all eight patients. Whiskers indicate the standard deviation. The significance level was reached for all six questions (the asterisk indicates p < 0.05).