Retinal prostheses are microelectronic devices, implanted into the back of the eye, with the goal of stimulating nonfunctioning neurons in the retina so as to restore a sense of vision to the blind. Despite much effort over several decades however, performance of these devices remains limited. While multiple factors are likely to contribute, a significant concern is thought to be the inability to create patterns of neural activity with the prosthetic that match the patterns that arise naturally in the healthy retina. Cochlear implants and deep brain stimulation, two neural implants that have been highly successful, are both thought to re-create some key elements of natural signaling even though they do not perfectly replicate natural signaling patterns, thus offering a roadmap for progress with retinal devices. Novel stimulation hardware, using both electrodes and other new stimulation modalities, along with novel stimulation strategies are under development and testing.
Background
In theory, each electrode from a retinal implant should activate a small number of nearby neurons; the resulting percept is referred to as a phosphene. Multiple phosphenes, produced by simultaneous stimulation from multiple electrodes, should then add in predictable ways to convey spatial information to the user. In reality however, individual electrodes are each surrounded by a different complement of neurons (there are ~40 different subtypes of retinal ganglion cells in the mammalian retina). As a result, the patterns of neural activity created by each electrode in nearby neurons can vary widely. In addition, it can be difficult to control the spread of activation arising from electrodes and thus the region of activity can be quite large, and in some cases overlap with that from other nearby electrodes. Such variability in the patterns of elicited neural activity is consistent with the variability in reports of phosphene appearance as well as the unreliable assembly of phosphenes into more complex percepts. Much of our effort here is devoted to studying the response of retinal neurons to artificial stimulation and then using the insights we gain from this work to develop new and more effective stimulation techniques.
In parallel, we collaborate with a number of researchers developing novel stimulation modalities. One example, is work with Chen Yang and Ji-Xin Cheng from BU who have developed an optoacoustic approach for retinal stimulation in which a pulsed light is illuminated on an absorber, resulting in transient heating and subsequent generation of acoustic waves at controlled ultrasonic frequencies. There are several advantages of an optoacoustic approach. For example, focused ultrasound can confine spatial resolution to regions as small as 25- 70 micrometers. In addition to the potential for high acuity, the use of precise stimulation helps to avoid cross talk between adjacent channels. Another advantage is the ability to take advantage of the parallel capability of optics using a DMD to generate patterned light. This enables parallel stimulation in up to 6850 pixels over a diameter of 7 mm, corresponding to a density of 178-pixels mm-2, comparing favorably to the highest density devices currently under development. Finally, the ultrasound penetration used in this approach is insensitive to scarring and inflammatory tissue that can form on the device post-surgery.
What we do
We continue to evaluate responses to artificial stimulation in retinal neurons using a combination of electrophysiology, immunochemistry, and computational modeling. Methodology includes single cell recordings in vitro, behavioral responses in vivo in both rodent and non-human primate. We also evaluate responses to optoacoustic and other novel stimulation techniques, investigating fundamental mechanisms of activation as well as effectiveness in replicating physiological spiking patterns.
Selected Publications
-
- Fried SI, Lasker AC, Desai NJ, Eddington DK & Rizzo JF (2009). Axonal sodium channel bands shape the response to electric stimulation in retinal ganglion cells. J. Neurophys. Apr;101(4):1972-87. PMID: 19193771.
-
- Freeman, DK, Eddington, DK, Rizzo, JF & Fried, SI (2010). Selective Activation of Neuronal Targets with Sinusoidal Electric Stimulation. J Neurophysiol. 2010 Nov;104(5):2778-91. PMID: 20810683
-
- Cai, C, Ren, Q, Desai, NJ, Rizzo, JF, Fried, SI (2011), Response variability to high rates of electric stimulation in retinal ganglion cells. J. Neurophysiology 2011 Jul;106(1):153-62. PMID: 21490287.
-
- Twyford P, Cai C, Fried S. (2014), Differential response to high-frequency electric stimulation. J Neural Eng. 2014 Feb 21;11(2):025001. PMID: 24556536.
-
- Guo, T, Lovell, NH, Tsai, D, Twyford, P, Fried, SI, Morley, JW, Suaning, G, Dokos, S, (2014), Selective activation of ON and OFF retinal ganglion cells to high frequency electric stimulation: a computational modeling study. Conf Proc IEEE Eng Med Biol Sci. 2014; 2014:6108-11. PMID: 25571391.
-
- Im, M, Fried, SI, (2016), Temporal properties of network-mediated responses to repetitive stimuli are dependent upon retinal ganglion cell type. J Neural Eng. 2016 Apr; 13(2):025002. PMID: 26905231.
-
- Ryu, Sang Baek; Fried, Shelley, (2018), Comparison of responses of visual cortical neurons in the mouse to intraocular and extraocular stimulation of the retina, Conf Proc IEEE Eng Med Biol Sci. PMID: 30440905.
