3

Looking into the Future: Novel Uses of Emerging Neurotechnologies with Potential Legal Applications

The BRAIN Initiative and other national-level brain projects around the world, as well as ventures launched by companies in the private sector, have ushered in what Khara Ramos called “a transformative time for the development of novel neurotechnologies,” which have enabled discovery of unexpected aspects of brain function. Although most of these new and emerging neurotechnologies are designed to better understand neurological deficits or provide benefits to patients—not to answer legal questions—they eventually may be applied in legal settings, said Sydney Cash, associate professor of neurology at Harvard Medical School, adding that 25

discussing the potential legal implications early in development of these technologies may be beneficial. Ramos agreed, adding that this workshop provides an opportunity to proactively consider the implications.

CLOSED-LOOP DEEP BRAIN STIMULATION

Aysegul Gunduz, director of the Brain Mapping Laboratory at the University of Florida, admitted that she had not considered the legal implications of her work until receiving an invitation to the workshop. Nonetheless, as she proceeded to describe the research conducted in her lab, several concepts emerged with potential application to the law.

Gunduz’s research focuses on the billions of neurons in the human brain that synchronize to create electrophysiological waves, or brain rhythms, which are modulated to control behavior. Both neurological and psychiatric disorders are characterized by distortions of normal modulations, she said. For example, clusters of neurons located deep in the brain, known as the subthalamic nucleus (STN), have been shown to play an important role in movement disorders such as Parkinson’s disease (PD) (Rossi et al., 2015). Gunduz and colleagues use deep brain stimulation (DBS) to target the STN with an electric pulse that works like a pacemaker to modulate the pathological rhythms generated in these circuits and bring the brain to a healthier state. In PD or essential tremor, for example, DBS can lead to a suppression of symptoms. The technology has also been used for other disorders such as dystonia and obsessive-compulsive disorder, and is under investigation for Alzheimer’s disease, Tourette’s syndrome, and drug addiction.

Gunduz and colleagues, with support from the BRAIN Initiative, are trying to make these devices smarter, so that they can deliver electricity to the brain on an as-needed, individualized regimen, to make treatment more effective, reduce side effects, and preserve battery life. To achieve this, they are working with individuals who have kinetic tremors (a subset of individuals with essential tremor), which appear only when a movement is initiated. If the intention to move can be detected, it may be possible to deliver DBS before the tremor begins, and stop the stimulation when intent is gone. They place a sensing electrode on top of the premotor cortex, which has been shown to be activated before a movement starts, and then attach that electrode to the device that delivers the DBS. In one patient given the task of pouring water from one cup to another, the patient’s system was able to detect the intent to move and deliver stimulation to the

deep brain before the movement began, reducing the tremor and allowing the task to be completed successfully.

DBS may have uses in legal settings. An example is for forensic evidence collection, said Gunduz, although those uses are far in the future and would need to overcome substantial ethical barriers. For example, placing electrodes deep in the brain could conceivably be used to detect lies, although fMRI studies show that the networks activated during the complicated act of lying cover a broad area of the brain not yet accessible to implantable electrodes, she said (Langleben et al., 2005). Gunduz noted that neuromodulation may also be used to promote truth telling. Studies have shown that damage to the dorsolateral prefrontal cortex leads to less honest behavior in tasks that pit honest motives against self-interest, raising the interesting yet highly speculative possibility that an area of the brain could be stimulated to make people more truthful (Zhu et al., 2014).

Gunduz also mentioned a group of projects funded by DARPA with the overall goal of developing a fully implantable closed-loop brain stimulator to restore active memory (DARPA RAM)¹ in soldiers who have suffered traumatic brain injuries. A recent study demonstrated that targeted stimulation of the lateral temporal cortex rescued poor memory encoding and improved later recall (Ezzyat et al., 2018); another proof-of-concept study recently demonstrated that electrical stimulation of an area of the hippocampus known to be important in memory formation resulted in improved short- and long-term retention of visual information (Hampson et al., 2018). While these studies show promise in restoring memory, Gunduz cautioned that electrical stimulation may also disrupt working memory rather than improve it; thus, she does not see it as either an ethical or scientifically probable approach to obtain truthful testimony.

CONTINUOUS RECORDING OF BRAIN ACTIVITY

Cash expanded on the idea of recording brain activity as a means of understanding the neural basis of volition. Cash’s lab has been trying to decode motor activity and language processing in persons with tetraplegia (partial or total paralysis in all four limbs and torso caused by illness or a spinal cord injury), such as those who have suffered devastating strokes that leave them paralyzed and often unable to communicate with voice.

___________________

¹ To learn more about DARPA RAM, see https://www.darpa.mil/program/restoring-active-memory (accessed April 28, 2018).

Working with the interdisciplinary research at BrainGate,² they have tested a microelectrode device that decodes motor activity from the brain via a set of electrodes implanted in the motor cortex, then uses that decoded information in real time to control a sophisticated multidimensional robotic arm (Hochberg et al., 2012). Essentially, said Cash, a patient can control the robotic arm simply by thinking about what he or she wants to do. This highly invasive approach facilitates the collection of very precise information from one location at extremely high temporal resolution, that is on the order of submilliseconds, enabling the investigators to decode the activities of individual neurons while the patient is performing a task.

The same technological approach can also be used to decode auditory information, said Cash. Using data captured from a microelectrode array implanted in the superior temporal gyrus brain, a multidisciplinary research team at Harvard–MIT and the University of California, San Diego, have shown that single neurons are tuned to specific phonemes, thus enabling spoken words to be decoded (Chan et al., 2014). Eventually, said Cash, this technology may move beyond decoding to facilitating the restoration of function, for example, by allowing non-verbal individuals to communicate by typing. Work is under way already, with help from BRAIN Initiative funding,³ to make these arrays fully implanted and fully wireless. Another approach would integrate implantable arrays with electrodes implanted into muscles to direct muscle activation in patients with tetraplegia (Ajiboye et al., 2017).

Ultimately, the goal would be to move beyond enabling people to complete prescribed tasks in controlled test environments to the point where they would be able to have complete volitional control in natural settings, said Cash. To move in this direction, Cash and the BrainGate team have been working experimentally with a Watch, Imagine, Act paradigm. Using implanted multielectrode arrays, they record neural activity at a single neuron level in participants with epilepsy or tetraplegia as they look at a movement, imagine doing the movement, and then attempt to perform the movement. The researchers have found that single neurons in the motor cortex fire at different rates depending on whether they are watching, imagining, or acting. This system thus provides information

___________________

² To learn more about BrainGate, see http://www.braingate.org (accessed April 28, 2018).

³ BRAIN UH2-NS095548, Hochberg and Nurmikko.

about volitional state with very high temporal resolution. Cash and colleagues are trying to look at continuous activity in patients doing whatever they wish to do.

A major challenge, he said, is analyzing the enormous amounts of multidimensional data collected in these experiments—not only the neural activity data but behavioral data as well. Video recording allows them to collect information on types of movement, joint position, contextual information, and audio information; to decode that information; and to integrate it with neural data using sophisticated analytical approaches.

Cash said these “embryonic” efforts to understand the neurological basis of volition, intent, recovery of memory, etc., are not yet far enough advanced to be mapped onto legal frameworks or to be used within the justice system. Their use in predicting outcomes or responsiveness following incarceration is especially fraught, he added. Moreover, coupling these systems to neuromodulation to prevent antisocial behaviors, while biologically feasible, raises major ethical, practical, and legal concerns, Cash said. Indeed, Hank Greely commented that these technologies, such as implanted chips that can make someone do things or stop someone from doing things, seem to be moving closer to the dystopian prospect of mind control, requiring careful thinking about appropriate limits.

This page intentionally left blank.