Coping with paralysis can be a psychologically and physiologically exacting process. Depending on the severity of the paralysis, even the most basic motor movements could become tough to perform. Many paralysed people also suffer from chronic depression as a result of their daily struggles and the feeling of loneliness. But what if we could actually and tangibly improve the quality of life of paralysed patients?
Reversing paralysis using Brain-to-Computer Interfaces and aiding communication in paralysed people with novel Neural interfaces
This is where the curative and palliative potential of Neurotechnology comes in, putting to use brain-to-computer interfaces. These interfaces can help paralysed people communicate at the speed of their thoughts, which is way faster than the speed with which we speak or write down thoughts. For people whose paralysis hasn’t impaired their eye movements or vocalization, devices such as eye-movement tracking keyboards and voice command readers can help them transcribe thoughts into words. For people afflicted with paralysis of the eyes and vocal chords, BCIs can come in handy, helping them to communicate by decoding neural patterns and brain activity by using non-invasive brain reading technology like EEGs. They can also help paralysed people with minor movements of hands or legs, to reach for and manipulate large objects, thereby restoring motor function in patients. BCIs can still not compete with traditional eye-trackers in the consumer market, mostly because typing using thought is a complex and laborious process, with a huge room for error in transcoding brain signals.
BCIs have been used to help a paralysed man to write text at speeds rivaling his able-bodied peers. The BCI converts mental handwriting movements from brain signals into on-screen words, by asking him to imagine writing with a pen on a piece of paper. AI machine learning algorithms coupled with a type of artificial neural network called ‘recurrent neural network’ or RNN, can be used to accurately predict sequential data like imagined handwriting or speech. Additionally RNNs use a technique called ‘data augmentation’ to produce artificial sentences by using the participant’s previously generated neural activity patterns. The training data can also be expanded by introducing artificial variability into neural signal patterns, to emulate changes that naturally occur in the human brain. Neural variability is defined as the temporal variability in neural activity during handwriting as opposed to one during drawing straight lines.
This is where invasive BCIs or iBCIs can help save the day, by basically implanting baby aspirin-sized electrodes in the motor cortex of the brain and thereby improve typing speeds upto 40 characters per minute. iBCIs can hugely improve both performance and functionality by decoding letters being imagined in the user’s brain. They use a machine learning algorithm repurposed from one originally designed for speech recognition, that allows scientists to read neural activity and estimate when, in real-time, the user starts to imagine writing a word.
BCIs can go a step further and aid paralysed people to operate everyday digital devices like a phone or a tablet by transcoding thoughts of making cursor movements and clicks.
Experimental BCIs called ‘Brain Gate’ has been shown to help patients surf the web, shop online, chat with friends on social media or play games on a tablet PC, decoding thoughts that arise in the motor cortex and reroute them to external, calibrated digital devices or modified Bluetooth interfaces. People who have lost the ability to move, from an illness like ALS and tetraplegia or a spinal chord injury can move a robotic arm using just their thoughts. However these assistive devices are limited in their scope, functionality, speed of use and flexibility of the interfaces. This can hinder efforts of neural restoration of motor activity in people suffering from chronic stroke or brain damage or locked-in syndrome or other forms of paralysis. However, the exponential advancements in Neurotech would soon reach an operational pinnacle that would easily outweigh its limitations.