One theory about why phantom limb occurs is that it happens when areas of the brain that used to control the amputated limb remain strongly connected to the mental image of the limb. To weaken this connection, Takufumi Yanagisawa, MD, PhD, and his research team at Osaka University, Japan, hypothesized that unconsciously training the brain regions that control the intact limb will also control the phantom limb. In the random crossover trial, the researchers were able to reduce phantom limb pain after only three days of training with a brain-computer-interface (BCI).
During training, magnetoencephalographic (MEG) signals were recorded while patients watched the images of the phantom hand that was controlled by real-time decoding of the recorded MEG signals.
Photographs courtesy of Osaka University.
“It is very difficult to intentionally activate the part of your brain that controls your right hand without actually thinking about moving that hand,” explains Yanagisawa. “Instead, we designed a system in which the patients did not even know they were using those parts of their brains.”
Twelve patients with chronic upper-limb phantom limb pain due to amputation or brachial plexus root avulsion participated. Using the BCI, the researchers recorded brain activity when the patients opened and closed their intact hands and used the pattern of brain activity as a template. Then the researchers continuously recorded brain activity related to the intact hands, but the patients were asked to try to control a virtual hand with their phantom hand.
The patients were trained to move the virtual hand image controlled by the BCI with a decoder, which was constructed to classify intact hand movements from motor cortical currents, by moving their phantom hands for three days (“real training”). Pain was evaluated using a visual analogue scale (VAS) before and after training, and at follow-up for an additional 16 days. As a control, patients engaged in “random training,” in which the same hand image was controlled by randomly changing values. The two trainings were randomly assigned to the patients, and patients thought they were controlling the virtual hand in both versions.
In the real training, the recorded brain activity was decoded based on the template and the image of the opening/closing virtual hand was adjusted accordingly. In the random training, the images of the virtual hand were randomly adjusted with no connection to brain activity. Patients trained for about 30 minutes per day for three days, and after each session they rated the intensity of their phantom limb pain.
The rate of pain reduction is shown for each day relative to the pain on day one.
The research team found that the participants’ pain was reduced by 30 percent even on the first day of training, and the effect lasted up to five days after training was complete. Importantly, only patients who received real training reported less phantom pain. They also found that after training, the mental image of the phantom hand was weakened in the brain regions that once controlled the amputated hand. Compared to VAS at day one, VAS at day four was reduced by 32 percent, and on day eight, it was reduced by 36 percent.
“These findings are promising,” said Yanagisawa, “especially given that alternatives like mirror training require a month of training to have the same effect. However, in order for this treatment to become truly practical, the cost must be reduced.”
The article, “BCI training to move a virtual hand reduces phantom limb pain: A randomized crossover trial,” was published in Neurology.