A study published in Nature Neuroscience on the 17th reported a brain computer interface device that can convert the activity of "trying to type" in the brain into actual text. This achievement is expected to provide paralyzed patients with a "keyboard" to communicate in a way that is closer to people's daily habits. Brain computer interface technology aims to establish a direct communication channel between the brain and external devices, and is an important research direction to help patients with severe motor dysfunction restore their communication ability. Previously, various technological paths have been developed in this field, such as non-invasive systems (such as those based on electroencephalography or eye tracking) that are safer, but have limited signal accuracy and speed; Invasive systems can directly record neural signals through implanted electrodes, enabling more precise decoding. In terms of communication function reconstruction, existing solutions mainly include controlling the cursor to select characters, decoding brain activity attempting to produce speech to synthesize speech, or decoding handwritten handwriting. However, these methods are either slower or have specific requirements for the patient's residual motor or vocal abilities. Many people, due to long-term usage habits, prefer the QWERTY keyboard (i.e. full keyboard, which is the most widely used keyboard layout) as an intuitive input method. To this end, a research team from Massachusetts General Hospital implanted a new brain computer interface into the motor cortex of two quadriplegic patients. Participants were asked to attempt finger movements while typing on a QWERTY keyboard, during which electrodes implanted in the anterior cingulate gyrus recorded their brain activity. Based on this data, the research team used deep neural networks to establish a model for predicting the characters that patients want to input. The results showed that one participant could output 110 characters (approximately 22 words) per minute, which is equivalent to 81% of the typing speed of a healthy person's smartphone, with an error rate of only 1.6%; Another participant can type 47 characters per minute. The system only requires about 30 sentences of training to start running effectively. Although further validation is needed in more patients, the research team believes that the device can help paralyzed patients achieve fast, accurate, and easier to use communication. Compared to speech to text systems, it also has advantages in protecting communication privacy. This study demonstrates a new pathway for efficient text output through decoding motion intentions, providing new communication possibilities for patients with severe paralysis caused by conditions such as amyotrophic lateral sclerosis and spinal cord injury. (New Society)