Fudan's brain-spine interface from Shenfu Jianxing: brain electrodes decode intent and drive spinal epidural stimulation in paraplegia. Four patients in a 2025 investigator-initiated study. Hardware specifications are not published; sources are Chinese-language releases and the ChiCTR registry.
Independent, source-linked catalog sheet. Not a manufacturer-issued datasheet, regulatory decision or instructions for clinical use. Human evidence does not establish approval. Source-specific restrictions, conflicts and missing specifications are retained below.
Shenfu Jianxing brain-spine interface (Fudan)
This sheet covers the brain-spine interface (脑脊接口) developed by Jia Fumin’s Fudan team and its company, 神复健行. It is an implanted brain-to-spinal-cord system rather than a cursor or arm controller. Sources are Fudan and hospital releases in Chinese and the ChiCTR registry. Hardware details are mostly unpublished, so many cells are blank.
Identity
Field
Value and source scope
Device
神复健行 brain-spine interface (脑脊接口, ‘three-in-one’ 三合一 system): brain electrodes, spinal epidural stimulation electrodes and a skull-implanted miniature device with AI decoding [1, 2, 3]
Manufacturer
神复健行(上海)医疗器械有限公司 (Shenfu Jianxing, Shanghai), a Fudan University translation company founded around Jia Fumin’s team at Fudan’s Institute of Science and Technology for Brain-Inspired Intelligence [1, 4]
Interface class
Combined intracranial recording and spinal epidural stimulation, brain to spinal cord closed loop [2, 4]
Origin
Jia Fumin’s team at Fudan University with Zhongshan Hospital and Huashan Hospital [1, 2]
First demonstrated
First patient implanted January 8, 2025 at Zhongshan Hospital; Fudan reports leg movement on the first day after power-on and walking with assistance within weeks [2]
First human implant
January 8, 2025 at Zhongshan Hospital, in a 34-year-old man with paraplegia of two years after a thoracic spinal injury; second patient February 5, 2025 and third February 25, 2025; fourth patient March 3, 2025 at Huashan Hospital (hospital release) [2]
Species studied
Human
Regulatory status
Investigator-initiated clinical study, not approved. Fudan reported in December 2025 that the product entered the US FDA Breakthrough Therapy pathway, described as the first domestic device of its kind (university statement; the FDA designation itself was not verified here) [1]. The team said in March 2025 it planned third-party type testing and registration-trial preparation [2]
Function
Decodes motor intent from brain electrodes and drives epidural electrical stimulation of the lumbar spinal cord to restore leg movement in paraplegia [2, 4]
Target tissue
Motor cortex (brain electrodes placed by stereotactic surgery) and the spinal cord epidural space [1, 2]
Geometry and architecture
Field
Value and source scope
Interface type
Stereotactically placed intracranial electrodes plus epidural spinal stimulation electrodes, placed in one operation [2]
Array layout
Electrode count
Pitch
Electrode lengths
Shank width and thickness
Tip and exposed site geometry
Contact coating
Insulation
Insertion method
One-stage minimally invasive surgery placing brain and spinal electrodes. Duration conflict: the March 2025 hospital release says the first operation took 2 hours [2]; the university’s December 2025 release and the March 2025 summary say 4 hours for simultaneous electrode implantation [1, 2]
The team describes a skull-implanted miniature unit it plans to refine (‘颅骨植入式’ device) [2]; dimensions unreported
MRI compatibility
Surgical complexity
First case 2 hours per the hospital release [2]; see the duration conflict under insertion method
Output connectors
Performance envelope
Field
Value and source scope
Acute yield
Chronic yield
Stability over time
Longevity
Revision and explant experience
Adverse events
Notable demonstrations
Hospital release, March 2025 (not peer reviewed): first patient raised a leg on day 1, controlled both legs by brain on day 3, tried stepping on a ceiling rail on day 10, walked more than 5 m in a standing frame under suspension on day 15 and walked with a walker under suspension on day 49.
Clinical and preclinical evidence
Field
Value and source scope
Human subjects
4 patients in the clinical proof of concept by March 3, 2025, all men born in the 1990s, chosen from more than 300 applicants [2]. The registry ChiCTR2500095359 lists a sample size of 3 [4]; the hospital release counts 4 including a Huashan case, and both are listed
Preclinical cohort
Follow-up duration
Registry execution period January 1 to December 31, 2025 [4]
Indications
Lower-limb paralysis after spinal cord injury; registry inclusion: ages 18 to 65, spinal cord injury 6 months or more earlier, injury level T12 or above, ASIA grade B, C or D, normal arm function [4]
Trials and registries
ChiCTR2500095359, prospective registration January 6, 2025, single arm, Zhongshan Hospital sponsor and leader Ding Jing, ethics approval B2024-536R dated December 31, 2024, funded by hospital development funds [4]
Primary outcomes
Registry primary outcomes: ASIA score, lower-limb iliopsoas and quadriceps strength, and gravity-assist ratio needed for standing [4]
Key limitations
Four patients in an investigator-initiated study. Reported results come from hospital and university releases, not a peer-reviewed paper found here. No hardware dimensions, channel counts or electrode specifications published in the sources read
Engineering tradeoffs
Field
Value and source scope
Strengths
Single operation for brain and spinal electrodes; direct brain-to-spinal-cord link rather than control of an external machine, per the team [1, 2]
Limitations
Two implantation sites; implant hardware not described in detail; evidence is a small proof of concept [2]
Scaling constraints
Version boundary
The configuration studied is the March 2025 clinical proof of concept, with brain electrodes and spinal epidural electrodes. The team describes a later ‘three-in-one’ skull-implanted miniature device as a planned refinement; that version is not described in the sources read.