[en] Objective: To assess the functional recovery in the patients with traumatic brain injury (TBI) after autologous implantation of neural progenitor cells, and 7 counterparts with matched age, injury location and extent were chosen as the control. Methods: Neural progenitor cells were isolated from exposed brain tissue and propagated for 25 to 30 d, then implanted the autologous neural progenitor cells at seven points around the traumatic regions with MRI-stereotactic guiding device for 7 patients. All recruited patients underwent ¹⁸F-fluorodeox-yglucose (FDG) PET imaging, function MRI (fMRI) and assessment of Glasgow outcome scale extended (GOSE) after operation for open brain trauma. The examinations were repeated one month after neural progenitor cell implantation and then repeated every 3 months during follow-up in the first year, and every 6 months in the second year. The same examinations were performed on untreated counterparts at similar intervals for avoiding deviations of spontaneous recovery. The data were analyzed with region of interest (ROI) and statistical parametric mapping (SPM). Results: At the third month of follow-up, mean tracer uptake in the damaged territory in implantation group increased significantly (P<0.01) from (71.1 ± 7.9)% to (157.5 ± 8.9)%, whereas the change in the control group was much less, which increased from (74.1 ± 9.7)% to (91.6 ± 17.6)%. Furthermore, it was notable that mean tracer uptake in ROI reached a plateau at the third month and persisted about nine months in 2 patients with long-term follow-up, and then went on elevating slightly. The tracer uptake in the damaged areas in control group increased slowly, plateaued at the ninth month, and no longer increased afterwards. SPM analysis showed that the metabolism of ¹⁸F-FDG in the top of precentral gyrus was significantly increased in implantation group, and the metabolism of ¹⁸F-FDG in the frontal lobe was significantly elevated postoperation according to paired SPM analysis. The activation in fMRI maps was seen in the motor cortex since the third month after implantation, whereas no active signals were detected before implantation or in control group. At the 6th month of follow-up, mean score of GOSE in the group of implantation was 6.63±0.52, whereas the mean score was 4.50 ±0.76 in control group (P<0.01). The increase of ¹⁸F-FDG uptake in the injured area was 3 months prior to the elevation of GOSE. Conclusions: The results of the study show that ¹⁸F-FDG PET and fMRI both showed significantly increased neurological function of injured region in implantation group compared with control group, and it was consistent with the change of GOSE. The increase of ¹⁸F-FDG uptake in the injured area was 3 months earier than the elevation of GOSE and with a plateau lasted nine months. (authors)