Abstract: BACKGROUND: Endoscopic fenestration and decompression of cervical lamina lead to changes in bony structure of cervical vertebrae, which in turn leads to biomechanical changes of responsible segments. In clinical rehabilitation, it is urgent to understand the biomechanical changes caused by structural changes so as to establish an accurate exercise rehabilitation program.
OBJECTIVE: To establish a three-dimensional solid finite element model for the treatment of cervical spondylotic myelopathy by using the finite element reverse engineering technology, and to establish the static biomechanical analysis of cervical physiological activities, and to formulate the strategy of postoperative accurate exercise rehabilitation training. 
METHODS: The thin slice plain scan data of cervical vertebrae of volunteers before and after two operations were obtained and imported into Mimics 20.0 software, respectively. Using the preoperative CT data, the finite element model M of C4-5 segment bone tissue was established. The fenestration model M1 of unilateral C4-5 lamina was established by using CT data after the first operation. The fenestration model M2 of bilateral C4-5 lamina was established by using CT data after the second operation. Three kinds of bone models were introduced into 3-matic for surface construction. A 75 N axial downward pressure was applied on the upper surface of the C4 vertebral body, and the intervertebral disc force and vertebral body displacement were recorded under different working conditions.   
RESULTS AND CONCLUSION: (1) Compared with model M, the stress in intervertebral disc b and d regions of model M1 increased by 10%-12% during the right bending activity, and there was no significant difference in other physiological activities. The stress in b and d regions of model M2 increased by 10%-13% when the right bend was active, and the stress in a and c regions increased by 9%-11% when the left bend was active, and there was no significant difference in other physiological activities. (2) Compared with model M, the displacement of model M1 increased by about 15% under the right bending physiological activity, and increased by about 13% in other directions. Under the physiological activities of left and right lateral bending, the displacement of model M2 increased by about 17%, and increased by about 14% in other directions. (3) It is concluded that full endoscopic fenestration discectomy for cervical spondylotic myelopathy has little effects on the biomechanical stability of the spine as a whole. An accurate exercise rehabilitation program was made according to the biomechanical changes caused by the changes of bone structure after operation. For the patients after unilateral full endoscopic fenestration discectomy, they should focus on the training of the cervical semispinalis muscle and the cephalic semispinalis muscle on the operative side. For patients after bilateral full endoscopic fenestration discectomy, it is necessary to train muscle tissues such as bilateral cervical semispinalis muscle and cephalic semispinalis muscle, so as to increase the muscle strength of the back of the cervical vertebra, and make the cervical vertebra better on the central axis through muscle strength, so as to achieve the purpose of stabilizing the cervical vertebra.

Key words: cervical spondylotic myelopathy, spinal endoscopy, minimally invasive decompression, precision sports rehabilitation, finite element analysis, biomechanics