Traffic accidents, high-altitude drops, mechanical shock and other accidents may lead to human spinal cord injury, interrupting the transmission channel of human nerve signals, resulting in loss of some limb function. Since there is no method of spinal cord injury site of nerve cell activity, while the conventional testing instruments such as X ray, CT and other traditional instruments cannot provide any nerve cell activity, doctors couldn’t have a clearer understanding of the patient’s condition, which may delay the best timing of treatment, thus causing lifelong paralysis. Spectroscopy method can be used in the detection of cell organization change while near infrared spectrum technology is a kind of fast measurement and simple operation of nondestructive measuring. Therefore, aiming at the shortcomings of the existing medical testing equipment, on the basis of animal experiment, this paper uses near infrared spectrum technology, based on the absorption characteristics of the material in near infrared wave band, and combined with near infrared spectrum analysis technique of qualitative determination and quantitative analysis of the characteristics. Using clustering algorithm for spinal cord injury site of neuron-specific nuclear protein and neurotransmitters classify and simulation, partial least squares (PLS) algorithm is used to calculate the content to achieve a spinal cord accurate detection of the site of injury nerve cell activity. This method provides theoretical basis for the detection of spinal cord injury, bringing hopes to patients with limb function reconstruction and rehabilitation, for clinical provides a method to detect the activity of nerve cells in the noninvasive.