The quality of Magnetic Resonance Image (MRI) determines the accuracy of clinical diagnosis. It provides information about the human soft tissue anatomy. MRI of spine is used by the physicians to evaluate any presence of diseases including slipped disk, herniated disk, trauma and disk degeneration. Existence of noises and artifacts can degrade the quality of the MR images. Thus, appropriate image processing techniques may help to improve the quality of the acquired image. Preprocessing is usually done to remove the noise, enhance an image boundary and adjust the image contrast. Current techniques to enhance and reduce noise in MRI human spine are discussed and a method using discrete wavelet transform to enhance the MRI of human spine is proposed. The resultant images are evaluated quantitatively. This study shows that the proposed method has better results as compared to other existing method based on evaluation tests.