To analyze the effectiveness and safety of zoledronic acid combined with chemotherapy for lung cancer spinal metastases, 96 patients with lung cancer spinal metastases were averagely classified into the experimental group (gemcitabine, cisplatin, and zoledronic acid) and the control group (gemcitabine and cisplatin). An optimized noise variance estimation algorithm (OMAPB) was proposed based on the maximum a posteriori Bayesian method (MAPB), and the algorithm was applied to the patient’s computed tomography (CT) scan. The results indicated that in terms of curative effect, the number of complete remission (CR), partial remission (PR) cases, effective rate, and clinical benefit rate of the test group was significantly higher than those of the control group. The number of progress disease (PD) cases was significantly lower than that of the control group ( P < 0.05 ). The disease progression time of the test group patients was 6.2 months, and the disease progression time of the control group patients was 3.7 months ( P < 0.05 ). The test group patients had 8 cases of bone marrow suppression and gastrointestinal reactions after treatment. In the test group, there were 8 cases of bone marrow suppression, 9 cases of gastrointestinal reaction, 3 cases of fever, 4 cases of pain, and 2 cases of hair loss. The patients in the control group were complicated with bone marrow suppression in 14 cases, gastrointestinal reaction in 17 cases, fever in 5 cases, pain in 4 cases, and hair loss in 6 cases. The difference was statistically significant ( P < 0.05 ). It showed that zoledronic acid combined with chemotherapy could effectively improve the treatment efficiency and clinical benefit rate of patients with lung cancer spinal metastases, prolong the progression of the disease, reduce the degree of bone tissue damage, and would not increase chemotherapy adverse events.