Patient outcomes of osteosarcoma vary because of tumor heterogeneity and treatment strategies. This study aimed to compare the performance of multiple machine learning (ML) models with the traditional Cox proportional hazards (CoxPH) model in predicting prognosis and explored the potential of ML models in clinical decision-making. From 2000 to 2018, 1243 patients with osteosarcoma were collected from the Surveillance, Epidemiology, and End Results (SEER) database. Three ML methods were chosen for model development (DeepSurv, neural multi-task logistic regression [NMTLR]) and random survival forest [RSF]) and compared them with the traditional CoxPH model and TNM staging systems. 871 samples were used for model training, and the rest were used for model validation. The models’ overall performance and predictive accuracy for 3- and 5-year survival were assessed by several metrics, including the concordance index (C-index), the Integrated Brier Score (IBS), receiver operating characteristic curves (ROC), area under the ROC curves (AUC), calibration curves, and decision curve analysis. The efficacy of personalized recommendations by ML models was evaluated by the survival curves. The performance was highest in the DeepSurv model (C-index, 0.77; IBS, 0.14; 3-year AUC, 0.80; 5-year AUC, 0.78) compared with other methods (C-index, 0.73–0.74; IBS, 0.16–0.17; 3-year AUC, 0.73–0.78; 5-year AUC, 0.72–0.78). There are also significant differences in survival outcomes between patients who align with the treatment option recommended by the DeepSurv model and those who do not (hazard ratio, 1.88; P < .05). The DeepSurv model is available in an approachable web app format at https://survivalofosteosarcoma.streamlit.app/. We developed ML models capable of accurately predicting the survival of osteosarcoma, which can provide useful information for decision-making regarding the appropriate treatment.

Background： This study aimed to compare the clinical results between the Wiltse approach and traditional approach in lumbar spine fusion. Methods: In this retrospective controlled study, from May 2016 to May 2017, 70 cases with lumbar spine fusion surgery for lumbar disc herniation or lumbar spondylolisthesis within Ⅱdegree were assigned to Wiltse approach (35 cases,18 males and 17 females; mean age 52 years) and traditional approach groups (35cases, 19 males, 16 females; 51 years). There were totally 38 cases of lumbar disc herniation and 32 cases of II- degree spondylolisthesis in two groups. The operation time, the amount of blood loss and post-operation drainage, the level of creatine phosphokinase (CK) and the reduction of cross sectional area (CSA) in magnetic resonance imaging (MRI) were recorded. Results: After the operation for the Wiltse approach group, there was less operation time, blood loss and post-operative drainage (p＜0.05) ;the peripheral blood CK levels was changed a lot in 1 day and 3 days after the operation. In the Wiltse group, the reduction of CSA was significantly lower than traditional approach group in MRI. Conclusion: For lumbar spinal fusion surgery, the Wiltse approach do have shorter operation time and smaller vertebral side muscle injury compared with the traditional approach. However, in determining the operation program, the performer should fully recognize that the Wiltse clearance may have different influence on the operation at different anatomical level.

Background：This study aim to compare the clinical results such as functional improvement, complications, intraoperative parameters etc. between the Wiltse approach and traditional approach in lumbar spine fusion. Methods: In this retrospective controlled study, from May 2016 to May 2017, 70 patients undergone lumbar spine fusion surgery for lumbar disc herniation or lumbar spondylolisthesis within degree II were randomly divided into 2 groups which include Wiltse approach group (N=35, M:F= 1.06:1, mean age= 52 years) and traditional approach group (N=35, M:F= 1.18:1, mean age= 51 years). Actually, all the 70 cases were consisted of 38 lumbar-disc-herniation patients and 32 degree II lumbar-spondylolisthesis patients. In order to specify the clinical curative effects of the two group patients, several indicators including the operation time, the amount of blood loss and post-operation drainage, the ODI, the concentration of peripheral blood creatine phosphokinase (CK) and the cross sectional area (CSA) in magnetic resonance imaging (MRI) were compared. Results: Compared with traditional group, the Wiltse approach group had a statistically significant small figure on operative time, intraoperative blood loss, post-operative drainage and post-operative ODI (p＜0.05).As for the peripheral blood CK levels, the concentration of the Wiltse approach group was lower in 1 day and 3 days after the operation(p＜0.05), while the difference of the CK concentration for 7 days post-operation has no statistical significance(p>0.05). Furthermore, for the MRI CSA three months after the operation , the Wiltse approach side,s was bigger than the traditional side,s in the Wiltse approach group; the Wiltse approach group,s was bigger than the traditional group,s of the two group,s non-decompression side(p＜0.05).Conclusion: For lumbar spinal fusion surgery, the Wiltse approach do have such advantages as shorter operation time, smaller vertebral side muscle injury and better lumbar functional recovery compared with the traditional approach, but as a surgeon, while determining operation program, we should take into account the may influence of the inconsistency of paravertebral muscle anatomy in different levels.