Search

Study Design This was a retrospective cohort study. Purpose We evaluated the predictive value of the Spinal Instability Neoplastic Score (SINS) system for spinal adverse events (SAEs) in patients with single spinal metastatic tumor. Overview of Literature The SINS system was developed to assess spinal instability in patients with single metastatic spinal tumor. However, the system’s potential predictive value for SAEs has been partially studied. Methods This system was applied to a retrospective cohort of 78 patients with single spinal metastatic tumors. The patients underwent surgical treatment and were postoperatively followed up for at least 2 years or until death. The attribution of each score and total SINS to SAE (vertebral compression fracture [VCF] and spinal cord compression [SCC]) occurrence was assessed using the Cox proportional hazards model. Results SAEs occurred on average 7 months after diagnosis of spinal metastasis. The mean survival rate post diagnosis was 43 months. Multivariate analysis using the Cox proportional hazards model revealed that the pain (p=0.029) and spinal alignment (p=0.001) scores were significantly related to VCF occurrence, whereas the pain (p=0.008) and posterolateral involvement (p=0.009) scores were related to SCC occurrence. Conclusions Among the components of the SINS system, while pain and spinal alignment showed a significant association with VCF occurrence, pain and posterolateral involvement showed association with SCC occurrence.

Techniques for detecting leakage in water pipe networks have been developed worldwide in order to reduce unaccounted-for water quantity and enhance the reliability of the pipe networks. In this paper computational algorithms utilizing principal component analysis (PCA) were developed so that the algorithms can be used in a realistic water pipe network management situation in which the daily flow data of a district metered area (DMA) are needed to be verified for a possible relation with a water leak incident. For the improvement of the algorithms, it was assumed that a manager of a water pipe network uses these algorithms every day to test if yesterday’s inflow data to a DMA were an outlier according to the PCA computational algorithm. The flow data used in this study were analyzed to determine the best flow data size for the field use of the developed PCA algorithm. For various flow data sets, which were defined as the smaller sizes of the flow measured in days than the whole data set available, a reference modeling for the PCA was applied to calculate the model outliers by moving the flow data sets day by day. For each DMA the effective outlier detection rates (EODRs) were calculated for the whole range of the defined time windows. The maximum effective outlier detection rate for a DMA was obtained as the maximum of the calculated EODRs. The process and results of the sensitivity analyses of the model parameters were used to suggest guidance on how to determine model parameters for a given flow data. [ABSTRACT FROM AUTHOR]