BackgroundElastomeric disc replacements have been developed to restore normal shock absorption and physiologic centers of rotation to the degenerated disc. The Physio-L Artificial Lumbar Disc is an elastomeric disc which uses a compliant polycarbonate-polyurethane core with enhanced endurance properties. The objective of this study was to evaluate the safety and efficacy of the Physio-L through a 12-month follow-up period in a prospective, nonrandomized clinical trial.MethodsTwelve patients who met the inclusion/exclusion criteria were enrolled in the study. Eight patients received a single implant (L5-S1) and 4 received a 2-level implantation (L4-5 and L5-S1). Patients were assessed preoperatively and postoperatively at 6 weeks and 3, 6, and 12 months. Primary outcomes included the VAS, ODI, a radiographic analysis of implant condition, incidence of major complications, and reoperations. Secondary outcomes included SF-36, ROM at index and adjacent levels and disc height.ResultsAll patients completed the 12-month follow-up evaluations. Through 12 months, the Physio-L devices have remained intact with no evidence of subsidence, migration, or expulsion. VAS low-back pain and ODI scores improved significantly at all follow-up periods compared to preoperative scores. The range of motion of 13.3° ± 5.5° at the index level was considered normal. Overall, patients were satisfied with an average score of 83.5 ± 26.8 mm. When comparing the device to other artificial discs, the current device showed a clinically relevant improvement in both ODI and VAS scores at all follow-up time points. Statistically significant improvements in both scores were observed at 12 months (P < .05).ConclusionThe Physio-L is safe and efficacious, as demonstrated by improved pain relief and functional recovery without any implant failures, significant device related complications, or adverse incidents. The clinical results for VAS and ODI were superior to other marketed artificial lumbar discs such as the Charité and ProDisc-L at the same follow-up timeframes.

Abstract. Building effective optimization heuristics is a challenging task which often takes developers several months if not years to complete. Predictive modelling has recently emerged as a promising solution, automatically constructing heuristics from training data. However, obtaining this data can take months per platform. This is becoming an ever more critical problem and if no solution is found we shall be left with out of date heuristics which cannot extract the best performance from modern machines. In this work, we present a low-cost predictive modelling approach for automatic heuristic construction which significantly reduces this training overhead. Typically in supervised learning the training instances are randomly selected to evaluate regardless of how much useful information they carry. This wastes effort on parts of the space that contribute little to the quality of the produced heuristic. Our approach, on the other hand, uses active learning to select and only focus on the most useful training examples. We demonstrate this technique by automatically constructing a model to determine on which device to execute four parallel programs at differing problem dimensions for a representative Cpu-Gpu based heterogeneous system. Our methodology is remarkably simple and yet effective, making it a strong candidate for wide adoption. At high levels of classification accuracy the average learning speed-up is 3x, as compared to the stateof-the-art.

Building effective optimization heuristics is a challenging task which often takes developers several months if not years to complete. Predictive modelling has recently emerged as a promising solution, automatically constructing heuristics from training data, however, obtaining this data can take months per platform. This is becoming an ever more critical problem as the pace of change in architecture increases. Indeed, if no solution is found we shall be left with out of date heuristics which cannot extract the best performance from modern machines.In this work, we present a low-cost predictive modelling approach for automatic heuristic construction which significantly reduces this training overhead. Typically in supervised learning the training instances are randomly selected to evaluate regardless of how much useful information they carry, but this wastes effort on parts of the space that contribute little to the quality of the produced heuristic. Our approach, on the other hand, uses active learning to select and only focus on the most useful training examples and thus reduces the training overhead.We demonstrate this technique by automatically creating a model to determine on which device to execute four parallel programs at differing problem dimensions for a representative Cpu-Gpu based system. Our methodology is remarkably simple and yet effective, making it a strong candidate for wide adoption. At high levels of classification accuracy the average learning speed-up is 3x, as compared to the state-of-the-art.

No abstract