RESUMO
Fifth Generation (5G) mobile networks introduce the concept of slicing to ensure isolation among the various supported heterogeneous services. The User Equipment (UE) can be connected to multiple slices simultaneously. Additionally, the notion of a Bandwidth Part (BWP) was also instigated to reduce power consumption. A BWP is a small chunk of the bandwidth scanned by the UE to retrieve its service data. Therefore, a UE connected to multiple services can be configured with multiple BWPs each associated with a given service. Such UEs find themselves scanning multiple BWPs, which can be time consuming and highly energy intensive. Hence, it is paramount to study the appropriate choice of the BWP configuration from an energy-efficiency perspective for multi-slice users depending on their battery level. In this paper, two energy-efficient BWP selection solutions are proposed for users connected to multiple slices. The first solution is based on a centralized approach where UEs are stirred optimally to the best BWP configuration, while the second solution relies on a user-centric distributed approach using non-cooperative game theory. The proposed schemes take into account the users' battery level and their sojourn time in the network as well as the scanned BWP size. Both solutions are compared with one another and against the legacy solution. Intensive simulation results demonstrate the efficiency of our proposition in terms of users' energy efficiency and quality of service.
RESUMO
BACKGROUND AND OBJECTIVES: The Ki-67 index has been incorporated into The World Health Organization's classification system of pancreatic neuroendocrine tumors. However, pathologists continue to question the utility of Ki-67 index over that of mitotic count as an indicator of proliferative activity. The intent of the current study is to compare K-i67 index with tumor size and mitotic rate for the association of each with lymph node metastasis and survival. METHODS: The current study is a review of 24 patients with pancreatic neuroendocrine tumors. RESULTS: Regional LNM were present in 100% of tumors with Ki-67 index >10%, while only 25% of tumors with <10% Ki-67 had LNM (P = 0.003). No tumors <2 cm had >10% Ki-67 labeling. Of patients with tumors showing ≥ 10% Ki-67 labeling, 80% died during the observation period of this study, while during the same time period, no patients with <10% Ki-67 labeling died. CONCLUSION: Ki-67 index of >10% is a sensitive indicator of malignant behavior and mortality. Future advances in management of pNETs will require development of staging guidelines with higher predictive value. Inclusion of Ki-67 labeling >10% as an indicator of aggressive disease may contribute to such improvements.