Search

Federated Learning (FL) enables multiple clients to train a shared model collaboratively without sharing any personal data. However, selecting a model and adapting it quickly to meet user expectations in a large-scale FL application with heterogeneous devices is challenging. In this paper, we propose a model selection and adaptation system for Federated Learning (FedMSA), which includes a hardware-aware model selection algorithm that trades-off model training efficiency and model performance base on FL developers' expectation. Meanwhile, considering the expected model should be achieved by dynamic model adaptation, FedMSA supports full automation in building and deployment of the FL task to different hardware at scale. Experiments on benchmark and real-world datasets demonstrate the effectiveness of the model selection algorithm of FedMSA in real devices (e.g., Raspberry Pi and Jetson nano).

Thousands of energy-aware sensors have been placed for monitoring in a variety of scenarios, such as manufacturing, control systems, disaster management, flood control and so on, requiring time-critical energy-efficient solutions to extend their lifetime. This paper proposes reinforcement learning (RL) based dynamic data streams for time-critical IoT systems in energy-aware IoT devices. The designed solution employs the Q-Learning algorithm. The proposed mechanism has the potential to adjust the data transport rate based on the amount of renewable energy resources that are available, to ensure collecting reliable data while also taking into account the sensor battery lifetime. The solution was evaluated using historical data for solar radiation levels, which shows that the proposed solution can increase the amount of transmitted data up to 23%, ensuring the continuous operation of the device.

Background & Aims: Cysts of the pancreas display a wide spectrum of histology, including inflammatory (pseudocysts), benign (serous), premalignant (mucinous), and malignant (mucinous) lesions. Endoscopic ultrasonography (EUS) may offer a diagnostic tool through the combination of imaging and guided, fine-needle aspiration (FNA). The purpose of this investigation was to determine the most accurate test for differentiating mucinous from nonmucinous cystic lesions., Methods: The results of EUS imaging, cyst fluid cytology, and cyst fluid tumor markers (CEA, CA 72-4, CA 125, CA 19-9, and CA 15-3) were prospectively collected and compared in a multicenter study using histology as the final diagnostic standard., Results: Three hundred forty-one (341) patients underwent EUS and FNA of a pancreatic cystic lesion; 112 of these patients underwent surgical resection, providing a histologic diagnosis of the cystic lesion (68 mucinous, 7 serous, 27 inflammatory, 5 endocrine, and 5 other). Receiver operator curve analysis of the tumor markers demonstrated that cyst fluid CEA (optimal cutoff of 192 ng/mL) demonstrated the greatest area under the curve (0.79) for differentiating mucinous vs. nonmucinous cystic lesions. The accuracy of CEA (88 of 111, 79%) was significantly greater than the accuracy of EUS morphology (57 of 112, 51%) or cytology (64 of 109, 59%) (P < 0.05). There was no combination of tests that provided greater accuracy than CEA alone (P < 0.0001)., Conclusions: Of tested markers, cyst fluid CEA is the most accurate test available for the diagnosis of mucinous cystic lesions of the pancreas.

The hCHK2 gene encodes the human homolog of the yeast Cds1 and Rad53 G2 checkpoint kinases, whose activation in response to DNA damage prevents cellular entry into mitosis. Here, it is shown that heterozygous germ line mutations in hCHK2 occur in Li-Fraumeni syndrome, a highly penetrant familial cancer phenotype usually associated with inherited mutations in the TP53 gene. These observations suggest that hCHK2 is a tumor suppressor gene conferring predisposition to sarcoma, breast cancer, and brain tumors, and they also provide a link between the central role of p53 inactivation in human cancer and the well-defined G2 checkpoint in yeast.