Portland General Electric (PGE) is a Fortune 1000 public utility based in Portland, Oregon. It distributes electricity to customers in parts of Multnomah, Clackamas, Marion, Yamhill, Washington, and Polk counties - 44% of the inhabitants of Oregon. Founded in 1888 as the Willamette Falls Electric Company, the company has been an independent company for most of its existence, though was briefly owned by the Houston-based Enron Corporation from 1997 until 2006 when Enron divested itself of PGE during its bankruptcy. Notably, PGE does not serve all of Portland. Its service territory comprises most of Portland west of the Willamette River, sharing most of the city east of the river with Pacific Power. PGE produces and purchases energy primarily from coal and natural gas plants, as well as hydroelectric power from dams on the Clackamas, Willamette and Deschutes rivers. Between 1976 and 1993, PGE operated Trojan, the only nuclear power plant in Oregon. Trojan was the subject of three Oregon initiatives to shut it down. The initiatives failed, but the company elected to close the plant twenty years early.
A multi-agent GenAI architecture is introduced to support the automation of ethical cloud security, solving the problem of scalability or adaptability, and compliance in dynamic cloud environments. This framework combines dedicated generative agents such as policy analysts, threat detectors, remediation organizers, and auditor agents which interact via common knowledge graph and can be explained by a decision log. The agents utilize context-based prompt generation, generation constraints, provenance management, and generation to generate security policies, anomaly detection, provide automated mitigation, and maintain human-in-the-loop control. Some of the ethical protections are bias audits, privacy-sensitive learning, least-privilege enforcement, and policy verifiability, to warrant the correctness of the decisions taken in compliance with regulatory or organizational limits. Testing with representative cloud work lines shows that there is shorter energy on discerning and correcting occurrences, elevated coverage of controls and signs when the automated actions are traced. The framework enables adjustable levels of trust and escalation measures to accommodate the autonomy versus governance. The method fosters usage of GenAI to deal with cloud security by offering modular agents, verifiable ethics controls, metrics-based assessment, hence fostering responsible automation that is also transparent, auditable, and considers changing threats.
Modern digital services are built on virtualized cloud infrastructures that are vulnerable to the fast-changing vulnerabilities made by complex software stacks and dynamic resource orchestration. Traditional vulnerability assessment methods have strong dependence on signature databases and reactive scanning tendencies that cannot predict zero-day exploits and configuration-based weaknesses. The paper presents a GenAI-based vulnerability prediction paradigm dedicated to a virtualized cloud setup, combining contextual infrastructure intelligence with the ability to generate models. It uses a scheme that trains on a massive vector of homogeneous security artifacts, such as configuration energies, runtime metadata, and past paths of vulnerabilities ambiguity, to predict latent outliers of risk distribution prior to the appearance of exploits. The approach actively determines high-risk virtual machines and hypervisor layers by combing possible attack patterns and adaptive threat situations. The design focuses on scalability, explainability and compatibility with multi-tenant cloud platforms. In comparison with traditional and learning-based vulnerability assessment techniques, it is shown to exhibit a better predictive coverage and less false alarms when subjected to dynamic workload conditions. The proposed paradigm will advance cloud security management by replacing vulnerability mitigation through reactive monitoring with proactive risk intelligence, enabling robust and secure operation of next-generation virtualized cloud system infrastructures.
AIMS:Quantitative stress perfusion (QP) cardiac magnetic resonance (CMR) can be performed using the dual sequence (DS) or dual bolus (DB) technique. DS does not require additional contrast and image acquisition but needs a research sequence. DB can be performed on all magnetic resonance imaging (MRI) scanners with standard perfusion sequences but requires additional contrast injection and image acquisition. Our aim was to compare the prognostic significance of DB and DS. METHODS AND RESULTS:DB and DS were performed on the same patient and the same examination. Analysts were blinded to clinical outcomes. Stress myocardial blood flow (MBF) and myocardial perfusion reserve (MPR) were quantified. The primary outcome was a composite of major adverse cardiovascular events (MACE) comprising acute coronary syndrome, stroke, heart failure (HF) hospitalization, late revascularization, and all-cause death. 570 patients (mean age: 63.2 ± 12.3 years; 61.2% male) were recruited. Median follow-up was 743 days; 54 events were documented. All QP CMR variables demonstrated significance in univariate Cox regression [DB stress MBF [HR = 0.53 (95%CI:0.35-0.78)], DB MPR [HR = 0.38 (95%CI:0.22-0.66)], DS stress MBF [HR = 0.27 (95%CI:0.18-0.40)] and DS MPR [HR = 0.19 (95%CI:0.13-0.29)]]. On multivariable Cox regression models, only DB MPR, DS MBF, and DS MPR remained significant for MACE (HR = 0.50 (95%CI:0.28-0.89), HR = 0.35 (95%CI:0.23-0.53); HR = 0.23 (95%CI 0.15-0.36), respectively). Harrell's C-index of DS MPR and DS stress MBF showed significantly better prognostication than their DB counterparts (P < 0.001 and P = 0.012, respectively). CONCLUSION:In this blinded comparison, DS stress MBF and MPR demonstrated better prognostication than DB stress MBF and MPR. Our findings support DS as the preferred approach where available.
