Harrison Medical Center

The rapid integration of Artificial Intelligence (AI) into everyday social systems has reshaped long-standing sociological understandings of power, identity, and human agency. This paper explores AI not merely as a technological tool but as an emerging social actor capable of influencing behaviours, shaping decision-making processes, and redefining institutional practices. Drawing on theories of symbolic interactionism, actor–network theory, and critical sociology, the study examines how AI systems mediate social interactions, produce new forms of algorithmic authority, and contribute to shifting power relations between individuals, organisations, and the state. The analysis highlights how AI-driven classifications, predictions, and automated decisions reshape identities—through profiling, personalisation, and digital surveillance—while also raising concerns over autonomy, inequality, and ethical accountability. By conceptualising AI as a socially embedded actor, the paper argues that AI technologies have begun to co-produce social realities, redistribute control, and challenge the boundaries between human and machine agency. This reconfiguration demands renewed sociological attention toward digital governance, transparency, and the societal impacts of algorithmic systems in increasingly automated environments.

This article implements rate-calculation methods in the open literature and evaluates each of them by comparing the calculated results to the upset rates observed on-orbit for missions from 1987 to the present-day. Within the radiation effects community, the upset rates are typically calculated to give a worst-case result. Keeping this in mind, the models are optimized to provide the greatest precision while still over-estimating the rate in 90% of the datasets. The results indicate that a number of simple models that approximate the device data and environment do a good job in calculating the upset rate. Further, more complicated methods are not necessarily better and produce calculations with wider error bars and be more likely to underestimate the rate. In all cases, the implementation of the model is as important as the models used.

Late gadolinium-enhanced (LGE) MRI has become a widely used technique to non-invasively image the left atrium prior to catheter ablation. However, LGE-MRI images are prone to variable image quality, with quality metrics that do not necessarily correlate to the image's diagnostic quality. In this study, we aimed to define consistent clinically relevant metrics for image and diagnostic quality in 3D LGE-MRI images of the left atrium, have multiple observers assess LGE-MRI image quality to identify key features that measure quality and intra/inter-observer variabilities, and train and test a CNN to assess image quality automatically. We identified four image quality categories that impact fibrosis assessment in LGE-MRI images and trained individuals to score 50 consecutive pre-ablation atrial fibrillation LGE-MRI scans from the University of Utah hospital image database. The trained individuals then scored 146 additional scans, which were used to train a convolutional neural network (CNN) to assess diagnostic quality. There was excellent agreement among trained observers when scoring LGE-MRI scans, with inter-rater reliability scores ranging from 0.65 to 0.76 for each category. When the quality scores were converted to a binary diagnostic/non-diagnostic, the CNN achieved a sensitivity of and a specificity of . The use of a training document with reference examples helped raters achieve excellent agreement in their quality scores. The CNN gave a reasonably accurate classification of diagnostic or non-diagnostic 3D LGE-MRI images of the left atrium, despite the use of a relatively small training set.

Pediatric fractures are common in Malawi, and surgical care, when needed, remains inaccessible to many. Understanding which children in Malawi receive surgery or nonsurgical treatment would help set priorities for trauma system development. We used multivariate logistic regression to evaluate associations between surgical treatment and age, sex, school enrollment, injury mechanism, fracture type, open fracture, referral status, hospital of presentation, delayed presentation (≥2 days), healthcare provider, and inpatient vs outpatient treatment. From 2016 to 2020, 10,400 pediatric fractures were recorded in the Malawi Fracture Registry. Fractures were most commonly of the wrist (26%), forearm (17%), and elbow (14%). Surgical fixation was performed on 4.0% of patients, and 24 (13.0%) open fractures were treated nonsurgically, without débridement or fixation. Fractures of the proximal and diaphyseal humerus (odds ratio [OR], 3.72; 95% confidence interval [CI], 2.36 to 5.87), knee (OR, 3.16; 95% CI, 1.68 to 5.95), and ankle (OR, 2.63; 95% CI, 1.49 to 4.63) had highest odds of surgery. Odds of surgical treatment were lower for children referred from another facility (OR, 0.62; 95% CI, 0.49 to 0.77). Most Malawian children with fractures are treated nonsurgically, including many who may benefit from surgery. There is a need to increase surgical capacity, optimize referral patterns, and standardize fracture management in Malawi.