Erratum: Author Correction: Human-machine partnership with artificial intelligence for chest radiograph diagnosis.

[This corrects the article DOI: 10.1038/s41746-019-0189-7.].

Human-machine partnership with artificial intelligence for chest radiograph diagnosis.

Human-in-the-loop (HITL) AI may enable an ideal symbiosis of human experts and AI models, harnessing the advantages of both while at the same time overcoming their respective limitations. The purpose of this study was to investigate a novel collective intelligence technology designed to amplify the diagnostic accuracy of networked human groups by forming real-time systems modeled on biological swarms. Using small groups of radiologists, the swarm-based technology was applied to the diagnosis of pneumonia on chest radiographs and compared against human experts alone, as well as two state-of-the-art deep learning AI models. Our work demonstrates that both the swarm-based technology and deep-learning technology achieved superior diagnostic accuracy than the human experts alone. Our work further demonstrates that when used in combination, the swarm-based technology and deep-learning technology outperformed either method alone. The superior diagnostic accuracy of the combined HITL AI solution compared to radiologists and AI alone has broad implications for the surging clinical AI deployment and implementation strategies in future practice.

Lagrangian simulation of oil trajectories in the Florida Straits.

A Lagrangian floating oil trajectory model is developed to simulate a continuous surface spill in the Florida Straits. The trajectories are initiated at locations representative of exploratory drilling sites around Cuba's Exclusive Economic Zone and integrated for ten days. The simulation is initialized for different seasons, using leeway-corrected observed winds, ocean currents, different climatologies, and a first-order Markov Lagrangian Stochastic Model. A Monte-Carlo scheme is used to parameterize oil weathering processes collectively. Overall, the simulations exhibit a strong seasonal dependence, where the Florida coast is most affected during the summer and the Cuban coast in the winter. Drilling sites at the center of the Straits show the largest impact on Florida. Cuba is most affected by shoreline locations. A significant amount of oil reaches the Florida coastline within two to ten days, and Cuba is potentially affected within hours. Many simulations project impacts for the Florida Keys, South Florida, and the north shore of Cuba.

Persistent low-back pain is real. However, diagnostic spinal injections are not helpful in its evaluation.

We endeavored to assess the short-term effects of intrathecal fentanyl and lidocaine in chronic-pain patients by ascertaining whether the opioid fentanyl, by virtue of its lack of sensory and motor paralysis, conferred any diagnostic advantages over lidocaine, a local anesthetic whose effects include sensory and motor paralysis. Neuraxial administration of fentanyl has been touted as an improved diagnostic tool to distinguish between peripheral and central pain, because the absence of sensory and motor effects may avert the patient's presumption of the onset of analgesia based on these cues. Twenty-two patients with persistent low-back pain, whose investigations had determined that they were not surgical candidates, were studied using a counter-balanced, placebo-controlled, and double-blinded crossover design. Each patient received three separate lumbar intrathecal injections of equal volume (1.4 ml): cerebrospinal fluid, fentanyl 25 micrograms, and lidocaine 70 mg. Pain and symptom assessments were performed preinjection (baseline), and at regular intervals up to and including 4 h postinjection. Pain was evaluated by verbal patient response using a numerical pain-rating system of 0 to 10. Duration of analgesia, sensation of warmth, and adverse effects were noted. Statistical analyses were performed using nonparametric tests. Subjects' average age was 56 years, with a median low-back pain duration of 16 years. There were no significant differences in the baseline median-pain scores among injection types. The baseline and best cerebrospinal fluid-pain scores were significantly different, suggesting a placebo effect. The best pain scores for fentanyl and lidocaine were superior to their own baseline levels and to the best cerebrospinal fluid scores.(ABSTRACT TRUNCATED AT 250 WORDS)

Chronic pain and spinal cord injury.

With the medical progress that has given spinal cord injured individuals greater longevity and better overall health, chronic pain has emerged as a major challenge in treating this population. Over the past 40 years, estimates of prevalence of severe/disabling chronic pain in spinal cord injury (SCI) patients have ranged from 18% to 63%. Beyond this finding, the extant literature is extremely limited. This review summarizes the empirical findings with regard to the prevalence and clinical significance of chronic pain in the SCI population. In spite of widespread clinical beliefs, there is little evidence that characteristics of the SCI such as the level, completeness, or etiology of the injury are associated with either the development or severity of pain. Until recently, psychosocial issues have been almost totally ignored in spite of the importance such variables have demonstrated in chronic pain in other populations. A major purpose of the present article is to expand the scope of inquiry to include these factors and to emphasize the importance of employing a biopsychosocial model. Evidence is reviewed which suggests that chronic pain is associated with psychosocial impairment in this population. It is concluded that rather than being a minor problem in comparison to the other limitations imposed by SCI, chronic pain represents a significant additional challenge to the SCI patient that may be best addressed by a multidisciplinary approach.

Drinking-related locus of control and the drinking status of urban Native Americans.

Although promising, the available data concerning drinking-related locus of control have been almost entirely limited to treatment samples of white men. The major purpose of this study was to extend our understanding by comparing the drinking-related expectancies of three groups of Native Americans: problem drinkers, nonproblem drinkers and recovered alcoholics. Multivariate analyses were employed to control statistically for important sociodemographic differences between groups and included a test of the possibility that sex moderated the relationship between drinking status and drinking expectancies. As predicted, problem drinkers reported significantly less personal control of alcohol use than either nonproblem drinkers or recovered alcoholics. There were no significant differences between nonproblem drinkers and recovered alcoholics. Native American men were found to hold significantly more external orientations towards drinking than did women. The results were discussed in terms of comparisons with the available literature, and future research needs were identified. The findings provide indirect support for current treatment philosophies that seek to modify patient perceptions of the controllability of drinking behavior in a more internal direction.