Automated Endotracheal Tube Placement Check Using Semantically Embedded Deep Neural Networks.

RATIONALE AND OBJECTIVES: To develop artificial intelligence (AI) system that assists in checking endotracheal tube (ETT) placement on chest X-rays (CXRs) and evaluate whether it can move into clinical validation as a quality improvement tool. MATERIALS AND METHODS: A retrospective data set including 2000 de-identified images from intensive care unit patients was split into 1488 for training and 512 for testing. AI was developed to automatically identify the ETT, trachea, and carina using semantically embedded neural networks that combine a declarative knowledge base with deep neural networks. To check the ETT tip placement, a "safe zone" was computed as the region inside the trachea and 3-7 cm above the carina. Two AI outputs were evaluated: (1) ETT overlay, (2) ETT misplacement alert messages. Clinically relevant performance metrics were compared against prespecified thresholds of >85% overlay accuracy and positive predictive value (PPV) > 30% and negative predictive value NPV > 95% for alerts to move into clinical validation. RESULTS: An ETT was present in 285 of 512 test cases. The AI detected 95% (271/285) of ETTs, 233 (86%) of these with accurate tip localization. The system (correctly) did not generate an ETT overlay in 221/227 CXRs where the tube was absent for an overall overlay accuracy of 89% (454/512). The alert messages indicating that either the ETT was misplaced or not detected had a PPV of 83% (265/320) and NPV of 98% (188/192). CONCLUSION: The chest X-ray AI met prespecified performance thresholds to move into clinical validation.

Lung Cancer Screening Update.

Since the release of the US Preventive Services Task Force and Centers for Medicare and Medicaid Services recommendations for lung cancer screening, low-dose chest computed tomography screening has moved from the research arena to clinical practice. Lung cancer screening programs must reach beyond image acquisition and interpretation and engage in a multidisciplinary effort of clinical shared decision-making, standardization of imaging and nodule management, smoking cessation, and patient follow-up. Standardization of radiologic reports and nodule management will systematize patient care, provide quality assurance, further reduce harm, and contain health care costs. Although the National Lung Screening Trial results and eligibility criteria of a heavy smoking history are the foundation for the standard guidelines for low-dose chest computed tomography screening in the United States, currently only 27% of patients diagnosed with lung cancer would meet US lung cancer screening recommendations. Current and future efforts must be directed to better delineate those patients who would most benefit from screening and to ensure that the benefits of screening reach all socioeconomic strata and racial and ethnic minorities. Further optimization of lung cancer screening program design and patient eligibility will assure that lung cancer screening benefits will outweigh the potential risks to our patients.

Lithium activates the Wnt and phosphatidylinositol 3-kinase Akt signaling pathways to promote cell survival in the absence of soluble survival factors.

Mouse proximal tubular cells (BUMPT), when cultured in the absence of growth factors, activate a default apoptotic pathway. Although Wnt signaling antagonizes the effect of proapoptotic triggers, its role in regulating the default pathway of apoptosis is less well defined. The present study examines the hypothesis that lithium (Li(+)) and (2'Z,3'E)-6-bromoindirubin-3'-oxime (BIO), two glycogen synthase kinase-3beta (GSK3beta) inhibitors, promote survival of growth factor-deprived renal epithelial cells by activating the Wnt pathway. These studies demonstrate that Li(+) and BIO activate Wnt signaling as indicated by the following changes: phosphorylation (inhibition) of GSK3beta; decreased phosphorylation of beta-catenin (a GSK3beta substrate); nuclear translocation of beta-catenin; specific transcriptional activation of Tcf/catenin-responsive pTopflash constructs; and an increase in the expression of cyclin D1 (indicative of a promitogenic cell response). In addition, Li(+) or BIO significantly increases the phosphorylation (activation) of Akt, an anti-apoptotic protein, and inhibits apoptosis (decreases both annexin-V staining and caspase-3 activation), during serum deprivation. Inhibition of phosphatidylinositol 3-kinase (responsible for Akt activation) either by wortmanin or LY-294002 prevented Li(+)- or BIO-induced Akt phosphorylation and reduces cell survival without altering the phosphorylation state of GSK3beta. Li(+) or BIO also increases the expression of insulin-like growth factor-II (IGF-II), a potent proliferative signaling protein. Li(+) or BIO-free conditioned medium harvested from Li(+)- or BIO-exposed cells also induced Akt phosphorylation, mimicking the protective effect of the two GSK3beta inhibitors on serum-starved cells. Furthermore, the effect of conditioned medium on Akt phosphorylation could be inhibited by either LY-294002 or IGF-binding protein. BIO, a specific GSK3beta inhibitor, replicated the protective effect of Li(+) on cell viability, suggesting that GSK3beta activation is important for initiating the apoptotic pathway. Taken together, these data suggest that Li(+) or BIO promotes renal epithelial cell survival by inhibiting apoptosis through GSK3beta-dependent activation of the Wnt pathway and subsequent release of IGF-II. Extracellular IGF-II serves as an autocrine survival factor that is responsible, in part, for activating the anti-apoptotic phosphatidylinositol-3-kinase-Akt pathway during serum deprivation.

Hsp72 interacts with paxillin and facilitates the reassembly of focal adhesions during recovery from ATP depletion.

The cytoprotective effect of heat stress proteins on epithelial cell detachment, an important cause of acute, ischemic renal failure, was examined after ATP depletion by evaluating focal adhesion complex (FAC) integrity. The intracellular distribution of FAC proteins (paxillin, talin, and vinculin) was assessed by immunohistochemistry before, during, and after exposure of renal epithelial cells to metabolic inhibitors. The resulting ATP depletion caused reversible re-distribution of all three proteins from focal adhesions to the cytosol. Paxillin, a key adaptor protein, was selected as a surrogate marker for FAC integrity in subsequent studies. Prior heat stress increased hsp72, a molecular chaperone, in both the Triton X-100-soluble and -insoluble protein fractions. Compared with ATP depleted control, heat stress significantly decreased paxillin and hsp72 shift from the Triton X-100 soluble to the insoluble protein fraction (an established marker of denaturation and aggregation); increased paxillin-hsp72 interaction detected by co-immunoprecipitation; enhanced paxillin extractability from Triton X-100-insoluble precipitates, increased the reformation of focal adhesions, and improved cell attachment (p < 0.05). To determine whether hsp72 mediates protection afforded by heat stress, cells were infected with adenovirus containing human hsp72 or empty vector. Hsp72 overexpression increased its interaction with paxillin and improved focal adhesion reformation during recovery, mimicking the protective effects of heat stress. These data suggest that hsp72 facilitates the reassembly of focal adhesions and improves cell attachment by reducing paxillin denaturation and increasing its re-solubilization after ATP depletion.