Development and validation of a machine learning framework for improved resource allocation in the emergency department.

OBJECTIVE: The Emergency Severity Index (ESI) is the most commonly used system in over 70% of all U.S. emergency departments (ED) that uses predicted resource utilization as a means to triage [1], Mistriage, which includes both undertriage and overtriage has been a persistent issue, affecting 32.2% of total ED visits [2]. Our goal is to develop a machine learning framework that predicts patients' resource needs, thereby improving resource allocation during triage. METHODS: This retrospective study analyzed ED visits from the Medical Information Mart for Intensive Care IV, dividing the data into training (80%) and testing (20%) cohorts. We utilized data available during triage, including patient vital signs, age, gender, mode of arrival, medication history, and chief complaint. Azure AutoML was used to create different machine learning models trained to predict the 144 target columns including laboratory panels and imaging modalities as well as medications required during patients' ED visits. The 144 models' performance was evaluated using the area under the receiver operating characteristic curve (AUROC), F1 score, accuracy, precision and recall. RESULTS: A total of 391,472 ED visits were analyzed. 144 Voting ensemble models were created for each target. All frameworks achieved on average an AUC score of 0.82 and accuracy of 0.76. We gathered the feature importance for each target and observed that 'chief complaint', among others, had a high aggregate feature importance across different targets. CONCLUSION: This study shows the high accuracy in predicting resource needs for patients in the ED using a machine learning model. This can greatly improve patient flow and resource allocation in already resource limited emergency departments.

Aging-associated decrease of PGC-1α promotes pain chronification.

Aging is generally associated with declining somatosensory function, which seems at odds with the high prevalence of chronic pain in older people. This discrepancy is partly related to the high prevalence of degenerative diseases such as osteoarthritis in older people. However, whether aging alters pain processing in the primary somatosensory cortex (S1), and if so, whether it promotes pain chronification is largely unknown. Herein, we report that older mice displayed prolonged nociceptive behavior following nerve injury when compared with mature adult mice. The expression of peroxisome proliferator-activated receptor-gamma coactivator-1α (PGC-1α) in S1 was decreased in older mice, whereas PGC-1α haploinsufficiency promoted prolonged nociceptive behavior after nerve injury. Both aging and PGC-1α haploinsufficiency led to abnormal S1 neural dynamics, revealed by intravital two-photon calcium imaging. Manipulating S1 neural dynamics affected nociceptive behavior after nerve injury: chemogenetic inhibition of S1 interneurons aggravated nociceptive behavior in naive mice; chemogenetic activation of S1 interneurons alleviated nociceptive behavior in older mice. More interestingly, adeno-associated virus-mediated expression of PGC-1α in S1 interneurons ameliorated aging-associated chronification of nociceptive behavior as well as aging-related S1 neural dynamic changes. Taken together, our results showed that aging-associated decrease of PGC-1α promotes pain chronification, which might be harnessed to alleviate the burden of chronic pain in older individuals.

The endocannabinoid N-arachidonoyl dopamine is critical for hyperalgesia induced by chronic sleep disruption.

Chronic pain is highly prevalent and is linked to a broad range of comorbidities, including sleep disorders. Epidemiological and clinical evidence suggests that chronic sleep disruption (CSD) leads to heightened pain sensitivity, referred to as CSD-induced hyperalgesia. However, the underlying mechanisms are unclear. The thalamic reticular nucleus (TRN) has unique integrative functions in sensory processing, attention/arousal and sleep spindle generation. We report that the TRN played an important role in CSD-induced hyperalgesia in mice, through its projections to the ventroposterior region of the thalamus. Metabolomics revealed that the level of N-arachidonoyl dopamine (NADA), an endocannabinoid, was decreased in the TRN after CSD. Using a recently developed CB1 receptor (cannabinoid receptor 1) activity sensor with spatiotemporal resolution, CB1 receptor activity in the TRN was found to be decreased after CSD. Moreover, CSD-induced hyperalgesia was attenuated by local NADA administration to the TRN. Taken together, these results suggest that TRN NADA signaling is critical for CSD-induced hyperalgesia.

Immune-stimulating antibody conjugates elicit robust myeloid activation and durable antitumor immunity.

Innate pattern recognition receptor agonists, including Toll-like receptors (TLRs), alter the tumor microenvironment and prime adaptive antitumor immunity. However, TLR agonists present toxicities associated with widespread immune activation after systemic administration. To design a TLR-based therapeutic suitable for systemic delivery and capable of safely eliciting tumor-targeted responses, we developed immune-stimulating antibody conjugates (ISACs) comprising a TLR7/8 dual agonist conjugated to tumor-targeting antibodies. Systemically administered human epidermal growth factor receptor 2 (HER2)-targeted ISACs were well tolerated and triggered a localized immune response in the tumor microenvironment that resulted in tumor clearance and immunological memory. Mechanistically, ISACs required tumor antigen recognition, Fcγ-receptor-dependent phagocytosis and TLR-mediated activation to drive tumor killing by myeloid cells and subsequent T-cell-mediated antitumor immunity. ISAC-mediated immunological memory was not limited to the HER2 ISAC target antigen since ISAC-treated mice were protected from rechallenge with the HER2- parental tumor. These results provide a strong rationale for the clinical development of ISACs.

Transzonular Triamcinolone-Moxifloxacin Versus Topical Drops for the Prophylaxis of Postoperative Inflammation After Cataract Surgery.

Purpose: To compare the safety and efficacy of an intraoperative transzonular injection of triamcinolone-moxifloxacin (Imprimis' Tri-Moxi 15 mg/1 mg/mL) to topical drops in patients undergoing cataract surgery. Methods: Between January 2017 and October 2017, patients undergoing phacoemulsification cataract surgery by a single surgeon at the Penn State Eye Center were offered a single intraoperative injection of transzonular triamcinolone-moxifloxacin in lieu of perioperative drops. Between November 2017 and July 2018, the transzonular injection was not offered, and all patients who underwent surgery by the same surgeon received a drop regimen consisting of polymyxin b/trimethoprim and prednisolone acetate 1%. Patients were seen 1 day, 1 week, and 6-9 weeks postoperatively. Outcome measures included incidence of intraoperative complications, postoperative breakthrough inflammation, cystoid macular edema, and infectious sequela. Results: Of the 198 eyes, 99 from 73 patients received the injection and 99 from 82 patients received topical drops. One (1%) intraoperative posterior capsule tear occurred in each group. Eleven (11.1%) eyes in the injection group and 3 (3%) in the drop group experienced symptomatic breakthrough inflammation necessitating treatment (P = 0.0488). One (1%) eye in the injection group and zero (0%) in the drop group developed clinically significant macular edema (P = 1.0). No instances of elevated intraocular pressure or infectious sequela occurred in either group (P = 1.0). Conclusions: Transzonular injection of triamcinolone-moxifloxacin may be associated with an increased incidence of breakthrough inflammation compared to topical drops.

Pharmacological inactivation does not support a unique causal role for intraparietal sulcus in the discrimination of visual number.

The "number sense" describes the intuitive ability to quantify without counting. Single neuron recordings in non-human primates and functional imaging in humans suggest the intraparietal sulcus is an important neuroanatomical locus of numerical estimation. Other lines of inquiry implicate the IPS in numerous other functions, including attention and decision making. Here we provide a direct test of whether IPS has functional specificity for numerosity judgments. We used muscimol to reversibly and independently inactivate the ventral and lateral intraparietal areas in two monkeys performing a numerical discrimination task and a color discrimination task, roughly equilibrated for difficulty. Inactivation of either area caused parallel impairments in both tasks and no evidence of a selective deficit in numerical processing. These findings do not support a causal role for the IPS in numerical discrimination, except insofar as it also has a role in the discrimination of color. We discuss our findings in light of several alternative hypotheses of IPS function, including a role in orienting responses, a general cognitive role in attention and decision making processes and a more specific role in ordinal comparison that encompasses both number and color judgments.

Automatic characterization of user errors in spirometry.

Spirometry plays a critical role in characterizing and improving outcomes related to chronic lung disease. However, patient error in performing the spirometry maneuver, such as from coughing or taking multiple breaths, can lead to clinically misleading results. As a result, spirometry must take place under the supervision of a trained specialist who can identify and correct patient errors. To reduce the need for specialists to coach patients during spirometry, we demonstrate the ability to automatically detect four common patient errors. Creating separate machine learning classifiers for each error based on features derived from spirometry data, we were able to successfully label errors on spirometry maneuvers with an F-score between 0.85 and 0.92. Our work is a step toward reducing the need for trained individuals to administer spirometry tests by demonstrating the ability to automatically detect specific errors and provide appropriate patient feedback. This will increase the availability of spirometry, especially in low resource and telemedicine contexts.

Structural basis of the ultrasensitive calcium indicator GCaMP6.

GCaMP is one of the most widely used calcium indicators in neuronal imaging and calcium cell biology. The newly developed GCaMP6 shows superior brightness and ultrasensitivity to calcium concentration change. In this study, we determined crystal structures of Ca(2+)-bound GCaMP6 monomer and dimer and presented detailed structural analyses in comparison with its parent version GCaMP5G. Our analyses reveal the structural basis for the outperformance of this newly developed Ca(2+) indicator. Three substitution mutations and the resulting changes of local structure and interaction explain the ultrasensitivity and increased fluorescence intensity common to all three versions of GCaMP6. Each particular substitution in the three GCaMP6 is also structurally consistent with their differential sensitivity and intensity, maximizing the potential of using GCaMP6 in solving diverse problems in neuronal research and calcium signaling. Our studies shall also be beneficial to further structure-guided optimization of GCaMP and facilitate the design of novel calcium indicators.