The Future of Collaborative Human-Artificial Intelligence Decision-Making for Mission Planning.

In an increasingly complex military operating environment, next generation wargaming platforms can reduce risk, decrease operating costs, and improve overall outcomes. Novel Artificial Intelligence (AI) enabled wargaming approaches, based on software platforms with multimodal interaction and visualization capacity, are essential to provide the decision-making flexibility and adaptability required to meet current and emerging realities of warfighting. We highlight three areas of development for future warfighter-machine interfaces: AI-directed decisional guidance, computationally informed decision-making, and realistic representations of decision spaces. Progress in these areas will enable development of effective human-AI collaborative decision-making, to meet the increasing scale and complexity of today's battlespace.

Autonomous Flying With Neuromorphic Sensing.

Autonomous flight for large aircraft appears to be within our reach. However, launching autonomous systems for everyday missions still requires an immense interdisciplinary research effort supported by pointed policies and funding. We believe that concerted endeavors in the fields of neuroscience, mathematics, sensor physics, robotics, and computer science are needed to address remaining crucial scientific challenges. In this paper, we argue for a bio-inspired approach to solve autonomous flying challenges, outline the frontier of sensing, data processing, and flight control within a neuromorphic paradigm, and chart directions of research needed to achieve operational capabilities comparable to those we observe in nature. One central problem of neuromorphic computing is learning. In biological systems, learning is achieved by adaptive and relativistic information acquisition characterized by near-continuous information retrieval with variable rates and sparsity. This results in both energy and computational resource savings being an inspiration for autonomous systems. We consider pertinent features of insect, bat and bird flight behavior as examples to address various vital aspects of autonomous flight. Insects exhibit sophisticated flight dynamics with comparatively reduced complexity of the brain. They represent excellent objects for the study of navigation and flight control. Bats and birds enable more complex models of attention and point to the importance of active sensing for conducting more complex missions. The implementation of neuromorphic paradigms for autonomous flight will require fundamental changes in both traditional hardware and software. We provide recommendations for sensor hardware and processing algorithm development to enable energy efficient and computationally effective flight control.

Abrupt darkening under high dynamic range (HDR) luminance invokes facilitation for high-contrast targets and grouping by luminance similarity.

When scanning across a scene, luminance can vary by up to 100,000-to-1 (high dynamic range, HDR), requiring multiple normalizing mechanisms spanning from the retina to the cortex to support visual acuity and recognition. Vision models based on standard dynamic range (SDR) luminance contrast ratios below 100-to-1 have limited ability to generalize to real-world scenes with HDR luminance. To characterize how orientation and luminance are linked in brain mechanisms for luminance normalization, we measured orientation discrimination of Gabor targets under HDR luminance dynamics. We report a novel phenomenon, that abrupt 10- to 100-fold darkening engages contextual facilitation, distorting the apparent orientation of a high-contrast central target. Surprisingly, facilitation was influenced by grouping by luminance similarity, as well as by the degree of luminance variability in the surround. These results challenge vision models based solely on activity normalization and raise new questions that will lead to models that perform better in real-world scenes.

A 100,000-to-1 high dynamic range (HDR) luminance display for investigating visual perception under real-world luminance dynamics.

BACKGROUND: Real-world illumination challenges both autonomous sensing and displays, because scene luminance can vary by up to 109-to-1, whereas vision models have limited ability to generalize beyond 100-to-1 luminance contrast. Brain mechanisms automatically normalize the visual input based on feature context, but they remain poorly understood because of the limitations of commercially available displays. NEW METHOD: Here, we describe procedures for setup, calibration, and precision check of an HDR display system, based on a JVC DLA-RS600U reference projector, with over 100,000-to-1 luminance dynamic range (636-0.006055 cd/m2), pseudo 11 bit grayscale precision, and 3 ms temporal precision in the MATLAB/Psychtoolbox software environment. The setup is synchronized with electroencephalography (EEG) and infrared eye-tracking measurements. RESULTS: We show display metrics including light scatter versus average display luminance (ADL), spatial uniformity, and spatial uniformity at high spatial frequency. We also show a luminance normalization phenomenon, contextual facilitation of a high contrast target, whose discovery required HDR display. COMPARISON WITH EXISTING METHODS: This system provides 100-fold greater dynamic range than standard 1000-to-1 contrast displays and increases the number of gray levels from 256 or 1024 (8 or 10 bits) to 2048 (pseudo 11 bits), enabling the study of mesopic-to-photopic vision, at the expense of spatial non-uniformities. CONCLUSIONS: This HDR research capability opens new questions of how visual perception is resilient to real-world luminance dynamics and will lead to improved visual modeling of dense urban and forest environments and of mixed indoor-outdoor environments such as cockpits and augmented reality. Our display metrics code can be found at https://github.com/USArmyResearchLab/ARL-Display-Metrics-and-Average-Display-Luminance.

EEGNet: a compact convolutional neural network for EEG-based brain-computer interfaces.

OBJECTIVE: Brain-computer interfaces (BCI) enable direct communication with a computer, using neural activity as the control signal. This neural signal is generally chosen from a variety of well-studied electroencephalogram (EEG) signals. For a given BCI paradigm, feature extractors and classifiers are tailored to the distinct characteristics of its expected EEG control signal, limiting its application to that specific signal. Convolutional neural networks (CNNs), which have been used in computer vision and speech recognition to perform automatic feature extraction and classification, have successfully been applied to EEG-based BCIs; however, they have mainly been applied to single BCI paradigms and thus it remains unclear how these architectures generalize to other paradigms. Here, we ask if we can design a single CNN architecture to accurately classify EEG signals from different BCI paradigms, while simultaneously being as compact as possible. APPROACH: In this work we introduce EEGNet, a compact convolutional neural network for EEG-based BCIs. We introduce the use of depthwise and separable convolutions to construct an EEG-specific model which encapsulates well-known EEG feature extraction concepts for BCI. We compare EEGNet, both for within-subject and cross-subject classification, to current state-of-the-art approaches across four BCI paradigms: P300 visual-evoked potentials, error-related negativity responses (ERN), movement-related cortical potentials (MRCP), and sensory motor rhythms (SMR). MAIN RESULTS: We show that EEGNet generalizes across paradigms better than, and achieves comparably high performance to, the reference algorithms when only limited training data is available across all tested paradigms. In addition, we demonstrate three different approaches to visualize the contents of a trained EEGNet model to enable interpretation of the learned features. SIGNIFICANCE: Our results suggest that EEGNet is robust enough to learn a wide variety of interpretable features over a range of BCI tasks. Our models can be found at: https://github.com/vlawhern/arl-eegmodels.

PFA-100-measured aspirin resistance is the predominant risk factor for hospitalized cardiovascular events in aspirin-treated patients: A 5-year cohort study.

WHAT IS KNOWN AND OBJECTIVE: Aspirin therapy is the clinical gold standard for the prevention of cardiovascular events. However, cardiovascular events still develop in some patients undergoing aspirin therapy. Many laboratory methods exist for measuring aspirin resistance. Using the platelet Function Analyzer (PFA)-100 system, we aimed to determine the effect of aspirin resistance on hospitalized cardiovascular events (hCVE) in a 5-year follow-up cohort. We also sought to determine the impact of aspirin resistance on the relationship between common cardiovascular risk factors and cardiovascular hospitalization. METHOD: Aspirin resistance was evaluated in aspirin-treated patients from the outpatient department. A total of 465 patients during a 5-year follow-up period were included in this study. The primary endpoint of the study was hospitalization for any acute cardiovascular event. The prevalence and associated risk factors of acute cardiovascular events were evaluated. RESULTS AND DISCUSSION: Aspirin resistance was prevalent in 91 (20.0%) of 465 patients. Prior hospitalization history of cardiovascular events was highly associated with aspirin resistance (P = .001). At the 5-year follow-up, cardiovascular events were found to have developed in 11 patients (8 stroke and 3 myocardial infarction) who exhibited aspirin resistance (12.1%) and in 9 (4 stroke and 5 myocardial infarction) patients who did not exhibit aspirin resistance (2.4%) (P < .001). At the 5-year follow-up, multivariate logistic regression analysis results showed a strong association between aspirin resistance and cardiovascular events (adjusted odds ratio 4.28; 95% CI: 1.64-11.20; P = .03). WHAT IS NEW AND CONCLUSION: PFA-100 measurements of aspirin resistance correlate with hCVE, as evidenced by both the past medical history and the 5-year follow-up. The logistic regression analysis results showed that aspirin resistance plays a larger role in hospitalized cardiovascular disease than do other cardiovascular risk factors.

Incorporating blood-based liquid biopsy information into cancer staging: time for a TNMB system?

Tissue biopsy is the standard diagnostic procedure for cancer. Biopsy may also provide material for genotyping, which can assist in the diagnosis and selection of targeted therapies but may fall short in cases of inadequate sampling, particularly from highly heterogeneous tumors. Traditional tissue biopsy suffers greater limitations in its prognostic capability over the course of disease, most obviously as an invasive procedure with potential complications, but also with respect to probable tumor clonal evolution and metastasis over time from initial biopsy evaluation. Recent work highlights circulating tumor DNA (ctDNA) present in the blood as a supplemental, or perhaps an alternative, source of DNA to identify the clinically relevant cancer mutational landscape. Indeed, this noninvasive approach may facilitate repeated monitoring of disease progression and treatment response, serving as a means to guide targeted therapies based on detected actionable mutations in patients with advanced or metastatic solid tumors. Notably, ctDNA is heralding a revolution in the range of genomic profiling and molecular mechanisms to be utilized in the battle against cancer. This review will discuss the biology of ctDNA, current methods of detection and potential applications of this information in tumor diagnosis, treatment, and disease prognosis. Conventional classification of tumors to describe cancer stage follow the TNM notation system, heavily weighting local tumor extent (T), lymph node invasion (N), and detectable metastasis (M). With recent advancements in genomics and bioinformatics, it is conceivable that routine analysis of ctDNA from liquid biopsy (B) may make cancer diagnosis, treatment, and prognosis more accurate for individual patients. We put forward the futuristic concept of TNMB tumor classification, opening a new horizon for precision medicine with the hope of creating better outcomes for cancer patients.

Incorporation of log odds of positive lymph nodes into the AJCC TNM classification improves prediction of survival in oral cancer.

OBJECTIVES: To assess the prognostic performance of a new N classification that incorporates the log odds of positive lymph nodes (LODDS) into the routinely used pathological N classification for oral squamous cell carcinoma (OSCC) patients. DESIGN: Retrospective cohort study utilising LODDS into pN category was performed, and the AJCC TNM stage and T-New N-M stage were compared with respect to 5-year disease-specific survival (DSS) rates. The discriminability was evaluated from the linear trend chi-square test, Akaike information criterion (AIC) and Harrell's c-statistic. SETTING: Medical centrer in Taiwan. PARTICIPANTS: A total of 463 patients received primary surgery and neck dissection between 2004 and 2013 for OSCC. MAIN OUTCOME MEASURES: The discriminability for 5-year DSS rates. RESULTS: The median follow-up period was 54 months, the mean patient age was 54 ± 11 years and 428 patients (92.4%) were male. The patients with higher LODDS had worse 5-year DSS rates. Incorporation of LODDS into the prognostic model based on the seventh edition of the TNM classification significantly improved discriminative performance for 5-year DSS with a lower AIC (1883 versus 1897), and higher prediction accuracy (Harrell's c-statistic: 0.768 versus 0.764). CONCLUSIONS: By utilising a merger of the LODDS and pN classifications to create a new N classification has better discriminatory and predictive ability than pathological TNM staging and could help identify high-risk patients for intense adjuvant therapy.

The impact of smoking on incident type 2 diabetes in a cohort with hepatitis B but not hepatitis C infection.

Smoking may be a risk factor for diabetes, and it has been suggested that viral hepatitis may predispose to diabetes. We studied diabetes and smoking histories in people with viral hepatitis. From 1997 to 2004, we studied the risk of incident diabetes in a community cohort with hyperendemic HBV and HCV infection in southern Taiwan. The cohort involved 3539 people (40-70 years old) without diabetes. Four hundred and twenty-three individuals developed diabetes. Those who were ≥65 years old, frequently consumed alcohol, had a BMI ≥25, had <9 years of education, were anti-HCV+ or smoked ≥1 pack per day were more likely to develop diabetes (P < 0.05). A cumulative hazard function test showed that the higher the smoking levels, the greater the cumulative incidence rate of diabetes in HBsAg+ participants only (P = 0.03 by log-rank test). A multiple Cox proportional hazards model analysis in different hepatitis statuses showed smoking levels were strong predictors of diabetes with a dose-response relationship for type 2 diabetes in those with HBsAg+ : hazard ratio (HR) = 3.8, (95% CI: 1.2, 12.3) for light smokers (<1 pack per day) and HR = 4.4 (95% CI: 1.5, 13.3) for heavy smokers (≥1 pack per day). Increasing BMI was a common predictor in all people. Smoking is a strong predictor for diabetes with a dose-response relationship in HBsAg+ individuals and a mild predictor for seronegative individuals but not significant in anti-HCV+ individuals.

Anterior debridement may not be necessary in the treatment of tuberculous spondylitis of the thoracic and lumbar spine in adults: a retrospective study.

AIM: Many aspects of the surgical treatment of patients with tuberculosis (TB) of the spine, including the use of instrumentation and the types of graft, remain controversial. Our aim was to report the outcome of a single-stage posterior procedure, with or without posterior decompression, in this group of patients. PATIENTS AND METHODS: Between 2001 and 2010, 51 patients with a mean age of 62.5 years (39 to 86) underwent long posterior instrumentation and short posterior or posterolateral fusion for TB of the thoracic and lumbar spines, followed by anti-TB chemotherapy for 12 months. No anterior debridement of the necrotic tissue was undertaken. Posterior decompression with laminectomy was carried out for the 30 patients with a neurological deficit. RESULTS: The mean kyphotic angle improved from 26.1° (- 1.8° to 62°) to 15.2° (-25° to 51°) immediately after the operation. At a mean follow-up of 68.8 months (30 to 144) the mean kyphotic angle was 16.9° (-22° to 54°), with a mean loss of correction of 1.6° (0° to 10°). There was a mean improvement in neurological status of 1.2 Frankel grades in those with a neurological deficit. Bony union was achieved in all patients, without recurrent infection. CONCLUSIONS: Long posterior instrumentation with short posterior or posterolateral fusion is effective in the treatment of TB spine. It controls infection, corrects the kyphosis, and maintains correction and neurological improvement over time. TAKE HOME MESSAGE: With effective anti-TB chemotherapy, a posterior only procedure without debridement of anterior lesion is effective in the treatment of TB spondylitis, and an anterior procedure can be reserved for those patients who have not improved after posterior surgery. Cite this article: Bone Joint J 2016;98-B:834-9.

Is removal of the implants needed after fixation of burst fractures of the thoracolumbar and lumbar spine without fusion? a retrospective evaluation of radiological and functional outcomes.

METHODS: In this study of patients who underwent internal fixation without fusion for a burst thoracolumbar or lumbar fracture, we compared the serial changes in the injured disc height (DH), and the fractured vertebral body height (VBH) and kyphotic angle between patients in whom the implants were removed and those in whom they were not. Radiological parameters such as injured DH, fractured VBH and kyphotic angle were measured. Functional outcomes were evaluated using the Greenough low back outcome scale and a VAS scale for pain. RESULTS: Between June 1996 and May 2012, 69 patients were analysed retrospectively; 47 were included in the implant removal group and 22 in the implant retention group. After a mean follow-up of 66 months (48 to 107), eight patients (36.3%) in the implant retention group had screw breakage. There was no screw breakage in the implant removal group. All radiological and functional outcomes were similar between these two groups. Although solid union of the fractured vertebrae was achieved, the kyphotic angle and the anterior third of the injured DH changed significantly with time (p < 0.05). DISCUSSION: The radiological and functional outcomes of both implant removal and retention were similar. Although screw breakage may occur, the implants may not need to be removed. TAKE HOME MESSAGE: Implant removal may not be needed for patients with burst fractures of the thoracolumbar and lumbar spine after fixation without fusion. However, information should be provided beforehand regarding the possibility of screw breakage.

Seven-year incidence of uncorrected refractive error among an elderly Chinese population in Shihpai, Taiwan: The Shihpai Eye Study.

PURPOSE: To report the 7-year incidence of uncorrected refractive error in a metropolitan Chinese elderly population. METHODS: The Shihpai Eye Study 2006 included 460/824 (55.8%) subjects (age range 72-94 years old) of 1361 participants in the 1999 baseline survey for a follow-up eye examination. Visual acuity was assessed using a Snellen chart, uncorrected refractive error was defined as presenting visual acuity (naked eye if without spectacles and with distance spectacles if worn) in the better eye of <6/12 that improved to no impairment (≥6/12) after refractive correction. RESULTS: The 7-year incidence of uncorrected refractive error was 10.5% (95% confidence interval (CI): 7.6-13.4%). 92.7% of participants with uncorrection and 77.8% with undercorrection were able to improve at least two lines of visual acuity by refractive correction. In multivariate analysis controlling for covariates, uncorrected refractive error was significantly related to myopia (relative risk (RR): 3.15; 95% CI: 1.31-7.58) and living alone (RR: 2.94; 95% CI 1.14-7.53), whereas distance spectacles worn during examination was protective (RR: 0.35; 95% CI: 0.14-0.88). CONCLUSION: Our study indicated that the incidence of uncorrected refractive error was high (10.5%) in this elderly Chinese population. Living alone and myopia are predisposing factors, whereas wearing distance spectacles at examination is protective.

Omental mesenteric myxoid hamartoma, a subtype of inflammatory myofibroblastic tumor? Considerations based on the histopathological evaluation of four cases.

Omental mesenteric myxoid hamartoma (OMH) is a distinctive myxoid lesion of infancy, characterized by a benign clinical behavior. In the current World Health Organization (WHO) classification of soft tissue tumors, it is considered as part of the morphologic spectrum of inflammatory myofibroblastic tumors (IMT), but this relationship with IMT is still subject to debate. Four lesions with histologic features of OMH occurring in newborns and toddlers are described and compared with classic, ALK-positive IMT. All OMH showed a peculiar dot-like immunostaining for ALK, which, in one of the cases, was cytogenetically found to be associated with an inversion of the ALK gene. While OMHs were positive for smooth muscle actin (SMA), desmin, WT1, podoplanin, and cytokeratins (CAM5.2 and AE1-3), IMT were consistently positive only for SMA (10 cases). ALK-1 displayed cytoplasmic staining in IMT and characteristic paranuclear dot-like staining in OMH.

Network Anisotropy Trumps Noise for Efficient Object Coding in Macaque Inferior Temporal Cortex.

How neuronal ensembles compute information is actively studied in early visual cortex. Much less is known about how local ensembles function in inferior temporal (IT) cortex, the last stage of the ventral visual pathway that supports visual recognition. Previous reports suggested that nearby neurons carry information mostly independently, supporting efficient processing (Barlow, 1961). However, others postulate that noise covariation effects may depend on network anisotropy/homogeneity and on how the covariation relates to representation. Do slow trial-by-trial noise covariations increase or decrease IT's object coding capability, how does encoding capability relate to correlational structure (i.e., the spatial pattern of signal and noise redundancy/homogeneity across neurons), and does knowledge of correlational structure matter for decoding? We recorded simultaneously from ∼80 spiking neurons in ∼1 mm(3) of macaque IT under light neurolept anesthesia. Noise correlations were stronger for neurons with correlated tuning, and noise covariations reduced object encoding capability, including generalization across object pose and illumination. Knowledge of noise covariations did not lead to better decoding performance. However, knowledge of anisotropy/homogeneity improved encoding and decoding efficiency by reducing the number of neurons needed to reach a given performance level. Such correlated neurons were found mostly in supragranular and infragranular layers, supporting theories that link recurrent circuitry to manifold representation. These results suggest that redundancy benefits manifold learning of complex high-dimensional information and that subsets of neurons may be more immune to noise covariation than others. SIGNIFICANCE STATEMENT: How noise affects neuronal population coding is poorly understood. By sampling densely from local populations supporting visual object recognition, we show that recurrent circuitry supports useful representations and that subsets of neurons may be more immune to noise covariation than others.

Corrigendum: Correlated activity supports efficient cortical processing.

[This corrects the article on p. 171 in vol. 8, PMID: 25610392.].

Determinants of cerebral white matter changes in patients with stroke.

BACKGROUND/AIM: Cerebral white matter changes (WMC) are commonly observed in magnetic resonance imaging (MRI) scans of elderly people. Information about the prevalence of WMC is limited, and little is known about site-specific risk factors for the subcortical and periventricular regions in patients with ischaemic stroke. The study aims to analyse the prevalence and severity of WMC and investigate the risk factors of periventricular WMC (PVWMC) and deep WMC (DWMC) separately in patients with ischaemic stroke. METHODS: The data were collected between January and December 2013 from a medical centre in southern Taiwan. Every patient underwent a cerebral MRI scan, and WMC was separately rated as PVWMC and DWMC by using the modified Fazekas scale. RESULTS: In total, 527 patients who had experienced ischaemic stroke were included. The mean age of the patients was 67.0 ± 12.5 years (range: 31-94) and 62% of them were men. The mean age was significantly different among the four grades of severity in both the PVWMC (P < 0.001) and DWMC (P < 0.001) groups after adjustments for sex and vascular risk factors. Hypertension was independently correlated with severity of DWMC (P = 0.032) but not with PVWMC (P = 0.222). In multiple logistic regressions model, hypertension was a significant independent indicator of DWMC (odds ratio = 4.30; 95% confidence interval = 1.70-10.89). CONCLUSION: Our results suggest a region-specific pathogenesis of cerebral white matter in Asian patients with ischaemic stroke that may differ from those in the general population.

Factors correlated with tuberculosis reported after death.

SETTING: Taiwan. OBJECTIVE: To determine factors associated with tuberculosis (TB) reported after death. DESIGN: A retrospective cohort study was conducted of TB patients newly registered between 2006 and 2008. The national TB database was linked with the Vital Registry System and National Health Insurance database for analysis. RESULTS: A total of 1409 (4.0%) TB cases were reported after death. Age ⩾75 years (OR 1.70), chest X-ray (CXR) unknown or not performed (OR 2.41), positive sputum bacteriology (OR 1.74), and comorbidities such as cancer (OR 1.21), chronic liver disease (OR 1.21) and chronic kidney disease (OR 2.58) were associated with the reporting of TB cases after death. More than 30% of TB deaths in elderly persons with chronic kidney disease were TB cases that were reported after death. Abnormal CXR and comorbidities such as chronic obstructive pulmonary disease, stroke and diabetes mellitus were less likely to be associated with the reporting of TB cases after death. CONCLUSIONS: Elderly patients with cancer, chronic liver disease or chronic kidney disease require more attention to prevent delayed diagnosis and early mortality from TB. Elderly patients with chronic kidney disease are at highest risk.

Two-dimensional aperiodic nonlinear photonic crystal in a dual-wavelength Nd:YVO4 laser for pulsed orange generation.

We report the design and construction of a highly integrated two-dimensional (2D) aperiodic nonlinear photonic crystal (NPC) for working in a diode-pumped, dual-wavelength (1064 and 1342 nm) Nd:YVO4 laser to demonstrate a compact, high-peak-power intracavity sum-frequency generator (ISFG) radiating at orange 593.5 nm. The 2D aperiodic NPC was built in quasi-phase-matched LiNbO3whose crystal domain was structured based on the aperiodic optical superlattice technique to best achieve its simultaneous performance of a dual-wavelength electro-optic Bragg Q-switch and a SFG in the Nd:YVO4 laser. When the NPC device was driven with a 350-V Q-switching voltage and a 1-kHz switching rate, we measured pulse energy of ~4.3 µJ (or peak power of ~531 W) at orange 593.5 nm from the constructed ISFG with 5.28-W diode power.

Electroluminescence enhancement in InGaN light-emitting diode during the electrical stressing process.

This study of the optoelectronic properties of blue light-emitting diodes under direct current stress. It is found that the electroluminescence intensity increases with duration of stress, and the efficiency droop curves illustrated that the peak-efficiency and the peak-efficiency-current increases and decreases, respectively. We hypothesize that these behaviors mainly result from the increased internal quantum efficiency.