Estimation of modified Zernike coefficients from turbulence-degraded multispectral imagery using deep learning.

We investigate how wavelength diversity affects the performance of a deep-learning model that predicts the modified Zernike coefficients of turbulence-induced wavefront error from multispectral images. The ability to perform accurate predictions of the coefficients from images collected in turbulent conditions has potential applications in image restoration. The source images for this work were a point object and extended objects taken from a character-based dataset, and a wavelength-dependent simulation was developed that applies the effects of isoplanatic atmospheric turbulence to the images. The simulation utilizes a phase screen resampling technique to emulate the simultaneous collection of each band of a multispectral image through the same turbulence realization. Simulated image data were generated for the point and extended objects at various turbulence levels, and a deep neural network architecture based on AlexNet was used to predict the modified Zernike coefficients. Mean squared error results demonstrate a significant improvement in predicting modified Zernike coefficients for both the point object and extended objects as the number of spectral bands is increased. However, the improvement with the number of bands was limited when using extended objects with additive noise.

Estimation of non-uniform motion blur using a patch-based regression convolutional neural network.

The non-uniform blur of atmospheric turbulence can be modeled as a superposition of linear motion blur kernels at a patch level. We propose a regression convolutional neural network (CNN) to predict angle and length of a linear motion blur kernel for varying sized patches. We analyze the robustness of the network for different patch sizes and the performance of the network in regions where the characteristics of the blur are transitioning. Alternating patch sizes per epoch in training, we find coefficient of determination scores across a range of patch sizes of R 2>0.78 for length and R 2>0.94 for angle prediction. We find that blur predictions in regions overlapping two blur characteristics transition between the two characteristics as overlap changes. These results validate the use of such a network for prediction of non-uniform blur characteristics at a patch level.

Deep learning estimation of modified Zernike coefficients and recovery of point spread functions in turbulence.

Recovering the turbulence-degraded point spread function from a single intensity image is important for a variety of imaging applications. Here, a deep learning model based on a convolutional neural network is applied to intensity images to predict a modified set of Zernike polynomial coefficients corresponding to wavefront aberrations in the pupil due to turbulence. The modified set assigns an absolute value to coefficients of even radial orders due to a sign ambiguity associated with this problem and is shown to be sufficient for specifying the intensity point spread function. Simulated image data of a point object and simple extended objects over a range of turbulence and detection noise levels are created for the learning model. The MSE results for the learning model show that the best prediction is found when observing a point object, but it is possible to recover a useful set of modified Zernike coefficients from an extended object image that is subject to detection noise and turbulence.

A Comparison of Topical Agents for Eschar Removal in a Porcine Model: Bromelain-enriched vs Traditional Collagenase Agents.

Surgical excision and grafting of deep partial-thickness (DPT) and full-thickness (FT) burns is a cornerstone of wound care. The use of commercially available topical enzymatic agents has been limited due to slower and less complete eschar removal than surgical excision. Using a porcine model of DPT and FT burns, we compared the eschar removal efficacy of a bromelain-enriched enzymatic agent derived from the stems of pineapple plants and a commercially available collagenase. We created 40 DPT and 40 FT burns on four anesthetized Yorkshire pigs. Eschar removal was initiated 24 hours later. Two pigs each were randomly assigned to collagenase or the bromelain-enriched agent. The bromelain-enriched agent was applied topically once for 4 hours followed by a 2-hour soaking. The collagenase was applied topically daily until complete removal of eschar or for up to 14 days. All bromelain-enriched treated FT burns underwent complete removal of the eschar after a single application while none of the collagenase-treated FT burns underwent complete removal of the eschar even after 14 days of treatment. All bromelain-enriched treated DPT burns had complete eschar removal after the single application. None of the collagenase-treated DPT burns experienced complete removal of eschar after 10 days; by day 14, 35% had complete eschar removal, 30% had >50% eschar removed, and 35% had <50% eschar removed. We conclude that eschar removal is quicker and more complete with the bromelain-enriched compared with collagenase debriding agent.

Clinical outcome of multidisciplinary treatment of vulvar necrotising fasciitis.

OBJECTIVE: Vulvar necrotising fasciitis (VNF) is a severe soft tissue infection associated with substantial morbidity and high mortality. At Stony Brook Medicine, US, patients with known or suspected VNF are treated by a structured multidisciplinary team consisting of members of the Departments of Emergency Medicine and Medicine, the Divisions of Gynecologic Oncology, Burn and Surgical Intensive Care Units, Infectious Disease and Plastic Surgery, and the nursing, nutrition, physical/occupational therapy and social work services. METHOD: This is a retrospective review of patients presenting to Stony Brook University Hospital with VNF over an 18-month period. RESULTS: A total of 10 patients were treated for VNF during the study period. All patients were treated by the structured multidisciplinary team, including extensive initial surgical debridement by the gynaecologic oncologists. All patients survived to discharge. CONCLUSION: The results of this review demonstrated that prompt diagnosis, rapid implementation of appropriate antibiotic coverage, surgical debridement of necrotic tissue, and comprehensive care delivered by a structured multidisciplinary team contributed to positive clinical outcomes and decreased the risk of death from VNF.

Recasting the (Synchrosqueezed) Short-Time Fourier Transform as an Instantaneous Spectrum.

In a previous work, we proposed a time-frequency analysis called instantaneous spectral analysis (ISA), which generalizes the notion of the Fourier spectrum and in which instantaneous frequency is utilized to the fullest extent. In this paper, we recast both the Fourier transform (FT) and filterbank (FB) interpretations of the short-time Fourier transform (STFT) as instantaneous spectra. We show that to recast the FB interpretation of STFT as an instantaneous spectrum with valid structure, frequency reassignment is a fundamental necessity, thus demonstrating that this IS is closely related to the synchrosqueezed STFT. This result provides a new theoretical motivation for the synchrosqueezed STFT. Finally, we illustrate through example the instantaneous spectra corresponding to the FT and FB interpretations of STFT using two closed-form examples.

Prosodic Entrainment in Conversations of Verbal Children and Teens on the Autism Spectrum.

Unusual speech prosody has long been recognized as a characteristic feature of the speech of individuals diagnosed with Autism Spectrum Disorders (ASD). However, research to determine the exact nature of this difference in speech prosody is still ongoing. Many individuals with verbal autism perform well on tasks testing speech prosody. Nonetheless, their expressive prosody is judged to be unusual by others. We propose that one aspect of this perceived difference in speech prosody in individuals with ASD may be due to a deficit in the ability to entrain-or become more similar-to their conversation partners in prosodic features over the course of a conversation. In order to investigate this hypothesis, 24 children and teens between the ages of 9 and 15 years participated in our study. Twelve of the participants had previously been diagnosed with ASD and the other 12 participants were matched to the ASD participants in age, gender, and non-verbal IQ scores. All participants completed a goal-directed conversation task, which was subsequently analyzed acoustically. Our results suggest (1) that youth diagnosed with ASD entrain less to their conversation partners compared to their neurotypical peers-in fact, children and teens diagnosed with ASD tend to dis-entrain from their conversation partners while their neurotypical peers tend to converge to their conversation partners' prosodic features. (2) Although age interacts differently with prosodic entrainment in youth with and without ASD, this difference is attributable to the entrainment behavior of the conversation partners rather than to those with ASD. (3) Better language skill is negatively correlated with prosodic entrainment for both youth with and without ASD. The observed differences in prosodic entrainment in children and teens with ASD may not only contribute to the perceived unusual prosody in youth with ASD but are also likely to be indicative of their difficulties in social communication, which constitutes a core challenge for individuals with ASD.

The Relationship between Prosodic Ability and Conversational Prosodic Entrainment.

Conversational entrainment or alignment-the convergence of conversation partners over the course of a conversation in a variety of linguistic features-is a well-attested conversational phenomenon. The research on prosodic entrainment has shown correlations between prosodic entrainment and several social dimensions of rapport between conversation partners. However, little is known about how skill-level in the entrainment domain affects the ability to converge during a conversation. The goal of the current study was to investigate whether skill-level of a speaker in receptive and expressive word, sentence, and emotional prosody is correlated with the amount of prosodic entrainment contributed at the conversational level. Twenty native speakers of American English were paired into ten dyads of seven female/female and three female/male conversation pairs. Conversations for each pair were recorded and analyzed. Test scores measuring word, sentence, and emotional prosody were correlated with the amount of fundamental frequency entrainment during conversations. The results indicate that a negative correlation exists between expressive prosody skill and the amount of f0 entrainment contributed by a speaker. This suggests that speakers with better expressive prosodic skills at the word and sentence level entrain less to their conversation partners. Receptive prosody ability was not correlated with conversational prosodic entrainment.

The association of patient and burn characteristics with itching and pain severity.

INTRODUCTION: Itch after burn injury causes significant distress to patients and can hamper functional recovery. Itching can persist on a time scale ranging from several weeks to even years after injury. In this study, we sought to determine predictors of itching after burn injury. METHODS: We compared itch and pain severity among patients included in a burn registry at a level 1 trauma center. Both itch and pain severity was based on a 5-point scale. ANOVA, chi-squared, and multivariate analyses were performed to determine predictors of itch and pain severity. RESULTS: Of the 1159 patients enrolled in the registry, 58% were male and 42% female, with a median age (IQR) of 27 (8-47) years. Most patients were diagnosed with 2nd degree superficial (41%) or deep (43%) burns. Upper extremities were the most common location of burn injury (59%), followed by lower extremities (31%), trunk (22%), and face/neck (20%). More than half (53%) of enrolled patients reported itching, ranging from minimal (19%) to severe (7%) itching. Multivariate analyses revealed age, sex (female), extent of burn injuries, and location (face/neck) to be predictors of itch after burn. Predictors of pain after burn were slightly different: age, extent of burn, and depth of burn. CONCLUSION: Pain and itch after burn injuries are predicted by slightly different variables, presumably secondary to different underlying mechanisms. We conclude that age, sex (female), extent of burn injuries (total body surface area %), and injuries to the face/neck predict itching of greater severity. Patients with burn injuries that match these parameters would require greater care and closer follow up to reduce itching after healing.

The Effects of Speech Compression Algorithms on the Intelligibility of Two Individuals With Dysarthric Speech.

Purpose Telemedicine, used to offset disparities in access to speech-language therapy, relies on technology that utilizes compression algorithms to transmit signals efficiently. These algorithms have been thoroughly evaluated on healthy speech; however, the effects of compression algorithms on the intelligibility of disordered speech have not been adequately explored. Method This case study assessed acoustic and perceptual effects of resampling and speech compression (i.e., transcoding) on the speech of 2 individuals with dysarthria. Forced-choice vowel identification and transcription tasks were utilized, completed by 20 naive undergraduate listeners. Results Results showed relative improvements and decrements in intelligibility, on various measures, based on the speakers' acoustic profiles. The transcoding of the speech compression algorithm resulted in an enlarged vowel space area and associated improvements in vowel identification for 1 speaker and a smaller vowel space area and decreased vowel identification for the other speaker. Interestingly, there was an overall decrease in intelligibility in the transcription task in this condition for both speakers. Conclusions There is a complex interplay between dysarthria and compression algorithms that warrants further exploration. The findings suggest that it is critical to be mindful of apparent changes in intelligibility secondary to compression algorithms necessary for practicing telemedicine. Supplemental Material https://doi.org/10.23641/asha.7291940.

Which Burn Outcomes Do Patients Anticipate as Most Likely to Be Important.

Burn therapies should focus on achieving outcomes that are most important to patients. The authors wanted to discover which outcomes newly burned patients would anticipate as most important to them and explored the association between demographic/burn characteristics and patient preferences. The authors surveyed 753 of 776 patients seen by our burn service from 2008 to 2013 during the initial encounter. Patients were asked to rate the anticipated importance of several burn outcomes including cosmetic appearance, resumption of normal function, and the lack of pain/itching on a four-item Likert scale (not important, somewhat important, important, and extremely important). The association between demographic and burn characteristics with patient's views on the importance of various outcomes was explored with χ and nonparametric tests. Patient mean (SD) age was 30 (22) years, 58% were males, 69% were white. Overall, function was extremely important to 96% of patients, lack of pain/itching was extremely important to 85% of patients, and cosmesis was extremely important to 59% of patients. Cosmesis was extremely important to more females than males (69 vs 52%; P < .001) and the mean age of patients in whom cosmesis was extremely important was lower than those in whom it was not (25 vs 40; P < .001). Cosmesis was more commonly extremely important in patients with head/neck than extremity burns (67 vs 57%; P < .001). Levels of importance for function and lack of pain/itching did not differ by gender, age, TBSA, or burn location. Thus, return to normal function and lack of pain and itching appear to be more commonly very important to burn patient than the cosmetic appearance of their burns. Cosmesis was of greater importance to younger patients, female patients, and those with head/neck burns.

Infrared Thermal Imaging Has the Potential to Reduce Unnecessary Surgery and Delays to Necessary Surgery in Burn Patients.

Clinical distinction between superficial and deep burns is problematic. The authors determined whether an infrared thermal imaging (IRTI) camera could predict burn depth. Burn depth was assessed by an experienced surgeon, and the burns were imaged with a portable, lightweight IRTI camera that measures heat emission from the skin using long infrared wavelengths (7.5-13 µm). Burns were considered "deep" if they were surgically excised and confirmed to be of full thickness on microscopic evaluation or if they did not heal spontaneously within 21 days of injury. All other burns were considered "nondeep." There were 39 burns that had both days 1 and 2 IRTI measurements and available outcome. Of these, 16 were "deep" burns and 23 were "nondeep." The mean temperatures of "nondeep" burns between days 1 and 2 increased from 30.6 ± 2.7 to 32.1 ± 3.0°C (difference of 1.5 ± 2.3°C). The mean temperatures of "deep" burns decreased from 32.3 ± 2.0 to 30.8 ± 1.3°C (difference of -1.5 ± 2.0°C) between days 1 and 2. Any decrease in temperatures between days 1 and 2 was predictive of a deep wound, and any increase between days 1 and 2 was predictive of a nondeep burn. Using the ultimate burn depth as the criterion standard, the overall accuracy of IRTI was considerably higher than that of clinical assessment; 87.2% (95% CI: 71.8-95.2) vs 54.1% (95% CI: 37.1-70.2). Any decrease in temperatures between days 1 and 2 was predictive of a deep wound. Our results suggest that thermography using IRTI is more accurate than clinical examination in predicting burn depth and could potentially reduce unnecessary surgery as well as reduce delays to surgery when necessary.

The Role of Necroptosis in Burn Injury Progression in a Rat Comb Burn Model.

OBJECTIVES: Progression of cell death after burn injury may occur by one of three mechanisms: passive necrosis, apoptosis, and programmed necroptosis that requires the receptor-interacting protein kinase-3 (RIP-3). The hypothesis was that RIP-3 is present in normal and burned skin; that necroptosis plays a role in burn injury progression; and that treatment with necrostatin-1, an inhibitor of necroptosis, would reduce burn progression. METHODS: Skin specimens from rats were examined for the presence of RIP-3. Using a 150-g brass comb preheated to 100°C, we created two comb burns (one on each side) consisting of four rectangular burns, separated by three unburned interspaces, on both sides of the backs of anesthetized male Sprague-Dawley rats (240 to 300 g). The interspaces represent the ischemic zones surrounding the central necrotic core. Left untreated, these areas undergo necrosis. In the first experiment, 10 rats each were randomized to 1.65 mg/kg necrostatin-1 or control given by intraperitoneal injection 1 hour after injury. In the second experiment, 10 rats each were randomized to two intravenous injections of 1.65 mg/kg necrostatin-1 or its vehicle at 1 and 4 hours after injury. The primary outcome was the percentage of interspaces undergoing necrosis within 7 days of injury. Binary data were compared with chi-square or Fishers' exact tests. RESULTS: All normal and burned skin specimens from rats stained positive for RIP-3. In the first experiment, nearly all unburned interspaces in both the experimental and the control rats underwent necrosis (47 of 48, 97.9% vs. 48 of 48, 100%; p = not significant [NS]). Similarly, in the second experiment, there was no difference in the percentage of unburned interspaces undergoing necrosis within 7 days of injury in rats treated with two doses of necrostatin-1 or the control vehicle (46 of 48, 95.8% vs. 48 of 48, 100%; p = NS). There were no wound infections noted in rats injected with necrostatin-1. CONCLUSIONS: The skin of rats contains RIP-3 necessary for necroptosis. Injection of rats with either a single intraperitoneal dose or two intravenous doses of necrostatin-1 failed to reduce burn injury progression in a rat comb burn model. This may be due to inactivity of necrostatin-1 or the lack of a role of necroptosis in burn injury progression in the rat comb burn model.

Association between burn characteristics and pain severity.

INTRODUCTION: Traditionally, full thickness burns have been thought to be painless due to destruction of underlying nerves. We explored the association between patient and burn characteristics and pain severity in burn patients and determined whether full thickness burns were less likely to be painful than more superficial burns. METHODS: We performed a structured review of medical records of patients presenting to a burn center between 2010 and 2013. Data abstracted included baseline patient and burn characteristics. The primary end point was pain severity on patient arrival to the emergency department using a verbal numeric score of 0 to 10. Univariate and multivariate analyses were used to explore the association between patient and burn characteristics and pain severity. RESULTS: There were 507 patients. Mean (SD) age was 29.9 (23.6); 38% were ages younger than 18, and 68% were males. The median (interquartile ranges) pain score was 5 (2-8). Of all patients, 7% had isolated full thickness burns. Median (interquartile ranges) pain scores in isolated full thickness burns were slightly lower than in more superficial burns: 4 (1-8) vs 6 (2-8), respectively, P = .09. Twenty-five percent of patients with isolated full thickness burns had pain scores of 0 compared with 18% of all others (P = .28). There was no correlation between total body surface area and pain severity, however, pain scores increased with the number of burns (P = .007). CONCLUSIONS: Pain severity is slightly less with full thickness burns; however, most patients have pain. The presence of pain should not be used to exclude full thickness burns.

Modeling Pathological Speech Perception From Data With Similarity Labels.

The current state of the art in judging pathological speech intelligibility is subjective assessment performed by trained speech pathologists (SLP). These tests, however, are inconsistent, costly and, oftentimes suffer from poor intra- and inter-judge reliability. As such, consistent, reliable, and perceptually-relevant objective evaluations of pathological speech are critical. Here, we propose a data-driven approach to this problem. We propose new cost functions for examining data from a series of experiments, whereby we ask certified SLPs to rate pathological speech along the perceptual dimensions that contribute to decreased intelligibility. We consider qualitative feedback from SLPs in the form of comparisons similar to statements "Is Speaker A's rhythm more similar to Speaker B or Speaker C?" Data of this form is common in behavioral research, but is different from the traditional data structures expected in supervised (data matrix + class labels) or unsupervised (data matrix) machine learning. The proposed method identifies relevant acoustic features that correlate with the ordinal data collected during the experiment. Using these features, we show that we are able to develop objective measures of the speech signal degradation that correlate well with SLP responses.

Characterizing the distribution of the quadrilateral vowel space area.

The vowel space area (VSA) has been studied as a quantitative index of intelligibility to the extent it captures articulatory working space and reductions therein. The majority of such studies have been empirical wherein measures of VSA are correlated with perceptual measures of intelligibility. However, the literature contains minimal mathematical analysis of the properties of this metric. This paper further develops the theoretical underpinnings of this metric by presenting a detailed analysis of the statistical properties of the VSA and characterizing its distribution through the moment generating function. The theoretical analysis is confirmed by a series of experiments where empirically estimated and theoretically predicted statistics of this function are compared. The results show that on the Hillenbrand and TIMIT data, the theoretically predicted values of the higher-order statistics of the VSA match very well with the empirical estimates of the same.

Automatic assessment of vowel space area.

Vowel space area (VSA) is an attractive metric for the study of speech production deficits and reductions in intelligibility, in addition to the traditional study of vowel distinctiveness. Traditional VSA estimates are not currently sufficiently sensitive to map to production deficits. The present report describes an automated algorithm using healthy, connected speech rather than single syllables and estimates the entire vowel working space rather than corner vowels. Analyses reveal a strong correlation between the traditional VSA and automated estimates. When the two methods diverge, the automated method seems to provide a more accurate area since it accounts for all vowels.

Selecting Disorder-Specific Features for Speech Pathology Fingerprinting.

The general aim of this work is to learn a unique statistical signature for the state of a particular speech pathology. We pose this as a speaker identification problem for dysarthric individuals. To that end, we propose a novel algorithm for feature selection that aims to minimize the effects of speaker-specific features (e.g., fundamental frequency) and maximize the effects of pathology-specific features (e.g., vocal tract distortions and speech rhythm). We derive a cost function for optimizing feature selection that simultaneously trades off between these two competing criteria. Furthermore, we develop an efficient algorithm that optimizes this cost function and test the algorithm on a set of 34 dysarthric and 13 healthy speakers. Results show that the proposed method yields a set of features related to the speech disorder and not an individual's speaking style. When compared to other feature-selection algorithms, the proposed approach results in an improvement in a disorder fingerprinting task by selecting features that are specific to the disorder.

Impact of intelligent intravenous infusion pumps on directing care toward evidence-based standards: a retrospective data analysis.

INTRODUCTION: Published literature has successfully demonstrated the impact of intravenous (IV) infusion pump safety software on improving the quality of health care delivery. Much of this literature has focused solely on the ability of these devices to prevent potential medication errors, while overlooking the devices' additional valuable advantages. One non-reported benefit is the ability of IV infusion pump safety software to consistently administer doses of IV medication, which are based on evidence. This article describes the process undertaken to implement and evaluate the impact of IV infusion pump safety software on driving care toward evidence-based standards. METHODS: An advisory group of expert users was convened for a 2-day session to develop consensus recommendations of best practices for IV infusion pump safety software. Using these recommendations, administrative data were collected from a community hospital to assess the endpoints identified by the advisory panel. RESULTS: Data analysis of rescue agents (ie, flumazenil, glucagon, and protamine sulfate) showed reductions in utilization in the post-implementation period of the safety software. The decreased requirement for blood transfusions in patients receiving heparin infusions suggests that heparin infusions were more safely administered in the post-implementation period. The decreased length of stay and mortality rate observed in patients with complex respiratory infections during the post-implementation period suggests that by correctly infusing antibiotics consistently, patient outcomes may be improved. Additionally, alert and edit data from the pumps demonstrated that the IV infusion pump safety software alerted to and influenced edits on many critical dose rate errors for benzodiazepines, heparin, and several antibiotics. CONCLUSION: Intravenous infusion pump safety software improves clinical outcomes through consistent application of evidence-based standards of dose rates for IV drugs.

Predictors of sustaining burn injury: does the use of common prevention strategies matter?

The incidence of burn injury has decreased over the past several decades. Although this has been largely attributed to increased prevention awareness, few studies evaluate the effectiveness of implementing standard burn prevention strategies in preventing burn injury. The authors hypothesized that patients who sustain burns use burn prevention strategies less frequently than those who do not. This was a case-control study composed of a prospective survey questionnaire and retrospective burn registry query, which was performed in a suburban academic medical center with a burn unit. All burn patients seen by the burn service in the year 2008 and a nonrandom sample of nonburned emergency department patients and visitors during the same time period were enrolled. Demographics included age, gender, income, education, house type, insurance status, and prevention strategy usage including smoke alarms, carbon monoxide detectors, fire extinguishers, and escape plans. The primary outcome of interest in this study was burn injury. Chi-square tests were used to compare rates, Student's t-tests were used to compare mean values of continuous variables between burn patients and others, and multivariate logistic regression was used to determine the strongest predictors of sustaining burn injury. One hundred ninety-four burn patients and 348 nonburned emergency department patients and visitors were surveyed. Burn patients reported the same rates of smoke alarm usage (96.9 vs 96.3%, P = .692), carbon monoxide detectors (75.3 vs 67.2%, P = .05), and higher rates of fire extinguisher ownership (80.4 vs 72.7%, P = .045) when compared with others. In multivariable analysis, the strongest predictor of sustaining burn injury was less than high school education (odds ratio [OR] 3.61, 95% confidence interval [CI] 1.27-10.27), whereas English as a primary language (OR 0.48, 95% CI 0.26-0.89), a graduate degree (OR 0.10, 95% CI 0.02-0.42), income >$50,000 (OR 0.46, 95% CI 0.29-0.72), and keeping flammable liquids in a locked place (OR 0.59, 95% CI 0.44-0.80) were protective against burn injury. Smoke alarms (OR 0.75, 95% CI 0.22-2.61), fire extinguishers (OR 1.34, 95% CI 0.80-2.32), and having an escape plan (OR 0.85, 95% CI 0.56-1.29) were not protective. Patients who sustain burn injury use burn prevention strategies at similar rates when compared with those who do not. When holding demographic characteristics constant, utilization of most burn prevention strategies is not protective of sustaining burn injury. Those with lower levels of education and income remain more susceptible to burn injury.