One crucial test for any quantitative model of the brain is to show that the model can be used to accurately decode information from evoked brain activity. Several recent neuroimaging studies have decoded the structure or semantic content of static visual images from human brain activity. Here we present a decoding algorithm that makes it possible to decode detailed information about the object and action categories present in natural movies from human brain activity signals measured by functional MRI. Decoding is accomplished using a hierarchical logistic regression (HLR) model that is based on labels that were manually assigned from the WordNet semantic taxonomy. This model makes it possible to simultaneously decode information about both specific and general categories, while respecting the relationships between them. Our results show that we can decode the presence of many object and action categories from averaged blood-oxygen level-dependent (BOLD) responses with a high degree of accuracy (area under the ROC curve > 0.9). Furthermore, we used this framework to test whether semantic relationships defined in the WordNet taxonomy are represented the same way in the human brain. This analysis showed that hierarchical relationships between general categories and atypical examples, such as organism and plant, did not seem to be reflected in representations measured by BOLD fMRI.

Given the extraordinary ability of humans and animals to recognize communication signals over a background of noise, describing noise invariant neural responses is critical not only to pinpoint the brain regions that are mediating our robust perceptions but also to understand the neural computations that are performing these tasks and the underlying circuitry. Although invariant neural responses, such as rotation-invariant face cells, are well described in the visual system, high-level auditory neurons that can represent the same behaviorally relevant signal in a range of listening conditions have yet to be discovered. Here we found neurons in a secondary area of the avian auditory cortex that exhibit noise-invariant responses in the sense that they responded with similar spike patterns to song stimuli presented in silence and over a background of naturalistic noise. By characterizing the neurons' tuning in terms of their responses to modulations in the temporal and spectral envelope of the sound, we then show that noise invariance is partly achieved by selectively responding to long sounds with sharp spectral structure. Finally, to demonstrate that such computations could explain noise invariance, we designed a biologically inspired noise-filtering algorithm that can be used to separate song or speech from noise. This novel noise-filtering method performs as well as other state-of-the-art de-noising algorithms and could be used in clinical or consumer oriented applications. Our biologically inspired model also shows how high-level noise-invariant responses could be created from neural responses typically found in primary auditory cortex.

With the success of deep learning in a wide variety of areas, many deep multi-task learning (MTL) models have been proposed claiming improvements in performance obtained by sharing the learned structure across several related tasks. However, the dynamics of multi-task learning in deep neural networks is still not well understood at either the theoretical or experimental level. In particular, the usefulness of different task pairs is not known a priori. Practically, this means that properly combining the losses of different tasks becomes a critical issue in multi-task learning, as different methods may yield different results. In this paper, we benchmarked different multi-task learning approaches using shared trunk with task specific branches architecture across three different MTL datasets. For the first dataset, i.e. Multi-MNIST (Modified National Institute of Standards and Technology database), we thoroughly tested several weighting strategies, including simply adding task-specific cost functions together, dynamic weight average (DWA) and uncertainty weighting methods each with various amounts of training data per-task. We find that multitask learning typically does not improve performance for a user-defined combination of tasks. Further experiments evaluated on diverse tasks and network architectures on various datasets suggested that multitask learning requires careful selection of both task pairs and weighting strategies to equal or exceed the performance of single task learning.INDEX TERMS Dynamic weighting average, multi-MNIST, multi-objective optimization, multi-task learning, uncertainty weighting.

Background: In pediatrics, communication often occurs through an intermediary such as a caregiver. The goal of this study is to assess caregiver communication expectations and determine if meeting expectations influences caregiver satisfaction or instruction retention. Methods: A survey study was performed at the Children's Hospital of Philadelphia. Before the visit, caregivers completed a survey on communication expectations, Caregiver Expected Kalamazoo Essential Elements Communication Checklist (Caregiver Expected KEECC). After the visit, caregivers were surveyed on their perception of physician communication (Caregiver Perceived KEECC) and satisfaction. Caregivers were contacted 1 week after the clinic visit to assess instruction retention. Meeting of caregiver expectation was calculated by the difference between Caregiver Expected and Caregiver Perceived KEECC scores. Results: 112 caregivers participated in the study. There was no significant difference in Caregiver Expected KEECC versus Caregiver Perceived KEECC score (4.39 vs 4.56). Caregiver communication expectations were exceeded in 51.5% of the visits. Communication expectations were exceeded more among caregivers with at a college education (p < 0.01) and more among White caregivers (p < 0.01). The average caregiver satisfaction score with the clinic visit was 4.67. Higher satisfaction scores were observed in caregivers who had their communication expectations met or exceeded (p < 0.01). Caregivers with communication expectations exceeded had higher percentage recall of physician instructions (p < 0.01). Conclusions: Caregiver communication expectations may be influenced by demographic factors. Communication expectation affects visit outcomes including caregiver satisfaction and instruction retention. Therefore, physicians need to be cognizant of caregiver communication expectations, which can impact quality of the healthcare experience.

Animals throughout the animal kingdom excel at extracting individual sounds from competing background sounds, yet current state-of-the-art signal processing algorithms struggle to process speech in the presence of even modest background noise. Recent psychophysical experiments in humans and electrophysiological recordings in animal models suggest that the brain is adapted to process sounds within the restricted domain of spectro-temporal modulations found in natural sounds. Here, we describe a novel single microphone noise reduction algorithm called spectro-temporal detection–reconstruction (STDR) that relies on an artificial neural network trained to detect, extract and reconstruct the spectro-temporal features found in speech. STDR can significantly reduce the level of the background noise while preserving the foreground speech quality and improving estimates of speech intelligibility. In addition, by leveraging the strong temporal correlations present in speech, the STDR algorithm can also operate on predictions of upcoming speech features, retaining similar performance levels while minimizing inherent throughput delays. STDR performs better than a competing state-of-the-art algorithm for a wide range of signal-to-noise ratios and has the potential for real-time applications such as hearing aids and automatic speech recognition.

Background Prior evaluations of automated speech recognition (ASR) to create hospital progress notes have not analyzed its effect on professional revenue billing codes. As ASR becomes a more common method of entering clinical notes, clinicians, hospital administrators, and payers should understand whether this technology alters charges associated with inpatient physician services. Objectives This study aimed to measure the difference in professional fee charges between using voice and keyboard to create inpatient progress notes. Methods In a randomized trial of a novel voice with ASR system, called voice-generated enhanced electronic note system (VGEENS), to generate physician notes, we compared 1,613 notes created using intervention (VGEENS) or control (keyboard with template) created by 31 physicians. We measured three outcomes, as follows: (1) professional fee billing levels assigned by blinded coders, (2) number of elements within each note domain, and (3) frequency of organ system evaluations documented in review of systems (ROS) and physical exam. Results Participants using VGEENS generated a greater portion of high-level (99233) notes than control users (31.8 vs. 24.3%, p < 0.01). After adjustment for clustering by author, the finding persisted; intervention notes were 1.43 times more likely (95% confidence interval [CI]: 1.14–1.79) to receive a high-level code. Notes created using voice contained an average of 1.34 more history of present illness components (95% CI: 0.14–2.54) and 1.62 more review of systems components (95% CI: 0.48–2.76). The number of physical exam components was unchanged. Conclusion Using this voice with ASR system as tested slightly increases documentation of patient symptom details without reliance on copy and paste and may raise physician charges. Increased provider reimbursement may encourage hospital and provider group to offer use of voice and ASR to create hospital progress notes as an alternative to usual methods.

Introduction The objective of this study was to describe interruptions in the pediatric ambulatory setting and to assess their impact on perceived physician communication, patient satisfaction and recall of provided physician instructions. Methods An observational study was performed at the Children’s Hospital of Philadelphia, Pediatric Gastroenterology clinic. Participation consisted of video recording the clinic visit and the caregiver completed post-visit surveys on communication and satisfaction. Video recordings were coded for interruptions, which were divided into 3 main categories: Visit Associated, Pediatric Associated, and Unanticipated. An interruption rate was calculated and correlated with the following outcome variables to assess the impact of interruptions: caregiver satisfaction, caregiver perception on the quality of physician communication, and caregiver instruction recall. Results There were 675 interruptions noted in the 81 clinic visits, with an average of 7.96 (σ = 7.68) interruptions per visit. Six visits had no interruptions. The Patient was the most frequent interrupter. Significantly higher interruption rates occurred in clinic visits with younger patients (<7 years old) with most of the interruptions being Pediatric Associated interruptions. There was minimal correlation between the clinic visit interruption rate and caregiver satisfaction with the communication, caregiver perception of quality of communication, or caregiver instruction recall rate. Conclusion The effect of interruptions on the pediatric visit remains unclear. Interruptions may be part of the communication process to ensure alignment of the patient’s agenda. Additional studies are needed to help determine the impact of interruptions and guide medical education on patient communication.

No abstract