Predicting Adverse Behavior in Individuals with Autism Spectrum Disorder Through Off-body Sleep Analysis

Poor sleep quality in Autism Spectrum Disorder (ASD) individuals is linked to severe daytime behaviors. This study explores the relationship between a prior night's sleep structure and its predictive power for next-day behavior in ASD individuals. The motion was extracted using a low-cost near-infrared camera in a privacy-preserving way. Over two years, we recorded overnight data from 14 individuals, spanning over 2,000 nights, and tracked challenging daytime behaviors, including aggression, self-injury, and disruption. We developed an ensemble machine learning algorithm to predict next-day behavior in the morning and the afternoon. Our findings indicate that sleep quality is a more reliable predictor of morning behavior than afternoon behavior the next day. The proposed model attained an accuracy of 74% and a F1 score of 0.74 in target-sensitive tasks and 67% accuracy and 0.69 F1 score in target-insensitive tasks. For 7 of the 14, better-than-chance balanced accuracy was obtained (p-value<0.05), with 3 showing significant trends (p-value<0.1). These results suggest off-body, privacy-preserving sleep monitoring as a viable method for predicting next-day adverse behavior in ASD individuals, with the potential for behavioral intervention and enhanced care in social and learning settings.

Indoor Group Identification and Localization Using Privacy-Preserving Edge Computing Distributed Camera Network

Social interaction behaviors change as a result of both physical and psychiatric problems, and it is important to identify subtle changes in group activity engagements for monitoring the mental health of patients in clinics. This work proposes a system to identify when and where group formations occur in an approximately 1700 m2 therapeutic built environment using a distributed edge-computing camera network. The proposed method can localize group formations when provided with noisy positions and orientations of individuals, estimated from sparsely distributed multiview cameras, which run a lightweight multiperson 2-D pose detection model. Our group identification method demonstrated an F1 score of up to 90% with a mean absolute error of 1.25 m for group localization on our benchmark dataset. The dataset consisted of seven subjects walking, sitting, and conversing for 35 min in groups of various sizes ranging from 2 to 7 subjects. The proposed system is low-cost and scalable to any ordinary building to transform the indoor space into a smart environment using edge computing systems. We expect the proposed system to enhance existing therapeutic units for passively monitoring the social behaviors of patients when implementing real-time interventions.

Graph Trilateration for Indoor Localization in Sparsely Distributed Edge Computing Devices in Complex Environments Using Bluetooth Technology

Spatial navigation patterns in indoor space usage can reveal important cues about the cognitive health of participants. In this work, we present a low-cost, scalable, open-source edge computing system using Bluetooth low energy (BLE) beacons for tracking indoor movements in a large, 1700 m2 facility used to carry out therapeutic activities for participants with mild cognitive impairment (MCI). The facility is instrumented with 39 edge computing systems, along with an on-premise fog server. The participants carry a BLE beacon, in which BLE signals are received and analyzed by the edge computing systems. Edge computing systems are sparsely distributed in the wide, complex indoor space, challenging the standard trilateration technique for localizing subjects, which assumes a dense installation of BLE beacons. We propose a graph trilateration approach that considers the temporal density of hits from the BLE beacon to surrounding edge devices to handle the inconsistent coverage of edge devices. This proposed method helps us tackle the varying signal strength, which leads to intermittent detection of beacons. The proposed method can pinpoint the positions of multiple participants with an average error of 4.4 m and over 85% accuracy in region-level localization across the entire study area. Our experimental results, evaluated in a clinical environment, suggest that an ordinary medical facility can be transformed into a smart space that enables automatic assessment of individuals' movements, which may reflect health status or response to treatment.

Heart murmur detection from phonocardiogram recordings: The George B. Moody PhysioNet Challenge 2022

Cardiac auscultation is an accessible diagnostic screening tool that can help to identify patients with heart murmurs, who may need follow-up diagnostic screening and treatment for abnormal cardiac function. However, experts are needed to interpret the heart sounds, limiting the accessibility of cardiac auscultation in resource-constrained environments. Therefore, the George B. Moody PhysioNet Challenge 2022 invited teams to develop algorithmic approaches for detecting heart murmurs and abnormal cardiac function from phonocardiogram (PCG) recordings of heart sounds. For the Challenge, we sourced 5272 PCG recordings from 1452 primarily pediatric patients in rural Brazil, and we invited teams to implement diagnostic screening algorithms for detecting heart murmurs and abnormal cardiac function from the recordings. We required the participants to submit the complete training and inference code for their algorithms, improving the transparency, reproducibility, and utility of their work. We also devised an evaluation metric that considered the costs of screening, diagnosis, misdiagnosis, and treatment, allowing us to investigate the benefits of algorithmic diagnostic screening and facilitate the development of more clinically relevant algorithms. We received 779 algorithms from 87 teams during the Challenge, resulting in 53 working codebases for detecting heart murmurs and abnormal cardiac function from PCG recordings. These algorithms represent a diversity of approaches from both academia and industry, including methods that use more traditional machine learning techniques with engineered clinical and statistical features as well as methods that rely primarily on deep learning models to discover informative features. The use of heart sound recordings for identifying heart murmurs and abnormal cardiac function allowed us to explore the potential of algorithmic approaches for providing more accessible diagnostic screening in resource-constrained environments. The submission of working, open-source algorithms and the use of novel evaluation metrics supported the reproducibility, generalizability, and clinical relevance of the research from the Challenge.

COVID-19 pneumonia chest radiographic severity score: variability assessment among experienced and in-training radiologists and creation of a multireader composite score database for artificial intelligence algorithm development

OBJECTIVE

The purpose was to evaluate reader variability between experienced and in-training radiologists of COVID-19 pneumonia severity on chest radiograph (CXR), and to create a multireader database suitable for AI development.

METHODS

In this study, CXRs from polymerase chain reaction positive COVID-19 patients were reviewed. Six experienced cardiothoracic radiologists and two residents classified each CXR according to severity. One radiologist performed the classification twice to assess intraobserver variability. Severity classification was assessed using a 4-class system: normal (0), mild (1), moderate (2), and severe (3). A median severity score (Rad Med) for each CXR was determined for the six radiologists for development of a multireader database (XCOMS). Kendal Tau correlation and percentage of disagreement were calculated to assess variability.

RESULTS

A total of 397 patients (1208 CXRs) were included (mean age, 60 years SD ± 1), 189 men). Interobserver variability between the radiologists ranges between 0.67 and 0.78. Compared to the Rad Med score, the radiologists show good correlation between 0.79-0.88. Residents show slightly lower interobserver agreement of 0.66 with each other and between 0.69 and 0.71 with experienced radiologists. Intraobserver agreement was high with a correlation coefficient of 0.77. In 220 (18%), 707 (59%), 259 (21%) and 22 (2%) CXRs there was a 0, 1, 2 or 3 class-difference. In 594 (50%) CXRs the median scores of the residents and the radiologists were similar, in 578 (48%) and 36 (3%) CXRs there was a 1 and 2 class-difference.

CONCLUSION

Experienced and in-training radiologists demonstrate good inter- and intraobserver agreement in COVID-19 pneumonia severity classification. A higher percentage of disagreement was observed in moderate cases, which may affect training of AI algorithms.

ADVANCES IN KNOWLEDGE

Most AI algorithms are trained on data labeled by a single expert. This study shows that for COVID-19 X-ray severity classification there is significant variability and disagreement between radiologist and between residents.

Inverse design of photonic nanostructures using dimensionality reduction: reducing the computational complexity

In this Letter, we present a deep-learning-based method using neural networks (NNs) for inverse design of photonic nanostructures. We show that by using dimensionality reduction in both the design and the response spaces, the computational complexity of the inverse design algorithm is considerably reduced. As a proof of concept, we apply this method to design multi-layer thin-film structures composed of consecutive layers of two different dielectrics and compare the results using our techniques to those using conventional NNs.

Toward understanding COVID-19 pneumonia: a deep-learning-based approach for severity analysis and monitoring the disease

We report a new approach using artificial intelligence (AI) to study and classify the severity of COVID-19 using 1208 chest X-rays (CXRs) of 396 COVID-19 patients obtained through the course of the disease at Emory Healthcare affiliated hospitals (Atlanta, GA, USA). Using a two-stage transfer learning technique to train a convolutional neural network (CNN), we show that the algorithm is able to classify four classes of disease severity (normal, mild, moderate, and severe) with the average Area Under the Curve (AUC) of 0.93. In addition, we show that the outputs of different layers of the CNN under dominant filters provide valuable insight about the subtle patterns in the CXRs, which can improve the accuracy in the reading of CXRs by a radiologist. Finally, we show that our approach can be used for studying the disease progression in a single patient and its influencing factors. The results suggest that our technique can form the foundation of a more concrete clinical model to predict the evolution of COVID-19 severity and the efficacy of different treatments for each patient through using CXRs and clinical data in the early stages of the disease. This use of AI to assess the severity and possibly predicting the future stages of the disease early on, will be essential in dealing with the upcoming waves of COVID-19 and optimizing resource allocation and treatment.