The Device-Object Pairing Problem: Matching IoT Devices with Video Objects in a Multi-Camera Environment.

IoT technologies enable millions of devices to transmit their sensor data to the external world. The device-object pairing problem arises when a group of Internet of Things is concurrently tracked by cameras and sensors. While cameras view these things as visual "objects", these things which are equipped with "sensing devices" also continuously report their status. The challenge is that when visualizing these things on videos, their status needs to be placed properly on the screen. This requires correctly pairing visual objects with their sensing devices. There are many real-life examples. Recognizing a vehicle in videos does not imply that we can read its pedometer and fuel meter inside. Recognizing a pet on screen does not mean that we can correctly read its necklace data. In more critical ICU environments, visualizing all patients and showing their physiological signals on screen would greatly relieve nurses' burdens. The barrier behind this is that the camera may see an object but not be able to see its carried device, not to mention its sensor readings. This paper addresses the device-object pairing problem and presents a multi-camera, multi-IoT device system that enables visualizing a group of people together with their wearable devices' data and demonstrating the ability to recover the missing bounding box.

Cooperative Sensing Data Collection and Distribution with Packet Collision Avoidance in Mobile Long-Thin Networks.

Mobile ad hoc networks (MANETs) have gained a lot of interests in research communities for the infrastructure-less self-organizing nature. A MANET with fleet cyclists using smartphones forms a two-tier mobile long-thin network (MLTN) along a common cycling route, where the high-tier network is composed of 3G/LTE interfaces and the low-tier network is composed of IEEE 802.11 interfaces. The low-tier network may consist of several path-like networks. This work investigates cooperative sensing data collection and distribution with packet collision avoidance in a two-tier MLTN. As numbers of cyclists upload their sensing data and download global fleet information frequently, serious bandwidth and latency problems may result if all members rely on their high-tier interfaces. We designed and analyzed a cooperative framework consisting of a distributed grouping mechanism, a group merging and splitting method, and a sensing data aggregation scheme. Through cooperation between the two tiers, the proposed framework outperforms existing works by significantly reducing the 3G/LTE data transmission and the number of 3G/LTE connections.

Improvements in dental care using a new mobile app with cloud services.

BACKGROUND/PURPOSE: Traditional dental care, which includes long-term oral hygiene maintenance and scheduled dental appointments, requires effective communication between dentists and patients. In this study, a new system was designed to provide a platform for direct communication between dentists and patients. METHODS: A new mobile app, Dental Calendar, combined with cloud services specific for dental care was created by a team constituted by dentists, computer scientists, and service scientists. This new system would remind patients about every scheduled appointment, and help them take pictures of their own oral cavity parts that require dental treatment and send them to dentists along with a symptom description. Dentists, by contrast, could confirm or change appointments easily and provide professional advice to their patients immediately. In this study, 26 dentists and 32 patients were evaluated by a questionnaire containing eight dental-service items before and after using this system. Paired sample t test was used for statistical analysis. RESULTS: After using the Dental Calendar combined with cloud services, dentists were able to improve appointment arrangements significantly, taking care of the patients with sudden worse prosthesis (p < 0.05). Patients also achieved significant improvement in appointment reminder systems, rearrangement of appointments in case of sudden worse prosthesis, and establishment of a direct relationship with dentists (p < 0.05). CONCLUSION: Our new mobile app, Dental Calendar, in combination with cloud services, provides efficient service to both dentists and patients, and helps establish a better relationship between them. It also helps dentists to arrange appointments for patients with sudden worsening of prosthesis function.

Sex differences in the association of long-term exposure to heat stress on kidney function in a large Taiwanese population study.

The incidence and prevalence of dialysis in Taiwan are high compared to other regions. Consequently, mitigating chronic kidney disease (CKD) and the worsening of kidney function have emerged as critical healthcare priorities in Taiwan. Heat stress is known to be a significant risk factor for CKD and kidney function impairment. However, differences in the impact of heat stress between males and females remains unexplored. We conducted this retrospective cross-sectional analysis using data from the Taiwan Biobank (TWB), incorporating records of the wet bulb globe temperature (WBGT) during midday (11 AM-2 PM) and working hours (8 AM-5 PM) periods based on the participants' residential address. Average 1-, 3-, and 5-year WBGT values prior to the survey year were calculated and analyzed using a geospatial artificial intelligence-based ensemble mixed spatial model, covering the period from 2010 to 2020. A total of 114,483 participants from the TWB were included in this study, of whom 35.9% were male and 1053 had impaired kidney function (defined as estimated glomerular filtration rate < 60 ml/min/1.73 m2). Multivariable analysis revealed that in the male participants, during the midday period, the 1-, 3-, and 5-year average WBGT values per 1 â„ƒ increase were significantly positively associated with eGFR < 60 ml/min/1.73 m2 (odds ratio [OR], 1.096, 95% confidence interval [CI] = 1.002-1.199, p = 0.044 for 1 year; OR, 1.093, 95% CI = 1.000-1.196, p = 0.005 for 3 years; OR, 1.094, 95% CI = 1.002-1.195, p = 0.045 for 5 years). However, significant associations were not found for the working hours period. In the female participants, during the midday period, the 1-, 3-, and 5-year average WBGT values per 1 â„ƒ increase were significantly negatively associated with eGFR < 60 ml/min/1.73 m2 (OR, 0.872, 95% CI = 0.778-0.976, p = 0.018 for 1 year; OR, 0.874, 95% CI = 0.780-0.978, p = 0.019 for 3 years; OR, 0.875, 95% CI = 0.784-0.977, p = 0.018 for 5 years). In addition, during the working hours period, the 1-, 3-, and 5-year average WBGT values per 1 â„ƒ increase were also significantly negatively associated with eGFR < 60 ml/min/1.73 m2 (OR, 0.856, 95% CI = 0.774-0.946, p = 0.002 for 1 year; OR, 0.856, 95% CI = 0.774-0.948, p = 0.003 for 3 years; OR, 0.853, 95% CI = 0.772-0.943, p = 0.002 for 5 years). In conclusion, our results revealed that increased WBGT was associated with impaired kidney function in males, whereas increased WBGT was associated with a protective effect against impaired kidney function in females. Further studies are needed to elucidate the exact mechanisms underlying these sex-specific differences.

MPVF: 4D Medical Image Inpainting by Multi-Pyramid Voxel Flows.

Generatinga detailed 4D medical image usually accompanies with prolonged examination time and increased radiation exposure risk. Modern deep learning solutions have exploited interpolation mechanisms to generate a complete 4D image with fewer 3D volumes. However, existing solutions focus more on 2D-slice information, thus missing the changes on the z-axis. This article tackles the 4D cardiac and lung image interpolation problem by synthesizing 3D volumes directly. Although heart and lung only account for a fraction of chest, they constantly undergo periodical motions of varying magnitudes in contrast to the rest of the chest volume, which is more stationary. This poses big challenges to existing models. In order to handle various magnitudes of motions, we propose a Multi-Pyramid Voxel Flows (MPVF) model that takes multiple multi-scale voxel flows into account. This renders our generation network rich information during interpolation, both globally and regionally. Focusing on periodic medical imaging, MPVF takes the maximal and the minimal phases of an organ motion cycle as inputs and can restore a 3D volume at any time point in between. MPVF is featured by a Bilateral Voxel Flow (BVF) module for generating multi-pyramid voxel flows in an unsupervised manner and a Pyramid Fusion (PyFu) module for fusing multiple pyramids of 3D volumes. The model is validated to outperform the state-of-the-art model in several indices with significantly less synthesis time.

Representative Data Selection for Efficient Medical Incremental Learning.

To train a deep neural network relies on a large amount of annotated data. In special scenarios like industry defect detection and medical imaging, it is hard to collect sufficient labeled data all at once. Newly annotated data may arrive incrementally. In practice, we also prefer our target model to improve its capability gradually as new data comes in by quick re-training. This work tackles this problem from a data selection prospective by constraining ourselves to always retrain the target model with a fix amount of data after new data comes in. A variational autoencoder (VAE) and an adversarial network are combined for data selection, achieving fast model retraining. This enables the target model to continually learn from a small training set while not losing the information learned from previous iterations, thus incrementally adapting itself to new-coming data. We validate our framework on the LGG Segmentation dataset for the semantic segmentation task.Clinical relevance- The proposed VAE-based data selection model combined with adversarial training can choose a representative and reliable subset of data for time-efficient medical incremental learning. Users can immediately see the improvement of the medical segmentation model whenever new annotated images are contributed (after a few minutes of model retraining).

Decision tree-based learning to predict patient controlled analgesia consumption and readjustment.

BACKGROUND: Appropriate postoperative pain management contributes to earlier mobilization, shorter hospitalization, and reduced cost. The under treatment of pain may impede short-term recovery and have a detrimental long-term effect on health. This study focuses on Patient Controlled Analgesia (PCA), which is a delivery system for pain medication. This study proposes and demonstrates how to use machine learning and data mining techniques to predict analgesic requirements and PCA readjustment. METHODS: The sample in this study included 1099 patients. Every patient was described by 280 attributes, including the class attribute. In addition to commonly studied demographic and physiological factors, this study emphasizes attributes related to PCA. We used decision tree-based learning algorithms to predict analgesic consumption and PCA control readjustment based on the first few hours of PCA medications. We also developed a nearest neighbor-based data cleaning method to alleviate the class-imbalance problem in PCA setting readjustment prediction. RESULTS: The prediction accuracies of total analgesic consumption (continuous dose and PCA dose) and PCA analgesic requirement (PCA dose only) by an ensemble of decision trees were 80.9% and 73.1%, respectively. Decision tree-based learning outperformed Artificial Neural Network, Support Vector Machine, Random Forest, Rotation Forest, and Naïve Bayesian classifiers in analgesic consumption prediction. The proposed data cleaning method improved the performance of every learning method in this study of PCA setting readjustment prediction. Comparative analysis identified the informative attributes from the data mining models and compared them with the correlates of analgesic requirement reported in previous works. CONCLUSION: This study presents a real-world application of data mining to anesthesiology. Unlike previous research, this study considers a wider variety of predictive factors, including PCA demands over time. We analyzed PCA patient data and conducted several experiments to evaluate the potential of applying machine-learning algorithms to assist anesthesiologists in PCA administration. Results demonstrate the feasibility of the proposed ensemble approach to postoperative pain management.

3D Gait Tracking by Acoustic Doppler Effects.

Rehabilitation and physical therapies can recover people suffering from neurological disorder. Due to limited medical personnels, there are not enough medical personnels help patients with their posture diagnosis. In this paper, we propose a 3D gait tracking method to help medical personnels monitor patients. Based on acoustic signals, our approach derives displacement by only one integration of velocity. When one walks, his feet move back and forth, causing relative movements to our acoustic sensors, which we call self-Doppler effect. We utilize three buzzers and one microphone mounted on feet to collect the frequency shifts caused by relative movements and measure 3D trajectories. We validate through simulations that this approach would perform very well. In real experiments, due to the existence of noise and the limitation of hardware, we observe an average error of 0.1669 m in step length estimation and 0.0867 m in step height estimation.