Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference最新文献

The success of deep learning methods has enabled many modern wearable health applications, but has also highlighted the critical caveat of their extremely data hungry nature. While the widely explored wrist and finger photoplethysmography (PPG) sites are less affected, given the large available databases, this issue is prohibitive to exploring the full potential of novel recording locations such as in-ear wearables. To this end, we assess the feasibility of transfer learning from finger PPG to in-ear PPG in the context of deep learning for respiratory monitoring. This is achieved by introducing an encoder-decoder framework which is set up to extract respiratory waveforms from PPG, whereby simultaneously recorded gold standard respiratory waveforms (capnography, impedance pneumography and air flow) are used as a training reference. Next, the data augmentation and training pipeline is examined for both training on finger PPG and the subsequent fine tuning on in-ear PPG. The results indicate that, through training on two large finger PPG data sets (95 subjects) and then retraining on our own small in-ear PPG data set (6 subjects), the model achieves lower and more consistent test error for the prediction of the respiratory waveforms, compared to training on the small in-ear data set alone. This conclusively demonstrates the feasibility of transfer learning from finger PPG to in-ear PPG, leading to better generalisation across a wide range of respiratory rates.

Computational models of neurons are valuable tools that allow researchers to form and evaluate hypotheses and minimize high-cost animal work. We soon plan to use computational modeling to explore the response of different sensory fiber types to long duration external stimulation to try to selectively block nociceptive C-fibers. In this work, we modified an existing C-fiber-specific axon model to additionally include concentration-dependent conductance changes, the contribution of longitudinal current flow to changes in local concentrations, and longitudinal currents generated by concentration gradients along the axon. Then, we examined the impact of these additional elements on the modeled action potential properties, activity-dependent latency increases, and concentration changes due to external stimulation. We found that these additional model elements did not significantly affect the action potential properties or activity-dependent behavior, but they did have a significant impact on the modeled response to external long duration stimulation.Clinical Relevance- This presents a computational model that can be used to help investigate and develop electrical stimulation therapies for pathological pain.

Federated learning (FL) is a machine learning framework that allows remote clients to collaboratively learn a global model while keeping their training data localized. It has emerged as an effective tool to solve the problem of data privacy protection. In particular, in the medical field, it is gaining relevance for achieving collaborative learning while protecting sensitive data. In this work, we demonstrate the feasibility of FL in the development of a deep learning model for screening diabetic retinopathy (DR) in fundus photographs. To this end, we conduct a simulated FL framework using nearly 700,000 fundus photographs collected from OPHDIAT, a French multi-center screening network for detecting DR. We develop two FL algorithms: 1) a cross-center FL algorithm using data distributed across the OPHDIAT centers and 2) a cross-grader FL algorithm using data distributed across the OPHDIAT graders. We explore and assess different FL strategies and compare them to a conventional learning algorithm, namely centralized learning (CL), where all the data is stored in a centralized repository. For the task of referable DR detection, our simulated FL algorithms achieved similar performance to CL, in terms of area under the ROC curve (AUC): AUC =0.9482 for CL, AUC = 0.9317 for cross-center FL and AUC = 0.9522 for cross-grader FL. Our work indicates that the FL algorithm is a viable and reliable framework that can be applied in a screening network.Clinical relevance- Given that data sharing is regarded as an essential component of modern medical research, achieving collaborative learning while protecting sensitive data is key.

Mild cognitive impairment (MCI) is considered the early stage of Alzheimer's disease, characterized as mild memory loss. A novel method of functional connectivity (FC) analysis can be used to detect MCI before memory is significantly impaired allowing for preventative measures to be taken. FC examines interactions between EEG channels to grant insight on underlying neural networks and analyze the effects of MCI. Applying FC method of weighted phase lag index (wPLI) to P300 ERPs provided insight on the link between the pathology of Alzheimer's disease and cognitive loss. wPLI was analyzed per frequency band (θ, α, μ, β) and by channel combination groups (intra-hemispheric short, intra-hemispheric long, inter-hemispheric short, inter-hemispheric long, transverse). MCI was found to have a statistically significant lower ΔwPLI_P300 compared to normal controls in the μ intra-hemispheric short (p = 0.0286), μ intra-hemispheric long (p = 0.0477), μ inter-hemispheric short (p = 0.0018) and the α intra-hemispheric short (p = 0.0423). Results indicate a possible deficiency in the dorsal visual processing pathway among MCI subjects as well as an unbalanced coordination between the two hemispheres.

This study presents fingertip strain plethysmography (SPG) as a visual trace of cardiac cycles in peripheral vessels. The setup includes a small, sensitive MEMS strain sensor attached to the fingertip to capture the pulsatile vibrations corresponding to cardiac cycles. SPG is evaluated on 10 healthy subjects for the estimation of heart rate (HR) and heart rate variability (HRV), as well as heartbeat-derived respiratory rate (RR) which is an HRV parameter. The estimated parameters are compared with a simultaneously-recorded electrocardiogram (ECG) for HR and HRV, and an inertial sensor placed on the chest wall for RR. Bland-Altman analyses suggest small estimation biases of 0.03 beats-per-minute (BPM) and 0.38 ms for HR and HRV respectively, demonstrating excellent agreement between fingertip SPG and ECG. The average estimation accuracies of 99.88% (± 0.04), 96.43% (± 1.44), and 92.64% (± 2.30) for HR, HRV, and RR respectively, prove the reliability of SPG for hemodynamic monitoring.Clinical Relevance- Conventional plethysmography sensors are either cumbersome or susceptible to skin color. This effort is a fundamental step towards the augmentation of conventional methods, thus ensuring stable, clinical-grade hemodynamic monitoring.

Neural connectivity describes how neuron populations coordinate and create cognitive and behavioral functions. Neural connectivity performs dynamics where its population spiking responses to stimuli or intention change over time. Brain-machine interface (BMI) provides a framework for studying dynamical neural connectivity. In BMI, point process is a powerful technique in analyzing the single neuronal tuning. And generalized linear mode (GLM) as an encoding model can incorporate the tuning in kinematics and the neural connectivity. Quantification and tracking of dynamic neural connectivity can contribute to the elucidation of the generation of brain functions in a computational way. However, most of the previous work focused on single neuronal adaptation to kinematics. When a neuron is significantly modulated by some other neurons in some tasks, the shape of the log likelihood function for single neuronal observations can be narrowed in some dimensions. And the existing gradient-based methods are not able to reach the optimum in a fast and adaptive searching way. In this work, to maximize the likelihood of observations and obtain the dynamic neural connectivity tuning parameters, we proposed a conjugate gradient-based encoding model (CGE). We illustrate CGE for likelihood function using the real experimental data under manual control and brain control. The results show that the proposed CGE has better performance in tracking the dynamic neural connectivity tuning parameters and modeling neural encoding.Clinical Relevance- Not directly related.

Inflammatory bowel disease (IBD) is characterized by chronic inflammation in the intestinal tract. There is currently no effective cure for IBD. The aim of this study was to evaluate the protective effect of low-intensity pulsed ultrasound (LIPUS) on lipopolysaccharide (LPS)-induced intestinal damage in a C57BL/6 mouse model. Colonic inflammation was induced by LPS injection (0.75 mg/kg, i.p.) for 7 days. A 1.0 MHz ultrasound transducer was used with a duty cycle of 5% and a repetition frequency of 1 Hz. LIPUS was applied to the abdominal region for 15 min/day from days 1 to 6 at both intensity of 0.5 W/cm² or 1.0 W/cm². Colonic samples were collected for macroscopic and westerm blotting analysis. First, the optimal dose of LPS for experiments was investigated. Our results demonstrated that LIPUS alleviates colonic damage by reducing colon shortening and increasing the levels of tight junction proteins such as Occludin and ZO-1. These findings show that abdominal LIPUS stimulation may be a novel therapeutic strategy for IBD through enhancement of tight junction protein levels and attenuation of colonic length.

Unilateral spatial neglect (USN) is defined as the inability to attend and see on one side, which seriously interferes with daily life. Clinically, patients with left USN commonly demonstrate a striking immediate capture of attention from ipsilesional, right-sided items as soon as a visual scene unfolds (i.e., magnetic attraction [MA]). Therefore, this preliminary study utilized a three-dimensional (3D) virtual environment to evaluate the effects of eliminating stimuli in the rightward space and directing attention to the left on neglect symptoms.

Methods: Seven patients with USN participated in this study, and two types of visual stimuli were created: the numbers and objects in the 3D virtual environment. To eliminate the visual stimuli on the right side, a moving slit was introduced in the virtual environment. During the experiment, patients were required to orally identify each object and number both in moving and nonmoving slit conditions.

Results: A statistical comparison of scores with and without the moving slit in the 3D virtual space indicated significant changes in the object stimuli condition; however, no statistically significant difference was observed in the number stimuli condition.

Conclusions: Masking the right side within the 3D virtual space increased the number of objects that can be recognized on the left side by patients with USN. The results may allow interventions in a virtual reality environment that closely resembles the patient's real-life space.Clinical Relevance-Magnetic attraction is a symptom seen in patients in clinical practice, but there is no method of rehabilitation. The proposed moving slit method is expected to be effective because it enables attention guidance in a three-dimensional space.

Colorectal cancer (CRC) is one of the most common causes of cancer and cancer-related mortality worldwide. Performing colon cancer screening in a timely fashion is the key to early detection. Colonoscopy is the primary modality used to diagnose colon cancer. However, the miss rate of polyps, adenomas and advanced adenomas remains significantly high. Early detection of polyps at the precancerous stage can help reduce the mortality rate and the economic burden associated with colorectal cancer. Deep learning-based computer-aided diagnosis (CADx) system may help gastroenterologists to identify polyps that may otherwise be missed, thereby improving the polyp detection rate. Additionally, CADx system could prove to be a cost-effective system that improves long-term colorectal cancer prevention. In this study, we proposed a deep learning-based architecture for automatic polyp segmentation called Transformer ResU-Net (TransResU-Net). Our proposed architecture is built upon residual blocks with ResNet-50 as the backbone and takes advantage of the transformer self-attention mechanism as well as dilated convolution(s). Our experimental results on two publicly available polyp segmentation benchmark datasets showed that TransResU-Net obtained a highly promising dice score and a real-time speed. With high efficacy in our performance metrics, we concluded that TransResU-Net could be a strong benchmark for building a real-time polyp detection system for the early diagnosis, treatment, and prevention of colorectal cancer. The source code of the proposed TransResU-Net is publicly available at https://github.com/nikhilroxtomar/TransResUNet.

Vision-based gait analysis can play an important role in the remote and continuous monitoring of the elderly's health conditions. However, most vision-based approaches compute gait spatiotemporal parameters using human pose information and provide average parameters. This study aimed to propose a straightforward method for stride-by-stride gait spatiotemporal parameters estimation. A total of 160 elderly individuals participated in this study. Data were gathered with a GAITRite system and a mobile camera simultaneously. Three deep learning networks were trained with a few RGB frames as input and a continuous 1D signal containing both spatial and temporal gait parameters as output. The trained networks estimated the stride lengths with correlations of 0.938 and more and detected gait events with F₁-scores of 0.914 and more.Clinical relevance- The proposed method showed excellent agreements with the GAITRite system in analyzing spatiotemporal gait parameters. Our approach can be applied to monitor the elderly's health conditions based on their gait parameters for early diagnosis of diseases, proper treatment, and timely intervention.