1. |
- Amirshahi, Alireza, et al.
(author)
-
M2D2 : Maximum-Mean-Discrepancy Decoder for Temporal Localization of Epileptic Brain Activities
- 2023
-
In: IEEE Journal of Biomedical and Health Informatics. - 2168-2194. ; 27:1, s. 202-214
-
Journal article (peer-reviewed)abstract
- Recent years have seen growing interest in leveraging deep learning models for monitoring epilepsy patients based on electroencephalographic (EEG) signals. However, these approaches often exhibit poor generalization when applied outside of the setting in which training data was collected. Furthermore, manual labeling of EEG signals is a time-consuming process requiring expert analysis, making fine-tuning patient-specific models to new settings a costly proposition. In this work, we propose the Maximum-Mean-Discrepancy Decoder (M2D2) for automatic temporal localization and labeling of seizures in long EEG recordings to assist medical experts. We show that M2D2 achieves 76.0% and 70.4% of F1-score for temporal localization when evaluated on EEG data gathered in a different clinical setting than the training data. The results demonstrate that M2D2 yields substantially higher generalization performance than other state-of-the-art deep learning-based approaches.
|
|
2. |
- Baghersalimi, Saleh, et al.
(author)
-
Decentralized Federated Learning for Epileptic Seizures Detection in Low-Power Wearable Systems
- 2024
-
In: IEEE Transactions on Mobile Computing. - 1536-1233. ; 23:5, s. 6392-6407
-
Journal article (peer-reviewed)abstract
- In healthcare, data privacy of patients regulations prohibits data from being moved outside the hospital, preventing international medical datasets from being centralized for AI training. Federated learning (FL) is a data privacy-focused method that trains a global model by aggregating local models from hospitals. Existing FL techniques adopt a central server-based network topology, where the server assembles the local models trained in each hospital to create a global model. However, the server could be a point of failure, and models trained in FL usually have worse performance than those trained in the centralized learning manner when the patient's data are not independent and identically distributed (Non-IID) in the hospitals. This paper presents a decentralized FL framework, including training with adaptive ensemble learning and a deployment phase using knowledge distillation. The adaptive ensemble learning step in the training phase leads to the acquisition of a specific model for each hospital that is the optimal combination of local models and models from other available hospitals. This step solves the non-IID challenges in each hospital. The deployment phase adjusts the model's complexity to meet the resource constraints of wearable systems. We evaluated the performance of our approach on edge computing platforms using EPILEPSIAE and TUSZ databases, which are public epilepsy datasets.
|
|
3. |
- Baghersalimi, Saleh, et al.
(author)
-
Personalized Real-Time Federated Learning for Epileptic Seizure Detection
- 2022
-
In: IEEE Journal of Biomedical and Health Informatics. - 2168-2194. ; 26:2
-
Journal article (peer-reviewed)abstract
- Epilepsy is one of the most prevalent paroxystic neurological disorders. It is characterized by the occurrence of spontaneous seizures. About 1 out of 3 patients have drug-resistant epilepsy, thus their seizures cannot be controlled by medication. Automatic detection of epileptic seizures can substantially improve the patient's quality of life. To achieve a high-quality model, we have to collect data from various patients in a central server. However, sending the patient's raw data to this central server puts patient privacy at risk and consumes a significant amount of energy. To address these challenges, in this work, we have designed and evaluated a standard federated learning framework in the context of epileptic seizure detection using a deep learning-based approach, which operates across a cluster of machines. We evaluated the accuracy and performance of our proposed approach on the NVIDIA Jetson Nano Developer Kit based on the EPILEPSIAE database, which is one of the largest public epilepsy datasets for seizure detection. Our proposed framework achieved a sensitivity of 81.25%, a specificity of 82.00%, and a geometric mean of 81.62%. It can be implemented on embedded platforms that complete the entire training process in 1.86 hours using 344.34 mAh energy on a single battery charge. We also studied a personalized variant of the federated learning, where each machine is responsible for training a deep neural network (DNN) to learn the discriminative electrocardiography (ECG) features of the epileptic seizures of the specific person monitored based on its local data. In this context, the DNN benefitted from a well-trained model without sharing the patient's raw data with a server or a central cloud repository. We observe in our results that personalized federated learning provides an increase in all the performance metric, with a sensitivity of 90.24%, a specificity of 91.58%, and a geometric mean of 90.90%.
|
|
4. |
- Constantin, Jeremy, et al.
(author)
-
An Ultra-Low-Power Application-Specific Processor for Compressed Sensing
- 2013
-
In: IFIP Advances in Information and Communication Technology. - Berlin, Heidelberg : Springer Berlin Heidelberg. - 9783642450723 - 9783642450730 ; 418, s. 88-106
-
Book chapter (peer-reviewed)abstract
- Compressed sensing (CS) is a universal low-complexity data compression technique for signals that have a sparse representation in some domain. While CS data compression can be done both in the analog- and digital domain, digital implementations are often used on low-power sensor nodes, where an ultra-low-power (ULP) processor carries out the algorithm on Nyquist-rate sampled data. In such systems an energy-efficient implementation of the CS compression kernel is a vital ingredient to maximize battery lifetime. In this paper, we propose an application-specific instruction-set processor (ASIP) processor that has been optimized for CS data compression and for operation in the subthreshold (sub-VT) regime. The design is equipped with specific sub-VT capable standard-cell based memories, to enable low-voltage operation with low leakage. Our results show that the proposed ASIP accomplishes 62× speed-up and 11.6× power savings with respect to a straightforward CS implementation running on the baseline low-power processor without instruction set extensions.
|
|
5. |
- Constantin, Jeremy, et al.
(author)
-
TamaRISC-CS: An Ultra-Low-Power Application-Specific Processor for Compressed Sensing
- 2012
-
In: IEEE/IFIP 20th International Conference on VLSI and System-on-Chip (VLSI-SoC), 2012. - 9781467326575 ; , s. 159-164
-
Conference paper (peer-reviewed)abstract
- Compressed sensing (CS) is a universal technique for the compression of sparse signals. CS has been widely used in sensing platforms where portable, autonomous devices have to operate for long periods of time with limited energy resources. Therefore, an ultra-low-power (ULP) CS implementation is vital for these kind of energy-limited systems. Sub-threshold (sub-VT) operation is commonly used for ULP computing, and can also be combined with CS. However, most established CS implementations can achieve either no or very limited benefit from sub-VT operation. Therefore, we propose a sub-VT application-specific instruction-set processor (ASIP), exploiting the specific operations of CS. Our results show that the proposed ASIP accomplishes 62x speed-up and 11.6x power savings with respect to an established CS implementation running on the baseline low-power processor.
|
|
6. |
- De Giovanni, Elisabetta, et al.
(author)
-
Real-Time Personalized Atrial Fibrillation Prediction on Multi-Core Wearable Sensors
- 2021
-
In: IEEE Transactions on Emerging Topics in Computing. - 2168-6750. ; 9:4, s. 1654-1666
-
Journal article (peer-reviewed)abstract
- In the recent Internet-of-Things (IoT) era where biomedical applications require continuous monitoring of relevant data, edge computing keeps gaining more and more importance. These new architectures for edge computing include multi-core and parallel computing capabilities that can enable prevention diagnosis and treatment of diseases in ambulatory or home-based setups. In this article, we explore the benefits of the parallelization capabilities and computing heterogeneity of new wearable sensors in the context of a personalized online atrial fibrillation (AF) prediction method for daily monitoring. First, we apply optimizations to a single-core design to reduce energy, based on patient-specific training models. Second, we explore multi-core and memory banks configuration changes to adapt the computation and storage requirements to the characteristics of each patient. We evaluate our methodology on the Physionet Prediction Challenge (2001) publicly available database, and assess the energy consumption of single-core (ARM Cortex-M3 based) and new ultra-low power multi-core architectures (open-source RISC-V based) for next-generation of wearable platforms. Overall, our exploration at the application level highlights that a parallelization approach for personalized AF in multi-core wearable sensors enables energy savings up to 24% with respect to single-core sensors. Moreover, including the adaptation of the memory subsystem (size and number of memory banks), in combination with deep sleep energy saving modes, can overall provide total energy savings up to 34%, depending on the specific patient.
|
|
7. |
- Forooghifar, Farnaz, et al.
(author)
-
Self-Aware Anomaly-Detection for Epilepsy Monitoring on Low-Power Wearable Electrocardiographic Devices
- 2021
-
In: 2021 IEEE 3rd International Conference on Artificial Intelligence Circuits and Systems, AICAS 2021. - 9781665419130
-
Conference paper (peer-reviewed)abstract
- Low-power wearable technologies offer a promising solution to pervasive epilepsy monitoring by removing the constraints concerning time and location, on one hand, and fulfilling long-term tracking, on the other hand. In the case of epileptic seizures, as the attacks infrequently occur, using an anomaly detection approach reduces the need to record long hours of data for each patient before detecting the successive coming seizures. In this work, by combining the concepts of self-aware system and anomaly detection, we propose an energy-efficient system to detect epileptic seizures on single-lead electrocardiographic signals, which is personalized after analyzing the first seizure of the patient. This system, then, uses a simple anomaly-detection model, whenever the model is deemed reliable, and uses a more complex model otherwise. We show that after the personalization, the number of patients, for which the method provides high sensitivity, can reach 26 out of 43 patients with the false alarm rate (FAR) of 4 alarms/day. Thus, the number of responders to the system is increased by 24%, while the FAR is only increased by one alarm/day, compared to the system that just uses the simple model. This benefit occurs while the system complexity decreases by 27.7% compared to the complex model. After adding the two-level (simple and complex) anomaly-detection, the complexity is tuned between 72.3% and 37.6% of the complex model. Similarly, the sensitivity is tuned between 66.5% and 60.3%.
|
|
8. |
- Huang, Baichuan, et al.
(author)
-
EpilepsyNet: Interpretable Self-Supervised Seizure Detection for Low-Power Wearable Systems
- 2023
-
In: IEEE 5th International Conference on Artificial Intelligence Circuits and Systems (AICAS). - 9798350332674
-
Conference paper (peer-reviewed)abstract
- Epilepsy is one of the most common neurological disorders that is characterized by recurrent and unpredictable seizures. Wearable systems can be used to detect the onset of a seizure and notify family members and emergency units for rescue. The majority of state-of-the-art studies in the epilepsy domain currently explore modern machine learning techniques, e.g., deep neural networks, to accurately detect epileptic seizures. However, training deep learning networks requires a large amount of data and computing resources, which is a major challenge for resource-constrained wearable systems. In this paper, we propose EpilepsyNet, the first interpretable self-supervised network tailored to resource-constrained devices without using any seizure data in its initial offline training. At runtime, however, once a seizure is detected, it can be incorporated into our self-supervised technique to improve seizure detection performance, without the need to retrain our learning model, hence incurring no energy overheads. Our self-supervised approach can reach a detection performance of 79.2%, which is on par with the state-of-the-art fully-supervised deep neural networks trained on seizure data. At the same time, our proposed approach can be deployed in resource-constrained wearable devices, reaching up to 1.3 days of battery life on a single charge.
|
|
9. |
- Masinelli, Giulio, et al.
(author)
-
Self-Aware Machine Learning for Multimodal Workload Monitoring during Manual Labor on Edge Wearable Sensors
- 2020
-
In: IEEE design & test. - 2168-2356 .- 2168-2364. ; 37:5, s. 58-66
-
Journal article (peer-reviewed)abstract
- Editor's notes: This article discusses self-awareness in wearable edge devices to enable real-time and long-term health monitoring. The authors use the notion of self-awareness to improve the battery life of edge wearable sensors for multimodal health and workload monitoring. This approach leads to a 27.6% lower energy consumption with less than 6% of performance loss. -Umit Y. Ogras, Arizona State University.
|
|
10. |
- Momeni, Niloofar, et al.
(author)
-
CAFS : Cost-Aware Features Selection Method for Multimodal Stress Monitoring on Wearable Devices
- 2022
-
In: IEEE Transactions on Biomedical Engineering. - 0018-9294. ; 69:3, s. 1072-1084
-
Journal article (peer-reviewed)abstract
- Objective: Today, stress monitoring on wearable devices is challenged by the tension between high-detection accuracy and battery lifetime driven by multimodal data acquisition and processing. Limited research has addressed the classification cost on multimodal wearable sensors, particularly when the features are cost-dependent. Thus, we design a Cost-Aware Feature Selection (CAFS) methodology that trades-off between prediction-power and energy-cost for multimodal stress monitoring. Methods: CAFS selects the most important features under different energy-constraints, which allows us to obtain energy-scalable stress monitoring models. We further propose a self-aware stress monitoring method that intelligently switches among the energy-scalable models, reducing energy consumption. Results: Using CAFS methodology on experimental data and simulation, we reduce the energy-cost of the stress model designed without energy constraints up to 94.37%. We obtain 90.98% and 95.74% as the best accuracy and confidence values, respectively, on unseen data, outperforming state-of-the-art studies. Analyzing our interpretable and energy-scalable models, we showed that simple models using only heart rate (HR) or skin conductance level (SCL), confidently predict acute stress for HR>93.30BPM and non-stress for SCL< 6.42 μS, but, outside these values, a multimodal model using respiration and pulse wave's features is needed for confident classification. Our self-aware acute stress monitoring proposal saves 10x energy and provides 88.72% of accuracy on unseen data. Conclusion: We propose a comprehensive solution for the cost-aware acute stress monitoring design addressing the problem of selecting an optimized feature subset considering their cost-dependency and cost-constraints. Significant: Our design framework enables long-term and confident acute stress monitoring on wearable devices.
|
|