| 1. |
- Ali, Muhaddisa Barat, 1986, et al.
(författare)
-
A Feasibility Study on Deep Learning Based Brain Tumor Segmentation Using 2D Ellipse Box Areas.
- 2022
-
Ingår i: Sensors (Basel, Switzerland). - : MDPI AG. - 1424-8220. ; 22:14
-
Tidskriftsartikel (refereegranskat)abstract
- In most deep learning-based brain tumor segmentation methods, training the deep network requires annotated tumor areas. However, accurate tumor annotation puts high demands on medical personnel. The aim of this study is to train a deep network for segmentation by using ellipse box areas surrounding the tumors. In the proposed method, the deep network is trained by using a large number of unannotated tumor images with foreground (FG) and background (BG) ellipse box areas surrounding the tumor and background, and a small number of patients (<20) with annotated tumors. The training is conducted by initial training on two ellipse boxes on unannotated MRIs, followed by refined training on a small number of annotated MRIs. We use a multi-stream U-Net for conducting our experiments, which is an extension of the conventional U-Net. This enables the use of complementary information from multi-modality (e.g., T1, T1ce, T2, and FLAIR) MRIs. To test the feasibility of the proposed approach, experiments and evaluation were conducted on two datasets for glioma segmentation. Segmentation performance on the test sets is then compared with those used on the same network but trained entirely by annotated MRIs. Our experiments show that the proposed method has obtained good tumor segmentation results on the test sets, wherein the dice score on tumor areas is (0.8407, 0.9104), and segmentation accuracy on tumor areas is (83.88%, 88.47%) for the MICCAI BraTS'17 and US datasets, respectively. Comparing the segmented results by using the network trained by all annotated tumors, the drop in the segmentation performance from the proposed approach is (0.0594, 0.0159) in the dice score, and (8.78%, 2.61%) in segmented tumor accuracy for MICCAI and US test sets, which is relatively small. Our case studies have demonstrated that training the network for segmentation by using ellipse box areas in place of all annotated tumors is feasible, and can be considered as an alternative, which is a trade-off between saving medical experts' time annotating tumors and a small drop in segmentation performance.
|
|
| 2. |
- Ali, Muhaddisa Barat, 1986, et al.
(författare)
-
A novel federated deep learning scheme for glioma and its subtype classification
- 2023
-
Ingår i: Frontiers in Neuroscience. - 1662-4548 .- 1662-453X. ; 17
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Deep learning (DL) has shown promising results in molecular-based classification of glioma subtypes from MR images. DL requires a large number of training data for achieving good generalization performance. Since brain tumor datasets are usually small in size, combination of such datasets from different hospitals are needed. Data privacy issue from hospitals often poses a constraint on such a practice. Federated learning (FL) has gained much attention lately as it trains a central DL model without requiring data sharing from different hospitals. Method: We propose a novel 3D FL scheme for glioma and its molecular subtype classification. In the scheme, a slice-based DL classifier, EtFedDyn, is exploited which is an extension of FedDyn, with the key differences on using focal loss cost function to tackle severe class imbalances in the datasets, and on multi-stream network to exploit MRIs in different modalities. By combining EtFedDyn with domain mapping as the pre-processing and 3D scan-based post-processing, the proposed scheme makes 3D brain scan-based classification on datasets from different dataset owners. To examine whether the FL scheme could replace the central learning (CL) one, we then compare the classification performance between the proposed FL and the corresponding CL schemes. Furthermore, detailed empirical-based analysis were also conducted to exam the effect of using domain mapping, 3D scan-based post-processing, different cost functions and different FL schemes. Results: Experiments were done on two case studies: classification of glioma subtypes (IDH mutation and wild-type on TCGA and US datasets in case A) and glioma grades (high/low grade glioma HGG and LGG on MICCAI dataset in case B). The proposed FL scheme has obtained good performance on the test sets (85.46%, 75.56%) for IDH subtypes and (89.28%, 90.72%) for glioma LGG/HGG all averaged on five runs. Comparing with the corresponding CL scheme, the drop in test accuracy from the proposed FL scheme is small (−1.17%, −0.83%), indicating its good potential to replace the CL scheme. Furthermore, the empirically tests have shown that an increased classification test accuracy by applying: domain mapping (0.4%, 1.85%) in case A; focal loss function (1.66%, 3.25%) in case A and (1.19%, 1.85%) in case B; 3D post-processing (2.11%, 2.23%) in case A and (1.81%, 2.39%) in case B and EtFedDyn over FedAvg classifier (1.05%, 1.55%) in case A and (1.23%, 1.81%) in case B with fast convergence, which all contributed to the improvement of overall performance in the proposed FL scheme. Conclusion: The proposed FL scheme is shown to be effective in predicting glioma and its subtypes by using MR images from test sets, with great potential of replacing the conventional CL approaches for training deep networks. This could help hospitals to maintain their data privacy, while using a federated trained classifier with nearly similar performance as that from a centrally trained one. Further detailed experiments have shown that different parts in the proposed 3D FL scheme, such as domain mapping (make datasets more uniform) and post-processing (scan-based classification), are essential.
|
|
| 3. |
- Ali, Muhaddisa Barat, 1986, et al.
(författare)
-
Domain Mapping and Deep Learning from Multiple MRI Clinical Datasets for Prediction of Molecular Subtypes in Low Grade Gliomas
- 2020
-
Ingår i: Brain Sciences. - : MDPI AG. - 2076-3425. ; 10:7, s. 1-20
-
Tidskriftsartikel (refereegranskat)abstract
- Brain tumors, such as low grade gliomas (LGG), are molecularly classified which require the surgical collection of tissue samples. The pre-surgical or non-operative identification of LGG molecular type could improve patient counseling and treatment decisions. However, radiographic approaches to LGG molecular classification are currently lacking, as clinicians are unable to reliably predict LGG molecular type using magnetic resonance imaging (MRI) studies. Machine learning approaches may improve the prediction of LGG molecular classification through MRI, however, the development of these techniques requires large annotated data sets. Merging clinical data from different hospitals to increase case numbers is needed, but the use of different scanners and settings can affect the results and simply combining them into a large dataset often have a significant negative impact on performance. This calls for efficient domain adaption methods. Despite some previous studies on domain adaptations, mapping MR images from different datasets to a common domain without affecting subtitle molecular-biomarker information has not been reported yet. In this paper, we propose an effective domain adaptation method based on Cycle Generative Adversarial Network (CycleGAN). The dataset is further enlarged by augmenting more MRIs using another GAN approach. Further, to tackle the issue of brain tumor segmentation that requires time and anatomical expertise to put exact boundary around the tumor, we have used a tight bounding box as a strategy. Finally, an efficient deep feature learning method, multi-stream convolutional autoencoder (CAE) and feature fusion, is proposed for the prediction of molecular subtypes (1p/19q-codeletion and IDH mutation). The experiments were conducted on a total of 161 patients consisting of FLAIR and T1 weighted with contrast enhanced (T1ce) MRIs from two different institutions in the USA and France. The proposed scheme is shown to achieve the test accuracy of 74.81% on 1p/19q codeletion and 81.19% on IDH mutation, with marked improvement over the results obtained without domain mapping. This approach is also shown to have comparable performance to several state-of-the-art methods.
|
|
| 4. |
- Ali, Muhaddisa Barat, 1986, et al.
(författare)
-
Prediction of glioma‑subtypes: comparison of performance on a DL classifier using bounding box areas versus annotated tumors
- 2022
-
Ingår i: BioMedical Engineering Online. - : Springer Science and Business Media LLC. - 1475-925X .- 2524-4426. ; 4
-
Tidskriftsartikel (refereegranskat)abstract
- Background: For brain tumors, identifying the molecular subtypes from magnetic resonance imaging (MRI) isdesirable, but remains a challenging task. Recent machine learning and deep learning (DL) approaches may help theclassification/prediction of tumor subtypes through MRIs. However, most of these methods require annotated datawith ground truth (GT) tumor areas manually drawn by medical experts. The manual annotation is a time consumingprocess with high demand on medical personnel. As an alternative automatic segmentation is often used. However, itdoes not guarantee the quality and could lead to improper or failed segmented boundaries due to differences in MRIacquisition parameters across imaging centers, as segmentation is an ill‑defined problem. Analogous to visual objecttracking and classification, this paper shifts the paradigm by training a classifier using tumor bounding box areas inMR images. The aim of our study is to see whether it is possible to replace GT tumor areas by tumor bounding boxareas (e.g. ellipse shaped boxes) for classification without a significant drop in performance.Method: In patients with diffuse gliomas, training a deep learning classifier for subtype prediction by employ‑ing tumor regions of interest (ROIs) using ellipse bounding box versus manual annotated data. Experiments wereconducted on two datasets (US and TCGA) consisting of multi‑modality MRI scans where the US dataset containedpatients with diffuse low‑grade gliomas (dLGG) exclusively.Results: Prediction rates were obtained on 2 test datasets: 69.86% for 1p/19q codeletion status on US dataset and79.50% for IDH mutation/wild‑type on TCGA dataset. Comparisons with that of using annotated GT tumor data fortraining showed an average of 3.0% degradation (2.92% for 1p/19q codeletion status and 3.23% for IDH genotype).Conclusion: Using tumor ROIs, i.e., ellipse bounding box tumor areas to replace annotated GT tumor areas for train‑ing a deep learning scheme, cause only a modest decline in performance in terms of subtype prediction. With moredata that can be made available, this may be a reasonable trade‑off where decline in performance may be counter‑acted with more data.
|
|
| 5. |
- Bagheri, Azam, et al.
(författare)
-
A Framework Based on Machine Learning for Analytics of Voltage Quality Disturbances
- 2022
-
Ingår i: Energies. - : MDPI. - 1996-1073. ; 15:4
-
Tidskriftsartikel (refereegranskat)abstract
- This paper proposes a machine-learning-based framework for voltage quality analytics, where the space phasor model (SPM) of the three-phase voltages before, during, and after the event is applied as input data. The framework proceeds along with three main steps: (a) event extraction, (b) event characterization, and (c) additional information extraction. During the first step, it utilizes a Gaussian-based anomaly detection (GAD) technique to extract the event data from the recording. Principal component analysis (PCA) is adopted during the second step, where it is shown that the principal components correspond to the semi-minor and semi-major axis of the ellipse formed by the SPM. During the third step, these characteristics are interpreted to extract additional information about the underlying cause of the event. The performance of the framework was verified through experiments conducted on datasets containing synthetic and measured power quality events. The results show that the combination of semi-major axis, semi-minor axis, and direction of the major axis forms a sufficient base to characterize, classify, and eventually extract additional information from recorded event data.
|
|
| 6. |
- Bagheri, Azam, 1982, et al.
(författare)
-
Estimation of frequency-dependent impedances in power grids by deep lstm autoencoder and random forest
- 2021
-
Ingår i: Energies. - : MDPI AG. - 1996-1073 .- 1996-1073. ; 14:13
-
Tidskriftsartikel (refereegranskat)abstract
- This paper proposes a deep-learning-based method for frequency-dependent grid impedance estimation. Through measurement of voltages and currents at a specific system bus, the estimate of the grid impedance was obtained by first extracting the sequences of the time-dependent features for the measured data using a long short-term memory autoencoder (LSTM-AE) followed by a random forest (RF) regression method to find the nonlinear map function between extracted features and the corresponding grid impedance for a wide range of frequencies. The method was trained via simulation by using time-series measurements (i.e., voltage and current) for different system parameters and verified through several case studies. The obtained results show that: (1) extracting the time-dependent features of the voltage/current data improves the performance of the RF regression method; (2) the RF regression method is robust and allows grid impedance estimation within 1.5 grid cycles; (3) the proposed method can effectively estimate the grid impedance both in steady state and in case of large transients like electrical faults.
|
|
| 7. |
- de Dios, Eddie, et al.
(författare)
-
Introduction to Deep Learning in Clinical Neuroscience
- 2022
-
Ingår i: Acta Neurochirurgica, Supplement. - Cham : Springer International Publishing. - 2197-8395 .- 0065-1419. ; 134, s. 79-89
-
Bokkapitel (övrigt vetenskapligt/konstnärligt)abstract
- The use of deep learning (DL) is rapidly increasing in clinical neuroscience. The term denotes models with multiple sequential layers of learning algorithms, architecturally similar to neural networks of the brain. We provide examples of DL in analyzing MRI data and discuss potential applications and methodological caveats. Important aspects are data pre-processing, volumetric segmentation, and specific task-performing DL methods, such as CNNs and AEs. Additionally, GAN-expansion and domain mapping are useful DL techniques for generating artificial data and combining several smaller datasets. We present results of DL-based segmentation and accuracy in predicting glioma subtypes based on MRI features. Dice scores range from 0.77 to 0.89. In mixed glioma cohorts, IDH mutation can be predicted with a sensitivity of 0.98 and specificity of 0.97. Results in test cohorts have shown improvements of 5–7% in accuracy, following GAN-expansion of data and domain mapping of smaller datasets. The provided DL examples are promising, although not yet in clinical practice. DL has demonstrated usefulness in data augmentation and for overcoming data variability. DL methods should be further studied, developed, and validated for broader clinical use. Ultimately, DL models can serve as effective decision support systems, and are especially well-suited for time-consuming, detail-focused, and data-ample tasks.
|
|
| 8. |
- de Oliveira, Roger Alves, et al.
(författare)
-
Visualizing The Results From Unsupervised Deep Learning For The Analysis Of Power-Quality Data
- 2021
-
Ingår i: Cired 2021 - The 26Th International Conference And Exhibition On Electricity Distribution. - : Institution of Engineering and Technology. ; , s. 653-657
-
Konferensbidrag (refereegranskat)abstract
- This paper presents a visualisation method, based on deep learning, to assist power engineers in the analysis of large amounts of power-quality data. The method assists in extracting and understanding daily, weekly and seasonal variations in harmonic voltage. Measurements from 10 kV and 0.4 kV in a Swedish distribution network are applied to the deep learning method to obtain daily harmonic patterns and their distribution over the week and the year. The results are presented in graphs that allow interpretation of the results without having to understand the mathematical details of the method. The inferences given by the results demonstrate that the method can become a new tool that compresses power quality big data in a form that is easier to interpret.
|
|
| 9. |
- Domagk, Max, et al.
(författare)
-
Automatic Identification of Different Types of Consumer Configurations by Using Harmonic Current Measurements
- 2021
-
Ingår i: Applied Sciences (Switzerland). - : MDPI AG. - 2076-3417. ; 11:8
-
Tidskriftsartikel (refereegranskat)abstract
- Power quality (PQ) is an increasing concern in the distribution networks of modern industrialized countries. The PQ monitoring activities of distribution system operators (DSO), and consequently the amount of PQ measurement data, continuously increase, and consequently new and automated tools are required for efficient PQ analysis. Time characteristics of PQ parameters (e.g., harmonics) usually show characteristic daily and weekly cycles, mainly caused by the usage behaviour of electric devices. In this paper, methods are proposed for the classification of harmonic emission profiles for typical consumer configurations in public low voltage (LV) networks using a binary decision tree in combination with support vector machines. The performance of the classification was evaluated based on 40 different measurement sites in German public LV grids. This method can support network operators in the identification of consumer configurations and the early detection of fundamental changes in harmonic emission behaviour. This enables, for example, assistance in resolving customer complaints or supporting network planning by managing PQ levels using typical harmonic emission profiles.
|
|
| 10. |
- Fürst, Kristoffer, 1990, et al.
(författare)
-
Hierarchical LSTM-Based Classification of Household Heating Types Using Measurement Data
- 2024
-
Ingår i: IEEE Transactions on Smart Grid. - 1949-3053 .- 1949-3061. ; 15:2, s. 2261-2270
-
Tidskriftsartikel (refereegranskat)abstract
- A lack of knowledge of the heating systems used by electricity consumers impedes distribution system operators in developing a sound grid upgrade plan and estimating potential demand flexibility from these consumers. The large-scale rollout of smart meters for electricity consumers provides an excellent opportunity to identify end users’ heating types. This paper proposed a hierarchically structured deep-learning framework for identifying heating types of individual electricity consumers. The main contributions of the paper are: (a) We propose an effective framework based on long short-term memory (LSTM) that offers an effective automatic feature learning from sequential electricity consumption data and weather conditions. (b) We apply the proposed deep-learning architecture for household heating type classification which is among the first few successful reports on this application. We evaluate the performance using hourly measurement data collected over four years from one and two-family dwellings with either district heating, exhaust air heat pumps or direct electric heating as the heating type. Good performance was shown from the test results using the proposed framework, with an average test accuracy of 94.2%. Comparisons with four existing machine learning algorithms using handcrafted features and a single-layer LSTM-based deep-learning algorithm have shown marked improvement of the proposed method.
|
|
| 11. |
- Fürst, Kristoffer, 1990, et al.
(författare)
-
Improved Peak Load Estimation from Single and Multiple Consumer Categories
- 2020
-
Ingår i: IET Conference Publications. - : Institution of Engineering and Technology (IET). - 2515-0855. ; 2020:CP767, s. 178-181
-
Konferensbidrag (refereegranskat)abstract
- Velander’s formula and coincidence factors have traditionally been used to estimate peak load for new connections in the distribution grid. By re-evaluating their underlying assumptions, this paper proposes two improved models for aggregated peak load estimation (PLE). For single-category load aggregation, the proposed coincidence factor model, by incorporating an average correlation coefficient, improves the model fitting by 76%–96% as compared to the standard Rusck model. For multiple-category load aggregation, the proposed joint Gaussian regression model reduces the PLE bias from 3%–34% to 0.2%–3% compared to the traditional approach. (see the full paper in 4 pages)
|
|
| 12. |
- Fürst, Kristoffer, 1990, et al.
(författare)
-
Support vector machine for classification of households' heating type using load curves
- 2023
-
Ingår i: IET Conference Proceedings. - 2732-4494. ; 2023:6, s. 3884-3888
-
Konferensbidrag (refereegranskat)abstract
- The distribution system operator lacks the knowledge of the heating system used by their customers to make sound grid planning decisions. Energy declaration from buildings and the large-scale rollout of smart meters provides an excellent opportunity to classify the heating system used. This paper proposes a machine-learning-based approach using a support vector machine (SVM) with daily load curves (mean and standard deviation of consumption) extracted from smart meter measurements. Three heating types are analysed: district heating, exhaust air heat pump, and direct electric heating. The performance was compared among the classifiers using daily load curves extracted over one year, for each month, each week, and each day of the year. The highest average accuracy of 92.6% was obtained for the SVM classifier using daily load curves extracted for each week of a year as features. Furthermore, the classifier showed a higher performance than using an ensemble of SVM or random forest classifiers (90.6%/90.5%) proposed in the literature. Lastly, an error analysis of the misclassification was carried out, including building characteristics and geographical analysis.
|
|
| 13. |
- Ge, Chenjie, 1991, et al.
(författare)
-
Deep Feature Clustering for Seeking Patterns in Daily Harmonic Variations
- 2021
-
Ingår i: IEEE Transactions on Instrumentation and Measurement. - : IEEE. - 0018-9456 .- 1557-9662. ; 70
-
Tidskriftsartikel (refereegranskat)abstract
- This article proposes a novel scheme for analyzing power system measurement data. The main question that we seek answers in this study is on “whether one can find some important patterns that are hidden in the large data of power system measurements such as variational data.” The proposed scheme uses an unsupervised deep feature learning approach by first employing a deep autoencoder (DAE) followed by feature clustering. An analysis is performed by examining the patterns of clusters and reconstructing the representative data sequence for the clustering centers. The scheme is illustrated by applying it to the daily variations of harmonic voltage distortion in a low-voltage network. The main contributions of the article include: 1) providing a new unsupervised deep feature learning approach for seeking possible underlying patterns of power system variation measurements and 2) proposing an effective empirical analysis approach for understanding the measurements through examining the underlying feature clusters and the associated reconstructed data by DAE.
|
|
| 14. |
- Ge, Chenjie, 1991, et al.
(författare)
-
Deep semi-supervised learning for brain tumor classification
- 2020
-
Ingår i: BMC Medical Imaging. - : Springer Science and Business Media LLC. - 1471-2342. ; 20:1
-
Tidskriftsartikel (refereegranskat)abstract
- Background: This paper addresses issues of brain tumor, glioma, classification from four modalities of Magnetic Resonance Image (MRI) scans (i.e., T1 weighted MRI, T1 weighted MRI with contrast-enhanced, T2 weighted MRI and FLAIR). Currently, many available glioma datasets often contain some unlabeled brain scans, and many datasets are moderate in size. Methods: We propose to exploit deep semi-supervised learning to make full use of the unlabeled data. Deep CNN features were incorporated into a new graph-based semi-supervised learning framework for learning the labels of the unlabeled data, where a new 3D-2D consistent constraint is added to make consistent classifications for the 2D slices from the same 3D brain scan. A deep-learning classifier is then trained to classify different glioma types using both labeled and unlabeled data with estimated labels. To alleviate the overfitting caused by moderate-size datasets, synthetic MRIs generated by Generative Adversarial Networks (GANs) are added in the training of CNNs. Results: The proposed scheme has been tested on two glioma datasets, TCGA dataset for IDH-mutation prediction (molecular-based glioma subtype classification) and MICCAI dataset for glioma grading. Our results have shown good performance (with test accuracies 86.53% on TCGA dataset and 90.70% on MICCAI dataset). Conclusions: The proposed scheme is effective for glioma IDH-mutation prediction and glioma grading, and its performance is comparable to the state-of-the-art.
|
|
| 15. |
- Ge, Chenjie, 1991, et al.
(författare)
-
Enlarged Training Dataset by Pairwise GANs for Molecular-Based Brain Tumor Classification
- 2020
-
Ingår i: IEEE Access. - 2169-3536 .- 2169-3536. ; 8:1, s. 22560-22570
-
Tidskriftsartikel (refereegranskat)abstract
- This paper addresses issues of brain tumor subtype classification using Magnetic Resonance Images (MRIs) from different scanner modalities like T1 weighted, T1 weighted with contrast-enhanced, T2 weighted and FLAIR images. Currently most available glioma datasets are relatively moderate in size, and often accompanied with incomplete MRIs in different modalities. To tackle the commonly encountered problems of insufficiently large brain tumor datasets and incomplete modality of image for deep learning, we propose to add augmented brain MR images to enlarge the training dataset by employing a pairwise Generative Adversarial Network (GAN) model. The pairwise GAN is able to generate synthetic MRIs across different modalities. To achieve the patient-level diagnostic result, we propose a post-processing strategy to combine the slice-level glioma subtype classification results by majority voting. A two-stage course-to-fine training strategy is proposed to learn the glioma feature using GAN-augmented MRIs followed by real MRIs. To evaluate the effectiveness of the proposed scheme, experiments have been conducted on a brain tumor dataset for classifying glioma molecular subtypes: isocitrate dehydrogenase 1 (IDH1) mutation and IDH1 wild-type. Our results on the dataset have shown good performance (with test accuracy 88.82%). Comparisons with several state-of-the-art methods are also included.
|
|
| 16. |
- Ge, Chenjie, 1991, et al.
(författare)
-
Unsupervised deep learning and analysis of harmonic variation patterns using big data from multiple locations
- 2021
-
Ingår i: Electric power systems research. - : Elsevier. - 0378-7796 .- 1873-2046. ; 194
-
Tidskriftsartikel (refereegranskat)abstract
- This paper addresses the issue of automatically seeking and identifying daily, weekly and seasonal patterns in harmonic voltage from measurement data at multiple locations. We propose a novel framework that employs deep autoencoder (DAE) followed by k-mean clustering. The DAE is used for extracting principal features from time series of harmonic voltages. A new strategy is used for training the encoder in DAE from data at one selected location that is effective for subsequent feature extraction from data at multiple locations. To analyze the patterns, several empirical analysis approaches are applied on the clustered principal features, including the distribution of daily patterns over the week and the year, representative waveform sequences of individual classes, and feature maps for visualizing high-dimensional feature space through low-dimensional embedding. The proposed scheme has been tested on a dataset containing harmonic measurements at 10 low-voltage locations in Sweden for the whole year of 2017. Results show distinct principal patterns for most harmonics that can be related to the use of equipment causing harmonic distortion. This information can assist network operators in finding the origin of harmonic distortion and deciding about mitigation actions. The proposed scheme is the first to provide a useful analysis tool and insight for finding and analyzing underlying patterns from harmonic variation data at multiple locations.
|
|
