Research Director
Line Eikvil
- Department Image analysis, machine learning and Earth observation
- Mobile phone +47 416 52 218
- Phone number +47 22 85 26 88
- E-mail eikvil@nr.stage.dekodes.no
Projects
- Image analysis
- Machine learning
Breast cancer detection with machine learning (MIM)
- Image analysis
- Machine learning
Intelligent cardiac ultrasounds (INCUS)
- Image analysis
- Machine learning
Deep learning for seismic data (DELI)
Publications
- 226 publications found
Eikvil, Line og Løland, Anders. (2026).
Industrielle problemer trenger fortsatt prediktiv kunstig intelligens.
Vis sammendrag
Generativ kunstig intelligens er imponerende, men ikke alltid så nyttig til å løse industrielle problemer.
Jenssen, Robert; Eikvil, Line; Solberg, Anne H Schistad; Solheim, Inger og Bjørklund, Petter. (2026).
Visual Intelligence Annual Report 2025.
Utseth, Ingrid; Vedal, Amund Hansen; Thomas, Sarina og Eikvil, Line. (2026).
Comparing Foundation Models for Medical Images: A Study on Limited Data and Generalization.
Vis sammendrag
In this study we have investigated how vision foundation models, pretrained on different domains, compete with a specialized model for classification as a function of the size of the labeled training set of medical images. Furthermore, we have looked into the different models' ability to generalize to difficult cases. Our experiments are conducted for cardiac ultrasound images and the downstream task of view recognition. Still, this classification task is meant to serve as a demonstrative example, where we think that the findings should be transferable to other classification tasks and other domains. Through these experiments we found that the foundation models were able to beat the performance of our task-specific supervised model when labelled training data were limited. This was true even for models trained on natural images and when using the simple linear probing method to create a classifier. We observed that more domain-specific foundation models achieved an even higher performance with limited data. On the other hand, the more general models showed a greater ability to generalize and perform well on difficult, out-of-distribution cases. Still, for typical in-domain cases with sufficient labeled data, a task-specific ResNet model was competitive with the foundation models, while also being both smaller and faster.
Utseth, Ingrid; Vedal, Amund Hansen; Thomas, Sarina og Eikvil, Line. (2026).
Comparing Foundation Models for Medical Images: A Study on Limited Data and Generalization. Universitetet i Tromsø
NVA
poster
Vis sammendrag
In this study we have investigated how vision foundation models, pretrained on different domains, compete with a specialized model for classification as a function of the size of the labeled training set of medical images. Furthermore, we have looked into the different models' ability to generalize to difficult cases. Our experiments are conducted for cardiac ultrasound images and the downstream task of view recognition. Still, this classification task is meant to serve as a demonstrative example, where we think that the findings should be transferable to other classification tasks and other domains. Through these experiments we found that the foundation models were able to beat the performance of our task-specific supervised model when labelled training data were limited. This was true even for models trained on natural images and when using the simple linear probing method to create a classifier. We observed that more domain-specific foundation models achieved an even higher performance with limited data. On the other hand, the more general models showed a greater ability to generalize and perform well on difficult, out-of-distribution cases. Still, for typical in-domain cases with sufficient labeled data, a task-specific ResNet model was competitive with the foundation models, while also being both smaller and faster.
Jenssen, Robert; Eikvil, Line; Solberg, Anne H Schistad; Solheim, Inger og Bjørklund, Petter. (2025).
Visual Intelligence Annual Report 2024.
Vis sammendrag
The Visual Intelligence Annual Report 2024 highlights the centre's progress, activities and achieved innovations for 2022. It describes new deep learning methods which address pressing societal needs in the fields of medicine and health, marine science, the energy sector, and earth observation.
Ordoñez, Alba; Dahl, Fredrik Andreas; Brautaset, Olav og Eikvil, Line. (2025).
Unsupervised Domain Adaptation for Breast Cancer Detection in a Multi-Scanner Environment: A Case-Study from Norway.
Vis sammendrag
Maintaining the performance of a deep learning model trained for breast cancer detection on a specific scanner type is challenging in a multi-scanner setting due to domain shifts caused by variations in imaging data. This often results in a performance drop when models trained on one scanner are tested on another. While re-training with labeled data from the new scanner is an option, delays in obtaining ground-truth labels make this approach impractical. To overcome this limitation, Unsupervised Domain Adaptation (UDA) offers a promising alternative by enabling models to adapt across scanners without requiring labeled target data. In this study, we investigated Conditional Domain-Adversarial Network (CDAN), an adversarial UDA approach, to adapt a classifier trained on Siemens scanner data using nearly 3 million mammograms from the Norwegian breast cancer screening program. We compared it to Maximum Mean Discrepancy (MMD), a simpler statistical feature alignment method, and evaluated histogram matching, which required no additional training. Our findings showed that the AUC drop on the target GE data (0.96 to 0.62) without adaptation was mitigated by histogram matching (AUC 0.84), but that was less effective than MMD (AUC 0.87), which performed competitively with CDAN. Further ablation with Domain-Adversarial Neural Network (DANN), the foundation of CDAN, suggested limitations in the domain discriminator. Unlike prior work focusing solely on performance, we paired UDA with explainability. This revealed how feature relevance shifted across scanner domains, offering novel insights into model generalizability in cancer detection.
Dahl, Fredrik Andreas; Vedal, Amund; Eikvil, Line; Thrun, Solveig; Kampffmeyer, Michael og Hofvind, Solveig Sand-Hanssen. (2025).
Modelling Uncertainty in Graph Convolutional Networks for Edge Detection in Mammograms.
Vis sammendrag
Delineation of structures and estimation of landmarks in mammograms is a critical step in the evaluation of image quality in breast cancer screening, but requires the estimation of the uncertainty of the predicted landmarks to refer uncertain cases to clinicians. Of particular importance – and the focus of this work – is on the pectoral muscle, where the variability in muscle visibility across images introduces significant uncertainty. While graph convolutional networks (GCN) have been demonstrated to accurately predict landmarks by explicitly leveraging structural relationships between landmarks, they typically lack the ability to provide accurate uncertainty estimates for the landmarks. To address this shortcoming, in this work we propose a novel GCN-based approach that not only locates key points along the muscle boundary but also provides accurate uncertainty estimates, capturing both the aleatoric and epistemic uncertainties. Our method was evaluated on in-house annotated mammograms demonstrating comparable accuracy to human annotators, while at the same time providing highly accurate uncertainty estimates, confirming its potential for identifying cases that require human review. We further validate our proposed approach on the publicly available CSAW-S and INBreast datasets, demonstrating its robustness to domain shift, as well as its potential to detect incorrect or untypical annotations.
Utseth, Ingrid; Sarmad, Muhammad; Eikvil, Line; Salberg, Arnt Børre; Ordonez, Alba og Brautaset, Olav. (2025).
Leveraging Data with Strong, Weak and No Labels in Marine Acoustics. SFI Visual Intelligence
NVA
poster
Utseth, Ingrid; Ordonez, Alba; Sarmad, Muhammad; Salberg, Arnt Børre; Eikvil, Line; Brautaset, Olav og Handegard, Nils Olav. (2025).
Deep learning for marine acoustics: Leveraging data with strong, weak and no labels. SFI Visual Intelligence
NVA
Faglig foredrag
Utseth, Ingrid; Brautaset, Olav; Eikvil, Line; Ordonez, Alba; Salberg, Arnt Børre; Sarmad, Muhammad; Holmin, Arne Johannes; Pala, Ahmet og Handegard, Nils Olav. (2025).
Deep learning methods for acoustic target classification. Iceland’s Marine and Freshwater Research Institute
NVA
Annen presentasjon
Martiniussen, Marit Almenning; Larsen, Marthe; Hovda, Tone; SANDHAUG, MERETE; Dahl, Fredrik Andreas; Eikvil, Line; Brautaset, Olav; Bjørnerud, Atle; Kristensen, Vessela N.; Bergan, Marie Burns og Hofvind, Solveig Sand-Hanssen. (2025).
Performance of Two Deep Learning–based AI Models for Breast Cancer Detection and Localization on Screening Mammograms from BreastScreen Norway.
Vis sammendrag
Two deep learning–based artificial intelligence (AI) models, one commercially available and one in-house, showed good performance for stand-alone cancer detection on retrospective mammography screening data. AI markings on the mammograms corresponded well to the true cancer location. Purpose - To evaluate cancer detection and marker placement accuracy of two artificial intelligence (AI) models developed for interpretation of screening mammograms. Materials and Methods - This retrospective study included data from 129 434 screening examinations (all female patients; mean age, 59.2 years ± 5.8 [SD]) performed between January 2008 and December 2018 in BreastScreen Norway. Model A was commercially available and model B was an in-house model. Area under the receiver operating characteristic curve (AUC) with 95% CIs were calculated. The study defined 3.2% and 11.1% of the examinations with the highest AI scores as positive, threshold 1 and 2, respectively. A radiologic review assessed location of AI markings and classified interval cancers as true or false negative. Results - The AUC value was 0.93 (95% CI: 0.92, 0.94) for model A and B when including screen-detected and interval cancers. Model A identified 82.5% (611 of 741) of the screen-detected cancers at threshold 1 and 92.4% (685 of 741) at threshold 2. Model B identified 81.8% (606 of 741) at threshold 1 and 93.7% (694 of 741) at threshold 2. The AI markings were correctly localized for all screen-detected cancers identified by both models and 82% (56 of 68) of the interval cancers for model A and 79% (54 of 68) for model B. At the review, 21.6% (45 of 208) of the interval cancers were identified at the preceding screening by either or both models, correctly localized and classified as false negative (n = 17) or with minimal signs of malignancy (n = 28). Conclusion - Both AI models showed promising performance for cancer detection on screening mammograms. The AI markings corresponded well to the true cancer locations.
Ordonez, Alba; Dahl, Fredrik Andreas; Brautaset, Olav og Eikvil, Line. (2025).
Unsupervised domain adaptation for breast cancer detection in a multi-scanner environment: A case-study from Norway. AIME2025
NVA
Vitenskapelig foredrag
Jenssen, Robert; Eikvil, Line; Solberg, Anne H Schistad; Solheim, Inger og Bjørklund, Petter. (2024).
Visual Intelligence Annual Report 2023.
Utseth, Ingrid; Vedal, Amund Hansen; Eikvil, Line og Waldeland, Anders U.. (2024).
Experiments with Foundation Models for Cardiac Ultrasound Images in Limited Data Scenarios. SFI Visual Intelligence
NVA
Faglig foredrag
Utseth, Ingrid; Vedal, Amund Hansen; Eikvil, Line og Waldeland, Anders U.. (2024).
Experiments with Foundation Models for
Cardiac Ultrasound Images in Limited Data
Scenarios. SFI Visual Intelligence
NVA
poster
Utseth, Ingrid; Vedal, Amund Hansen; Eikvil, Line og Waldeland, Anders U.. (2024).
Experiments with foundation models for cardiac ultrasound images.
NVA
Rapport
Utseth, Ingrid; Vedal, Amund Hansen og Eikvil, Line. (2024).
Detection of timing events in cardiac ultrasound.
NVA
Rapport
Løland, Anders og Eikvil, Line. (2024).
«In the 90s, everybody knew that AI didn't work.» Hvorfor hadde Sergey Brin rett da han sa det?
NVA
Programdeltagelse
Dahl, Fredrik Andreas; Brautaset, Olav; Holden, Marit; Eikvil, Line; Larsen, Marthe; Martiniussen, Marit Almenning og Hofvind, Solveig Sand-Hanssen. (2024).
En to-trinns kunstig intelligens modell for deteksjon av brystkreft i mammogrammer. Norsk radiologisk forening
NVA
Faglig foredrag
Jenssen, Robert; Eikvil, Line; Solberg, Anne H Schistad og Solheim, Inger. (2023).
Visual Intelligence Annual Report 2022.
Dahl, Fredrik Andreas; Eikvil, Line og Vedal, Amund Hansen. (2023).
Training GCN on breast positioning using partially annotated mammograms. Visual Intelligence (SFI)
NVA
poster
Dahl, Fredrik Andreas; Holden, Marit; Brautaset, Olav og Eikvil, Line. (2023).
Utilizing earlier images in mammography cancer detection. Visual Intelligence (SFI)
NVA
poster
Dahl, Fredrik Andreas; Eikvil, Line og Vedal, Amund Hansen. (2023).
Training GCN on breast positioning using partially annotated mammograms. Visual Intelligence (SFI)
NVA
Vitenskapelig foredrag
Brautaset, Olav; Utseth, Ingrid; Eikvil, Line; Salberg, Arnt-Børre og Handegard, Nils Olav. (2023).
Learning from weakly labelled marine acoustic data. Visual Intelligence
NVA
poster
Brautaset, Olav; Utseth, Ingrid; Eikvil, Line; Salberg, Arnt-Børre og Handegard, Nils Olav. (2023).
Learning from weakly labelled marine acoustic data. Visual Intelligence
NVA
Vitenskapelig foredrag
Vedal, Amund Hansen og Eikvil, Line. (2023).
Context-Aware Landmark Detection for 2D Cardiac Ultrasound using Graph Convnet.
NVA
Rapport
Dahl, Fredrik Andreas; Brautaset, Olav; Holden, Marit; Eikvil, Line; Larsen, Marthe og Hofvind, Solveig Sand-Hanssen. (2023).
A two-stage mammography classification model using explainable-AI for ROI detection.
Vis sammendrag
This study introduces an enhanced version of a two-stage modelling approach using artificial intelligence (AI) for breast cancer detection in mammography screening. Leveraging a large dataset of 2,863,175 mammograms from the BreastScreen Norway, the approach uses two convolutional neural networks. The first one is trained to classify whole images, and an explainable-AI method is applied to this network to identify a region of interest (ROI). The second neural network subsequently classifies the ROI for malignancy. While a prior method used simple gradient saliency maps to identify ROIs, a key enhancement of the present methodology is the application of Layered GradCam, which identifies cancerous areas more consistently and allows smaller ROIs. Layered GradCam is also used to display identified cancers to the user. By the AUC criterion, our model performs well, 0.974 for screen-detected and 0.931 for all cancers (screen-detected and interval), compared to a commercial program; 0.959 and 0.918, respectively. Comparisons with the radiologist scores indicate that the model has equal performance with two radiologists, and superior performance to one, for the detection of all cancers (screening- and interval type). Our tests indicate that our model generalizes well for different breast centers, but so far only images from a single manufacturer have been tested.
Dahl, Fredrik Andreas; Eikvil, Line; Tvete, Ingunn Fride; Lison, Pierre; Pilán, Ildikó; Fuglerud, Kristin Skeide og Leister, Wolfgang. (2023).
Helse-effektivisering - et mulig satsningsområde for NR.
NVA
Rapport
Salberg, Arnt-Børre og Eikvil, Line. (2023).
Out-of-distribution detection in deep neural networks applied to marine data. Visual Intelligence
NVA
poster
Jenssen, Robert; Eikvil, Line; Solberg, Anne H Schistad og Solheim, Inger. (2022).
Visual Intelligence Annual Report 2021.
Vis sammendrag
The Visual Intelligence Annual Report 2021 highlights the centre's progress, activities and achieved innovations for 2021. It describes new deep learning methods which address pressing societal needs in the fields of medicine and health, marine science, the energy sector, and earth observation.
Dahl, Fredrik Andreas; Vedal, Amund Hansen og Eikvil, Line. (2022).
Analyzing breast positioning in mammograms with graph convolutional networks. Visual Intelligence (SFI)
NVA
Vitenskapelig foredrag
Dahl, Fredrik Andreas; Holden, Marit; Brautaset, Olav og Eikvil, Line. (2022).
A mammography classification model trained from image labels only. Visual Intelligence center for research-based innovation and Uinversity of Tromsø
NVA
poster
Petterson, Jarle; Eikvil, Line; Fuglerud, Kristin Skeide; Fredrik, Dahl og Halbach, Till. (2022).
Kunstig intelligens avlaster travle spesialister.
NVA
Intervju
Eikvil, Line; Waldeland, Anders U.; Barker, Daniel Martin L; Holden, Marit; Hauge, Ragnar og Salberg, Arnt Børre. (2022).
Deep learning in seismic interpretation,
Development and experiments 2021-2022.
NVA
Rapport
Ordonez, Alba; Eikvil, Line; Salberg, Arnt-Børre; Harbitz, Alf og Elvarsson, Bjarki Thor. (2022).
Automatic Fish Age Determination across Different Otolith Image Labs Using Domain Adaptation.
Vis sammendrag
The age determination of fish is fundamental to marine resource management. This task is commonly done by analysis of otoliths performed manually by human experts. Otolith images from Greenland halibut acquired by the Institute of Marine Research (Norway) were recently used to train a convolutional neural network (CNN) for automatically predicting fish age, opening the way for requiring less human effort and availability of expertise by means of deep learning (DL). In this study, we demonstrate that applying a CNN model trained on images from one lab (in Norway) does not lead to a suitable performance when predicting fish ages from otolith images from another lab (in Iceland) for the same species. This is due to a problem known as dataset shift, where the source data, i.e., the dataset the model was trained on have different characteristics from the dataset at test stage, here denoted as target data. We further demonstrate that we can handle this problem by using domain adaptation, such that an existing model trained in the source domain is adapted to perform well in the target domain, without requiring extra annotation effort. We investigate four different approaches: (i) simple adaptation via image standardization, (ii) adversarial generative adaptation, (iii) adversarial discriminative adaptation and (iv) self-supervised adaptation. The results show that the performance varies substantially between the methods, with adversarial discriminative and self-supervised adaptations being the best approaches. Without using a domain adaptation approach, the root mean squared error (RMSE) and coefficient of variation (CV) on the Icelandic dataset are as high as 5.12 years and 28.6%, respectively, whereas by using the self-supervised domain adaptation, the RMSE and CV are reduced to 1.94 years and 11.1%. We conclude that careful consideration must be given before DL-based predictors are applied to perform large scale inference. Despite that, domain adaptation is a promising solution for handling problems of dataset shift across image labs.
Thorarinsdottir, Thordis Linda; Roksvåg, Thea Julie Thømt og Eikvil, Line. (2022).
Birdcam: Automatic monitoring of birds near wind parks.
NVA
Rapport
Vedal, Amund Hansen; Eikvil, Line; Holden, Marit og Gilbert, Andrew. (2022).
Context-Aware Landmark Detection for 2D Cardiac Ultrasound using Graph Convnet. GE Vingmed Ultrasound
NVA
Vitenskapelig foredrag
Utseth, Ingrid; Ordonez, Alba; Eikvil, Line; Brautaset, Olav; Salberg, Arnt Børre og Handegard, Nils Olav. (2022).
Improving marine acoustic target classification with context information. International Council for the Exploration of the Sea
NVA
poster
Dahl, Fredrik Andreas; Holden, Marit; Brautaset, Olav og Eikvil, Line. (2022).
A mammography classification model trained from image labels only.
Vis sammendrag
The Cancer Registry of Norway organises a population-based breast cancer screening program, where 250 000 women participate each year. The interpretation of the screening mammograms is a manual process, but deep neural networks are showing potential in mammographic screening. Most methods focus on methods trained from pixel-level annotations, but these require expertise and are time-consuming to produce. Through the screenings, image level annotations are however readily available. In this work we present a few models trained from image level annotations from the Norwegian dataset: a holistic model, an attention model and an ensemble model. We compared their performance with that of pretrained models based on pixel-level annotations, trained on international datasets. From this we found that models trained on our local data with image-level annotation gave considerably better performance than the pretrained models from external data, although based on pixel-level annotations.
Dahl, Fredrik Andreas; Holden, Marit; Brautaset, Olav og Eikvil, Line. (2022).
A two-stage mammography classification model using XAI for ROI detection. Visual Intelligence (SFI)
NVA
poster
Eikvil, Line. (2022).
Slik bruker vi norske mammogrammer for å utvikle et maskinlæringssystem. Kreftregisteret
NVA
Faglig foredrag
Vedal, Amund Hansen; Eikvil, Line og Holden, Marit. (2022).
Context-Aware Landmark Detection for 2D Cardiac Ultrasound using a Graph Convolution Network.
NVA
poster
Ordonez, Alba; Vedal, Amund Hansen; Dahl, Fredrik Andreas og Eikvil, Line. (2022).
Explainability and uncertainty for classification of mammograms. Visual Intelligence centre for research-based innovation and university of Tromsø
NVA
poster
Dahl, Fredrik Andreas; Holden, Marit; Brautaset, Olav og Eikvil, Line. (2022).
Machine learning for screening mammography.
NVA
Rapport
Eikvil, Line; Dahl, Fredrik Andreas; Holden, Marit; Brautaset, Olav; Hofvind, Solveig Sand-Hanssen; Aglen, Camilla Flåt og Larsen, Marthe. (2022).
Procedures for mammographic screening with machine learning.
NVA
Rapport
Jenssen, Robert; Eikvil, Line; Solberg, Anne H Schistad og Solheim, Inger. (2021).
Visual Intelligence Annual Report 2020.
Vis sammendrag
The Visual Intelligence Annual Report 2020 highlights the centre's progress, activities and achieved innovations for 2020. It describes new deep learning methods which address pressing societal needs in the fields of medicine and health, marine science, the energy sector, and earth observation.
Choi, Changkyu; Kampffmeyer, Michael; Handegard, Nils Olav; Salberg, Arnt Børre; Brautaset, Olav; Eikvil, Line og Jenssen, Robert. (2021).
Semi-supervised target classification in multi-frequency echosounder data.
Vis sammendrag
Acoustic target classification in multi-frequency echosounder data is a major interest for the marine ecosystem and fishery management since it can potentially estimate the abundance or biomass of the species. A key problem of current methods is the heavy dependence on the manual categorization of data samples. As a solution, we propose a novel semi-supervised deep learning method leveraging a few annotated data samples together with vast amounts of unannotated data samples, all in a single model. Specifically, two inter-connected objectives, namely, a clustering objective and a classification objective, optimize one shared convolutional neural network in an alternating manner. The clustering objective exploits the underlying structure of all data, both annotated and unannotated; the classification objective enforces a certain consistency to given classes using the few annotated data samples. We evaluate our classification method using echosounder data from the sandeel case study in the North Sea. In the semi-supervised setting with only a tenth of the training data annotated, our method achieves 67.6% accuracy, outperforming a conventional semi-supervised method by 7.0 percentage points. When applying the proposed method in a fully supervised setup, we achieve 74.7% accuracy, surpassing the standard supervised deep learning method by 4.7 percentage points.
Jenssen, Robert; Solberg, Anne H Schistad og Eikvil, Line. (2021).
Norge er verdensledende
på bildeanalyse med
kunstig intelligens.
NVA
Intervju
Eikvil, Line. (2021).
Deep Learning for Norwegian Breast Cancer Screening. Computational Radiology & Artificial Intelligence
NVA
Faglig foredrag
Ordonez, Alba; Utseth, Ingrid; Eikvil, Line og Handegard, Nils Olav. (2021).
Using model averaging ensembles in semantic segmentation of marine echosounder data for acoustic classification of species. Norsk Forening for Bildebehandling og Maskinlæring
NVA
Vitenskapelig foredrag
Ordonez, Alba; Eikvil, Line og Holden, Marit. (2021).
Learning motion of seismic structures without human labelling. Visual Intelligence centre for research-based innovation
NVA
Vitenskapelig foredrag
Vis sammendrag
Recent research published by Li and Abubakar (2020) investigated if the sediment interpretation could be done based on an idea inspired from video analysis, without involving human labelling. The aim of the study was understand how the proposed approach worked and investigate whether we could use it for detecting discontinuities in seismic structures.
Utseth, Ingrid; Ordonez, Alba; Eikvil, Line; Brautaset, Olav; Salberg, Arnt-Børre og Handegard, Nils Olav. (2021).
Improving marine acoustic target classification with context information. Visual Intelligence centre for research-based innovation
NVA
poster
Vis sammendrag
The aim of the study was to investigate whether context information related to acoustic data resolution could improve automatic acoustic classification of species using echosounder observations.
Handegard, Nils Olav; Eikvil, Line; Jenssen, Robert; Kampffmeyer, Michael; Salberg, Arnt Børre og Malde, Ketil. (2021).
Machine Learning + Marine Science: Critical Role of Partnerships in Norway.
Vis sammendrag
In this essay, we review some recent advances in developing machine learning (ML) methods for marine science applications in Norway. We focus mostly on deep learning (DL) methods and review the challenges we have faced in the process, including data preparation, (lack of) labelled training data, and interpretability. We also present the partnerships that have been formed between e-science institutions and marine science
institutions in Norway. These partnerships have been instrumental in moving this effort forward and have been fuelled by grants from the Norwegian Research Council. The last addition to this collaboration is the recent centres for research-based innovation in Marine Acoustic Abundance Estimation and Backscatter Classification (CRIMAC) and
Visual Intelligence (VI).
Ordonez, Alba; Harbitz, Alf; Elvarsson, Bjarki; Eikvil, Line og Salberg, Arnt-Børre. (2021).
Deep domain adaptation applied to automatic fish age prediction. Visual Intelligence centre for research-based innovation
NVA
Vitenskapelig foredrag
Vis sammendrag
During the presentation, we investigated how a model trained for automatic fish age prediction on otolith images from one lab could generalize to novel images from another lab. We identified Unsupervised Domain Adaptation based on coGAN (Liu and Tuzel, 2016) and Generate to Adapt (Sankaranarayanan et al., 2018) frameworks as promising methods.
Eikvil, Line; Holden, Marit og Ordonez, Alba. (2021).
Machine learning for image-based interpretation of non-verbal communication - Initial version.
NVA
Rapport
Eikvil, Line; Holden, Marit; Holden, Lars og Brautaset, Olav. (2021).
Machine learning for screening mammography.
NVA
Rapport
Handegard, Nils Olav; Algrøy, Tonny; Eikvil, Line; Hammersland, Hege; Tenningen, Maria og Ona, Egil. (2021).
Smart Fisheries in Norway: Partnership between Science, Technology, and the Fishing Sector.
NVA
Fagartikkel
Vis sammendrag
The Research Council of Norway recently
funded the Centre for Research-based
Innovation in Marine Acoustic Abundance
Estimation and Backscatter Classification
(CRIMAC). The centre has smart fisheries as
one of its main pillars, and will address reliable
catch quantification, sizing, and species
identification of individual fish. It builds on a
long tradition of science-industry partnership
in Norway. In this essay, we review how this
science-industry partnership has formed
through long standing collaborations between
marine science organizations, innovative
fishing companies, and the marine-tech
industry. The results from this collaboration
have led to instruments and methods being
used worldwide for smart fishing operations
and monitoring the marine environment
Choi, Changkyu; Kampffmeyer, Michael; Handegard, Nils Olav; Salberg, Arnt Børre; Eikvil, Line og Jenssen, Robert. (2021).
Semi-supervised Semantic Segmentation in Multi-frequency Echosounder Data. Robert Jenssen, Inger Solheim
NVA
poster
Eikvil, Line; Holden, Marit og Ordonez, Alba. (2021).
Machine learning for image-based interpretation of non-verbal communication.
NVA
Rapport
Gilbert, Andrew David; Holden, Marit; Eikvil, Line; Rakhmail, Mariia; Babic, Aleksandar; Aase, Svein Arne; Samset, Eigil og Mcleod, Kristin. (2020).
User-Intended Doppler Measurement Type Prediction Combining CNNs With Smart Post-Processing.
Vis sammendrag
Spectral Doppler measurements are an important part of the standard echocardiographic examination. These measurements give insight into myocardial motion and blood flow providing clinicians with parameters for diagnostic decision making. Many of these measurements are performed automatically with high accuracy, increasing the efficiency of the diagnostic pipeline. However, full automation is not yet available because the user must manually select which measurement should be performed on each image. In this work, we develop a pipeline based on convolutional neural networks (CNNs) to automatically classify the measurement type from cardiac Doppler scans. We show how the multi-modal information in each spectral Doppler recording can be combined using a meta parameter post-processing mapping scheme and heatmaps to encode coordinate locations. Additionally, we experiment with several architectures to examine the tradeoff between accuracy, speed, and memory usage for resource-constrained environments. Finally, we propose a confidence metric using the values in the last fully connected layer of the network and show that our confidence metric can prevent many misclassifications. Our algorithm enables a fully automatic pipeline from acquisition to Doppler spectrum measurements. We achieve 96% accuracy on a test set drawn from separate clinical sites, indicating that the proposed method is suitable for clinical adoption.
Eikvil, Line; Ordonez, Alba og Brautaset, Olav. (2020).
Feasibility study: Non-verbal communication interpreter - Image-based methodological approaches.
NVA
Rapport
Brautaset, Olav; Waldeland, Anders Ueland; Johnsen, Espen; Malde, Ketil; Eikvil, Line; Salberg, Arnt-Børre og Handegard, Nils Olav. (2020).
Acoustic classification in multifrequency echosounder data using deep convolutional neural networks.
Vis sammendrag
Acoustic target classification is the process of assigning observed acoustic backscattering intensity to an acoustic category. A deep learning strategy for acoustic target classification using a convolutional network is developed, consisting of an encoder and a decoder, which allow the network to use pixel information and more abstract features. The network can learn features directly from data, and the learned feature space may include both frequency response and school morphology. We tested the method on multifrequency data collected between 2007 and 2018 during the Norwegian sandeel survey. The network was able to distinguish between sandeel schools, schools of other species, and background pixels (including seabed) in new survey data with an F1 score of 0.87 when tested against manually labelled schools. The network separated schools of sandeel and schools of other species with an F1 score of 0.94. A traditional school classification algorithm obtained substantially lower F1 scores (0.77 and 0.82) when tested against the manually labelled schools. To train the network, it was necessary to develop sampling and preprocessing strategies to account for unbalanced classes, inaccurate annotations, and biases in the training data. This is a step towards a method to be applied across a range of acoustic trawl surveys.
Ordonez, Alba; Eikvil, Line; Salberg, Arnt-Børre; Harbitz, Alf; Murray, Sean Meling og Kampffmeyer, Michael. (2020).
Explaining decisions of deep neural networks used for fish age prediction.
Vis sammendrag
Age-reading of fish otoliths (ear stones) is important for the sustainable management of fish resources. However, the procedure is challenging and requires experienced readers to carefully examine annual growth zones. In a recent study, convolutional neural networks (CNNs) have been demonstrated to perform reasonably well on automatically predicting fish age from otolith images. In the present study, we carefully investigate the prediction rule learned by such neural networks to provide insight into the features that identify certain fish age ranges. For this purpose, a recent technique for visualizing and analyzing the predictions of the neural networks was applied to different versions of the otolith images. The results indicate that supplementary knowledge about the internal structure improves the results for the youngest age groups, compared to using only the contour shape attribute of the otolith. However, the contour shape and size attributes are, in general, sufficient for older age groups. In addition, within specific age ranges we find that the network tends to focus on particular areas of the otoliths and that the most discriminating factors seem to be related to the central part and the outer edge of the otolith. Explaining age predictions from otolith images as done in this study will hopefully help build confidence in the potential of deep learning algorithms for automatic age prediction, as well as improve the quality of the age estimation.
Malde, Ketil; Handegard, Nils Olav; Eikvil, Line og Salberg, Arnt Børre. (2020).
Machine intelligence and the data-driven future of marine science.
Vis sammendrag
Oceans constitute over 70% of the earth’s surface, and the marine environment and ecosystems are central to many global challenges. Not only are the oceans an important source of food and other resources, but they also play a important roles in the earth’s climate and provide crucial ecosystem services. To monitor the environment and ensure sustainable exploitation of marine resources, extensive data collection and analysis efforts form the backbone of management programmes on global, regional, or national levels. Technological advances in sensor technology, autonomous platforms, and information and communications technology now allow marine scientists to collect data in larger volumes than ever before. But our capacity for data analysis has not progressed comparably, and the growing discrepancy is becoming a major bottleneck for effective use of the available data, as well as an obstacle to scaling up data collection further. Recent years have seen rapid advances in the fields of artificial intelligence and machine learning, and in particular, so-called deep learning systems are now able to solve complex tasks that previously required human expertise. This technology is directly applicable to many important data analysis problems and it will provide tools that are needed to solve many complex challenges in marine science and resource management. Here we give a brief review of recent developments in deep learning, and highlight the many opportunities and challenges for effective adoption of this technology across the marine sciences.
Ordonez, Alba; Eikvil, Line; Salberg, Arnt Børre og Harbitz, Alf. (2020).
Understanding a Neural Network by Visualization: Application to Fish Age Prediction. Universitetet i Trømsø
NVA
poster
Eikvil, Line; Waldeland, Anders U.; Barker, Daniel Martin L; Holden, Marit; Hauge, Ragnar og Salberg, Arnt Børre. (2020).
Deep learning in seismic interpretations - Development and experiments 2020.
NVA
Rapport
Eikvil, Line. (2020).
Google har utviklet kunstig intelligens som gjenkjenner brystkreft bedre enn radiologer - Kreftregisteret og Norsk Regnesentral lager norsk variant.
NVA
Intervju
Eikvil, Line; Holden, Marit; Holden, Lars og Brautaset, Olav. (2020).
Machine learning for screening mammography - Initial analyses on a first limited dataset.
NVA
Rapport
Ordonez, Alba; Eikvil, Line og Salberg, Arnt Børre. (2019).
Visualization of deep neural networks applied to fish age prediction.
NVA
Rapport
Salberg, Arnt Børre; Eikvil, Line; Malde, Ketil og Handegard, Nils Olav. (2019).
The COGMAR project. Institute of Marine Research
NVA
Vitenskapelig foredrag
Murray, Sean Meling; Eikvil, Line og Salberg, Arnt Børre. (2019).
Automatic interpretation of otoliths with deep learning - explaining predictions.
NVA
Rapport
Eikvil, Line og Holden, Marit. (2019).
Deep learning for ultrasound images - next steps.
NVA
Rapport
Eikvil, Line; Waldeland, Anders U.; Holden, Marit; Salberg, Arnt Børre; Hauge, Ragnar og Barker, Daniel Martin L. (2019).
Deep learning in seismic interpretation.
NVA
Rapport
Gilbert, Andrew David; Holden, Marit; Eikvil, Line; Aase, Svein Arne; Samset, Eigil og Mcleod, Kristin. (2019).
Automated Left Ventricle Dimension Measurement in 2D Cardiac Ultrasound via an Anatomically Meaningful CNN Approach.
Vis sammendrag
Two-dimensional echocardiography (2DE) measurements of left ventricle (LV) dimensions are highly significant markers of several cardiovascular diseases. These measurements are often used in clinical care despite suffering from large variability between observers. This variability is due to the challenging nature of accurately finding the correct temporal and spatial location of measurement endpoints in ultrasound images. These images often contain fuzzy boundaries and varying reflection patterns between frames. In this work, we present a convolutional neural network (CNN) based approach to automate 2DE LV measurements. Treating the problem as a landmark detection problem, we propose a modified U-Net CNN architecture to generate heatmaps of likely coordinate locations. To improve the network performance we use anatomically meaningful heatmaps as labels and train with a multi-component loss function. Our network achieves 13.4%, 6%, and 10.8% mean percent error on intraventricular septum (IVS), LV internal dimension (LVID), and LV posterior wall (LVPW) measurements respectively. The design outperforms other networks and matches or approaches intra-analyser expert error.
Salberg, Arnt Børre og Eikvil, Line. (2019).
The COGMAR project: Ubiquitous cognitive computer vision for marine services. Capgemini
NVA
Faglig foredrag
Eikvil, Line; Holden, Marit; Hauge, Ragnar og Kvernelv, Vegard Berg. (2019).
Estimation of rock cuttings size from images.
NVA
Rapport
Eikvil, Line. (2018).
Intelligent og automatisk bildeanalyse med deep learning. OsloMet
NVA
Faglig foredrag
Eikvil, Line og Holden, Lars. (2018).
Bruk av Deep learning and Big Data i Mammografiprogrammet. Mammografiprogrammet
NVA
Faglig foredrag
Salberg, Arnt Børre og Eikvil, Line. (2016).
Deep learning: machine learning meets big data.
NVA
Vitenskapelig foredrag
Salberg, Arnt Børre og Eikvil, Line. (2016).
Deep learning: a powerful tool to solve your computer vision problems.
NVA
Vitenskapelig foredrag
Haugen, Marion; Eikvil, Line; Tvete, Ingunn Fride og Kvaal, Sigrid Ingeborg. (2016).
Development of improved methods or basis for medical age assessments of minors and young adults.
Vis sammendrag
This final report describes the results of the research conducted in the project «Development of improved methods or basis for medical age assessments of minors and young adults». The aim of the project has been to contribute to develop improved methods for measuring physical developments of a child or young adult and for assessing the chronological age based on such measurements, giving an improved basis for such assessments. The project has brought together various international experts, both from the medical and odontology fields. The project has been an international collaboration with a group of partners with experience from both practical age estimation and research, from both of the currently used radiographic methods in Norway (wrist and teeth) and new MRI approaches. Several disciplines have been incorporated in the project: odontology, pediatrics, radiography, statistical modeling and image analysis. Additional descriptions of methods and results from the statistical analyses and image analysis can be found in four additional notes. The project is funded by The Norwegian Directorate of Immigration.
Eikvil, Line; Jenssen, Tor-Kristian og Holden, Marit. (2015).
Multi-focus cluster labeling.
Vis sammendrag
Document collections resulting from searches in the biomedical literature, for instance, in PubMed, are often so large that some organization of the returned information is necessary. Clustering is an efficient tool for organizing search results. To help the user to decide how to continue the search for relevant documents, the content of each cluster can be characterized by a set of representative keywords or cluster labels. As different users may have different interests, it can be desirable with solutions that make it possible to produce labels from a selection of different topical categories. We therefore introduce the concept of multi-focus cluster labeling to give users the possibility to get an overview of the contents through labels from multiple viewpoints.
The concept for multi-focus cluster labeling has been established and has been demonstrated on three different document collections. We illustrate that multi-focus visualizations can give an overview of clusters along axes that general labels are not able to convey. The approach is generic and should be applicable to any biomedical (or other) domain with any selection of foci where appropriate focus vocabularies can be established. A user evaluation also indicates that such a multi-focus concept is useful.
Eikvil, Line og Holden, Marit. (2015).
Utvalg av deskriptorer for oppbygging av database.
NVA
Rapport
Eikvil, Line og Holden, Marit. (2015).
Bruk av fargedeskriptorer for gjenkjenning av ukeblader.
NVA
Rapport
Eikvil, Line og Holden, Marit. (2014).
Evaluation of Binary Descriptors for Fast and Fully Automatic Identification.
Vis sammendrag
In this study we evaluate the potential of local binary descriptors for automatic sorting in an industrial context. This problem is different from that of retrieval for human handling as we need to identify the one correct class, rather than finding all the similar classes. We have looked at classes of objects that need to be identified by their cover or label, rather than their shape. Challenges for this application are that the process needs to be very fast and the approach must be able to distinguish between a large number of classes, where the classes can be quite similar and have identical elements. We have studied various combinations of detectors and binary descriptors in combination with approximate nearest neighbor (ANN) searches in such contexts. Our conclusion is that these approaches are well suited for this type of automatic sorting, and our experiments show that for the best performing combinations we are able to obtain a 99% recognition rate on a database of 80,000 images using an average of less than 0.5 seconds per image.
Eikvil, Line og Holden, Marit. (2014).
Evaluation of binary descriptors for fast and fully automatic identification. IAPR
NVA
Vitenskapelig foredrag
Holden, Lars; Eikvil, Line; Holden, Marit og Boudko, Svetlana. (2014).
Historisk informasjon fra avisarkiv.
NVA
Rapport
Eikvil, Line og Salberg, Arnt Børre. (2013).
Signal and image processing for ultrasound camera.
NVA
Rapport