| 1. |
- Espinoza, Fredrik, et al.
(författare)
-
Analysis of Open Answers to Survey Questions throughInteractive Clustering and Theme Extraction
- 2018
-
Ingår i: Proceedings of Conference on Human Information Interaction & Retrieval. - New York, New York, USA : ACM Digital Library. ; , s. 317-320
-
Konferensbidrag (refereegranskat)abstract
- This paper describes design principles for and the implementation of Gavagai Explorer—a new application which builds on interactive text clustering to extract themes from topically coherent text sets such as open text answers to surveys or questionnaires.An automated system is quick, consistent, and has full coverage over the study material. A system allows an analyst to analyze more answers in a given time period; provides the same initial results regardless of who does the analysis, reducing the risks of inter-rater discrepancy; and does not risk miss responses due to fatigue or boredom. These factors reduce the cost and increase the reliability of the service. The most important feature, however, is relieving the human analyst from the frustrating aspects of the coding task, freeing the effort to the central challenge of understanding themes. Gavagai Explorer is available on-line at http://explorer.gavagai.se
|
|
| 2. |
- Sahlgren, Magnus, et al.
(författare)
-
The Gavagai Living Lexicon
- 2016
-
Ingår i: 10th edition of the Language Resources and Evaluation Conference, 23-28 May 2016, Portorož (Slovenia). - : ELRA.
-
Konferensbidrag (refereegranskat)abstract
- This paper presents the Gavagai Living Lexicon, which is an online distributional semantic model currently available in 14 different languages. We describe the underlying distributional semantic model, and how we have solved some of the challenges in applying such a model to large amounts of streaming data. We also describe the architecture of our implementation, and discuss how we deal with continuous quality assurance of the lexicon.
|
|
| 3. |
- Evertsson, Maria, et al.
(författare)
-
Frequency- and Phase-Sensitive Magnetomotive Ultrasound Imaging of Superparamagnetic Iron Oxide Nanoparticles
- 2013
-
Ingår i: IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control. - 0885-3010. ; 60:3, s. 481-491
-
Tidskriftsartikel (refereegranskat)abstract
- It has recently been demonstrated that superparamagnetic iron oxide nanoparticles can be used as magnetomotive ultrasound contrast agents. A time-varying external magnetic field acts to move the particles and, thus, the nanoparticle-laden tissue. However, the difficulty of distinguishing this magnetomotive motion from undesired movement induced in regions without nanoparticles or other motion artifacts has not been well reported. Using a high-frequency linear-array system, we found that displacements outside nanoparticle-laden regions can be similar in magnitude to those in regions containing nanoparticles. We also found that the displacement outside the nanoparticle regions had a phase shift of approximately p radians relative to that in the nanoparticle regions. To suppress signals arising from undesirable movements, we developed an algorithm based on quadrature detection and phase gating at the precise frequency of nanoparticle displacement. Thus, clutter at other frequencies can be filtered out, and the processed signal can be color-coded and superimposed on the B-mode image. The median signal-to-clutter ratio improvement using the proposed algorithm was 36 dB compared with simply summing the movement energy at all frequencies. This clutter rejection is a crucial step to move magnetomotive ultrasound imaging of nanoparticles toward in vivo investigations.
|
|
| 4. |
- Evertsson, Maria, et al.
(författare)
-
Magnetomotive Ultrasound Imaging Of Rat Lymph Nodes In Situ: Assessment Of Imaging Parameters
- 2013
-
Ingår i: 2013 IEEE International Ultrasonics Symposium (IUS). - 9781467356855 ; , s. 600-603
-
Konferensbidrag (refereegranskat)abstract
- Detection and removal sentinel lymph nodes (SLN) is important in the diagnosis and treatment of breast cancer and malignant melanoma. The SLN is the first regional lymph node draining the cancer tumor and if the cancer has spread it is most likely to find cancer cells in the SLN. In this study we have been able to detect multimodal superparamagnetic iron oxide nanoparticles (SPIO-NP) in rat SLNs in situ using magnetomotive ultrasound imaging (MMUS). In MMUS a time-varying external magnetic field acts to move the NPs and, thus, the NP-laden tissue. This movement can be detected by proper processing of ultrasound data. We have recently developed an MMUS algorithm, based on quadrature detection and phase gating at the frequency of NP displacement, and this is the first study where the algorithm is evaluated in animals. For both higher NP-concentration, as well as smaller NPs, we found that the MMUS data showed a larger displacement (1.56 +/- 0.43 and 1.94 +/- 0.54 times larger, respectively). The MMUS displacement also increased with a lower excitation frequency (1.95 +/- 0.64 times larger for 5 Hz compared to 15 Hz) and higher excitation voltage (2.95 +/- 1.44 times larger for 30V compared to 10V). The results from this study show that the MMUS technique has potential to be used as bedside guidance during SLN surgery, well as being used as standalone technique in a number of other applications such as stem cell tracking and cardiovascular research.
|
|
| 5. |
- Evertsson, Maria, et al.
(författare)
-
Normalization of Magnetic Field Effects: Towards Quantitative Magnetomotive Ultrasound Imaging
- 2012
-
Ingår i: 2011 IEEE International Ultrasonics Symposium (IUS). - 9781457712531 ; , s. 775-778
-
Konferensbidrag (refereegranskat)abstract
- In magnetomotive ultrasound (MMUS) imaging superparamagnetic iron oxide nanoparticles (NP) are used as contrast agents and a time-varying external magnetic field acts to move the particles and thereby the nanoparticle-laden tissue. Recently we proposed a frequency and phase sensitive algorithm to reduce motion artifacts. However, the method is not quantitative as the particle movement induced is dependent not only on the field strength, but also on the field gradient, plus material parameters. Here we assess the measured nanoparticle movement across the image plane in comparison with simulations of the magnetic force, to evaluate the potential for image normalization of magnetic field effects. We found that the movement decreased with distance to the iron core tip, from which the magnetic field was extending, and approaches zero at the transducer face. This finding did not coincide with the simulation and may make it difficult to enable quantification. The coefficient of variation between measurements on the homogeneous phantom was typically in the order of 15% for all frequencies, indicating the expected accuracy for quantitative measurements.
|
|
| 6. |
- Holst, Maria, et al.
(författare)
-
Phase-locked magnetomotive ultrasound imaging of superparamagnetic iron-oxide nanoparticles
- 2010
-
Ingår i: 2010 IEEE International Ultrasonics Symposium, IUS 2010. - 9781457703829 - 9781457703812 ; , s. 1007-1010
-
Konferensbidrag (refereegranskat)abstract
- It has recently been shown that superparamagnetic iron oxide nanoparticles (SPIO-NP) can be used as ultrasound contrast agents, by moving the particles using a time-varying external magnetic field and thereby moving nanoparticle laden tissue. However, the difficulty to distinguish magneto-motive motion from indirect motion induced outside areas containing nanoparticles, or other motion artifacts, which restrict the use in vivo, is not well reported. Using a high-frequency linear array system we found that displacements outside nanoparticle-laden areas could be in the same range as the displacement in areas containing nanoparticles. We also found in our setup that the movement in areas between the nanoparticles areas were approximately radians out of phase relative the nanoparticles areas. To suppress undesirable signals we developed an algorithm based on frequency-tracking and phase-locking, where the energy of the processed signal is color-coded and overlaid on the B-mode image. The signal-to-clutter ratio showed a 14.5 dB improvement between the image color-coded with the total power and the image coded with regard to phase and frequency. The capability to distinguish areas containing SPIO-NP from those that do not is a crucial step for detection of nanoparticles in vivo.
|
|
| 7. |
|
|
| 8. |
|
|
| 9. |
|
|
| 10. |
|
|
