| 1. |
- Vu, Minh Hoang, 1988-
(författare)
-
Resource efficient automatic segmentation of medical images
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Cancer is one of the leading causes of death worldwide. In 2020, there were around 10 million cancer deaths and nearly 20 million new cancer cases in the world. Radiation therapy is essential in cancer treatments because half of the cancer patients receive radiation therapy at some point. During a radiotherapy treatment planning (RTP), an oncologist must manually outline two types of areas of the patient’s body: target, which will be treated, and organs-at-risks (OARs), which are essential to avoid. This step is called delineation. The purpose of the delineation is to generate a sufficient dose plan that can provide adequate radiation dose to a tumor and limit the radiation exposure to healthy tissue. Therefore, accurate delineations are essential to achieve this goal.Delineation is tedious and demanding for oncologists because it requires hours of concentrating work doing a repeated job. This is a RTP bottleneck which is often time- and resource-intensive. Current software, such as atlasbased techniques, can assist with this procedure by registering the patient’s anatomy to a predetermined anatomical map. However, the atlas-based methods are often slowed down and erroneous for patients with abnormal anatomies.In recent years, deep learning (DL) methods, particularly convolutional neural networks (CNNs), have led to breakthroughs in numerous medical imaging applications. The core benefits of CNNs are weight sharing and that they can automatically detect important visual features. A typical application of CNNs for medical images is to automatically segment tumors, organs, and structures, which is assumed to save radiation oncologists much time when delineating. This thesis contributes to resource efficient automatic segmentation and covers different aspects of resource efficiency.In Paper I, we proposed a novel end-to-end cascaded network for semantic segmentation in brain tumors in the multi-modal magnetic resonance imaging challenge in 2019. The proposed method used the hierarchical structure of the tumor sub-regions and was one of the top-ranking teams in the task of quantification of uncertainty in segmentation. A follow-up work to this paper was ranked second in the same task in the same challenge a year later.We systematically assessed the segmentation performance and computational costs of the technique called pseudo-3D as a function of the number of input slices in Paper II. We compared the results to typical two-dimensional (2D) and three-dimensional (3D) CNNs and a method called triplanar orthogonal 2D. The typical pseudo-3D approach considers adjacent slices to be several image input channels. We discovered that a substantial benefit from employing multiple input slices was apparent for a specific input size.We introduced a novel loss function in Paper III to address diverse issues, including imbalanced datasets, partially labeled data, and incremental learning. The proposed loss function adjusts to the given data to use all accessible data, even if some lack annotations. We show that the suggested loss function also performs well in an incremental learning context, where an existing model can be modified to incorporate the delineations of newly appearing organs semi-automatically.In Paper IV, we proposed a novel method for compressing high-dimensional activation maps, which are the primary source of memory use in modern systems. We examined three distinct compression methods for the activation maps to accomplishing this. We demonstrated that the proposed method induces a regularization effect that acts on the layer weight gradients. By employing the proposed technique, we reduced activation map memory usage by up to 95%.We investigated the use of generative adversarial networks (GANs) to enlarge a small dataset by generating synthetic images in Paper V. We use the real and generated data during training CNNs for the downstream segmentation tasks. Inspired by an existing GAN, we proposed a conditional version to generate high-dimensional and high-quality medical images of different modalities and their corresponding label maps. We evaluated the quality of the generated medical images and the effect of this augmentation on the performance of the segmentation task on six datasets.
|
|
| 2. |
- Karlsson, Kåre
(författare)
-
Health problems and work-related stress in Swedish ambulance personnel
- 2020
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Background: Previous studies have shown a high incidence of both acute and post-traumatic stress among ambulance personnel. It has been shown that ambulance personnel are at a higher risk of being affected of heart disease, high blood pressure and cancer. Studies have also seen a higher incidence of substance abuse and suicide. One cause of these health problems can be work-related stress. No previous research has focused on the body's physical reactions in the form of changes in heart rate and stress hormones in ambulance personnel in connection with work-related stress. Nor is there any overall picture of what actually affects Swedish ambulance personnel in terms of reported morbidity.Aim: The overall aim of the thesis is to investigate health problems in Swedish ambulance personnel and to study if there are any factors related to the work environment and the special conditions occurring in the prehospital environment that can be linked to the findings that appear.The questions that should be answered are: Are there health problems that affect Swedish ambulance personnel to a higher extent than other professions in Sweden? Are there any factors relating to morbidity that can be linked to the profession and can be regarded as potentially dangerous? If so, are there methods to prevent health problems that can be implemented in daily work?Methods: Studies I, II and III were all studies where stress markers (heart rate and cortisol levels) were measured during different conditions linked to the profession. Study I was a study where this was measured during physical exertion e.g. to carry a stretcher. Study II was a validated theoretical stress test to see how the ambulance personnel reacted to unknown factors. Study III meant measuring stress markers during priority-1 alarms. Study IV was carried out as a longitudinal register study where data about ICD-codes was collected from Statistics Sweden and the National Board of Health and Welfare.Results: In study I it was shown that the use of lifting aids reduced the measurable stress in the form of both reduced heart rate and decreased cortisol levels. Study II showed that personnel were stressed of the unknown test though women had the highest salivary cortisol levels before the Trier social stress test while the highest value for men occurred 10 to 20 minutes after the test. Study III showed that there was an increase in heart rate during priority-1 alarms that could not be linked to physical activity. It also indicated/showed that traffic accidents, patients with heart attacks in need of acute PCI or thrombectomy and alarms regarding children generate the highest stress onset seen as prolonged elevation of cortisol levels. This was seen regardless of gender, age, education or experience in all three studies. Study IV showed that Swedish ambulance personnel run the risk of being affected by certain diseases such as cardiac arrhythmias and injuries as arthrosis of the knee, dorsopathies and intervertebral disc disorders to a greater extent compared to other health care workers and other professions in Sweden.Conclusion: Swedish ambulance personnel have a higher incidence of certain health problems and diseases such as paroxysmal tachycardia, atrial fibrillation and flutter, other cardiac arrhythmias, high blood pressure, of the knee and dorsopathies and intervertebral disc disorders compared to other professions. These conditions can be caused by work-related stress, although they are not the only cause. Some factors that cause a stress reaction that can be linked to the ambulance profession have emerged in the studies. But the extent to which these factors alone cause the health problems that Swedish ambulance personnel suffer from is more difficult to determine with certainty based on these studies. Using shoulder straps reduces both heart rate and cortisol secretion.Clinical implications: Greater use of aids both in terms of lifting, moving and carrying heavy loads should mean that the physical load on the body would be reduced. These aids could also prevent some musculoskeletal problems that have emerged in this thesis by relieving and distributing the burden throughout the body. An important aspect in preventing work-related stress is that sufficient time is given for recovery. The organizations need to be dimensioned so that there is sufficient standby time so that there is time for recovery and reflection but also time for education and training. This is something that is far from reality in many Swedish ambulance organizations. It is also time to seriously discuss what is an actual reality in many countries, namely that employees in the ambulance service can benefit from a lower retirement age after a certain number of years of service.
|
|
| 3. |
- Jonsson, R. Daniel, 1972-
(författare)
-
Analysing Sustainable Urban Transport and Land-Use : Modelling tools and appraisal frameworks
- 2008
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Sustainable development and climate change is high on the agenda for most cities around the world today. Urban transport is at the heart of these changes. Increasingly, it is recognised that not only is the emission of pollutants and greenhouse gases a problem, but also the detrimental effects of congestion and social exclusion. In order to address these issues, it will be necessary for cities to make strategic long term decisions regarding the future infrastructure and land use, not only in terms of what will be built, but also on measures that affect how these systems are used.This thesis is focused on the decision support tools that we need in order to make well informed decisions. Models that predict the performance of future scenarios, and appraisal frameworks that help evaluate whether these outcomes are desirable or not. The first two papers experiment with different ways of bringing some aspects of sustainability into the appraisal frameworks used to analyse long term strategies. Paper I addresses intergenerational fairness, and Paper II focuses on the emission of greenhouse gases. Paper III develops a model, Scapes, that can help us to better understand the daily travel behaviour, through an activity based approach. By explicitly modelling space-time constraints, and travel time uncertainty in a microeconomic framework, we can get a better understanding of how people can respond to, and value, changes in the transport system. Papers IV and V describe a new integrated land use and transport model, LandScapes.The policy implications from the studies in Papers I, II, and V are that it will be very difficult for Stockholm to reduce its emissions of CO2. Particularly, predicted economic and population growth will inevitably lead to more transport. It is likely that a range of different policies will be necessary to solve that problem. At the same time, we must not forget that decreasing CO2 emissions, although important, is not the only objective Stockholm has. To cope with the increasing travel demand from a growing population, it may well be necessary to build new infrastructure as well. This thesis does not prescribe any such relative valuation between conflicting objectives. It only helps bring them to the fore.
|
|
| 4. |
- Adjeiwaah, Mary, 1980-
(författare)
-
Quality assurance for magnetic resonance imaging (MRI) in radiotherapy
- 2019
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The use of Magnetic Resonance Imaging (MRI) in the radiotherapy (RT) treatment planning workflow is increasing. MRI offers superior soft-tissue contrast compared to Computed Tomography (CT) and therefore improves the accuracy in target volume definitions. There are, however concerns with inherent geometric distortions from system- (gradient nonlinearities and main magnetic field inhomogeneities) and patient-related sources (magnetic susceptibility effect and chemical shift). The lack of clearly defined quality assurance (QA) procedures has also raised questions on the ability of current QA protocols to detect common image quality degradations under radiotherapy settings. To fully implement and take advantage of the benefits of MRI in radiotherapy, these concerns need to be addressed.In Papers I and II, the dosimetric impact of MR distortions was investigated. Patient CTs (CT) were deformed with MR distortion vector fields (from the residual system distortions after correcting for gradient nonlinearities and patient-induced susceptibility distortions) to create distorted CT (dCT) images. Field parameters from volumetric modulated arc therapy (VMAT) treatment plans initially optimized on dCT data sets were transferred to CT data to compute new treatment plans. Data from 19 prostate and 21 head and neck patients were used for the treatment planning. The dCT and CT treatment plans were compared to determine the impact of distortions on dose distributions. No clinically relevant dose differences between distorted CT and original CT treatment plans were found. Mean dose differences were < 1.0% and < 0.5% at the planning target volume (PTV) for the head and neck, and prostate treatment plans, respectively. Strategies to reduce geometric distortions were also evaluated in Papers I and II. Using the vendor-supplied gradient non-linearity correction algorithm reduced overall distortions to less than half of the original value. A high acquisition bandwidth of 488 Hz/pixel (Paper I) and 488 Hz/mm (Paper II) kept the mean geometric distortions at the delineated structures below 1 mm. Furthermore, a patient-specific active shimming method implemented in Paper II significantly reduced the number of voxels with distortion shifts > 2 mm from 15.4% to 2.0%.B0 maps from patient-induced magnetic field inhomogeneities obtained through direct measurements and by simulations that used MR-generated synthetic CT (sCT) data were compared in Paper III. The validation showed excellent agreement between the simulated and measured B0 maps.In Paper IV, the ability of current QA methods to detect common MR image quality degradations under radiotherapy settings were investigated. By evaluating key image quality parameters, the QA protocols were found to be sensitive to some of the introduced degradations. However, image quality issues such as those caused by RF coil failures could not be adequately detected.In conclusion, this work has shown the feasibility of using MRI data for radiotherapy treatment planning as distortions resulted in a dose difference of less than 1% between distorted and undistorted images. The simulation software can be used to produce accurate B0 maps, which could then be used as the basis for the effective correction of patient-induced field inhomogeneity distortions and for the QA verification of sCT data. Furthermore, the analysis of the strengths and weaknesses in current QA tools for MRI in RT contribute to finding better methods to efficiently identify image quality errors.
|
|
| 5. |
- Simkó, Attila, 1995-
(författare)
-
Contributions to deep learning for imaging in radiotherapy
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Purpose: The increasing importance of medical imaging in cancer treatment, combined with the growing popularity of deep learning gave relevance to the presented contributions to deep learning solutions with applications in medical imaging.Relevance: The projects aim to improve the efficiency of MRI for automated tasks related to radiotherapy, building on recent advancements in the field of deep learning.Approach: Our implementations are built on recently developed deep learning methodologies, while introducing novel approaches in the main aspects of deep learning, with regards to physics-informed augmentations and network architectures, and implicit loss functions. To make future comparisons easier, we often evaluated our methods on public datasets, and made all solutions publicly available.Results: The results of the collected projects include the development of robust models for MRI bias field correction, artefact removal, contrast transfer and sCT generation. Furthermore, the projects stress the importance of reproducibility in deep learning research and offer guidelines for creating transparent and usable code repositories.Conclusions: Our results collectively build the position of deep learning in the field of medical imaging. The projects offer solutions that are both novel and aim to be highly applicable, while emphasizing generalization towards a wide variety of data and the transparency of the results.
|
|
| 6. |
- Haller, Henrik, 1977-
(författare)
-
Appropriate Technologies for Soil Remediation in Low Prioritized Region : Developing Countries and Sparsely Populated Regions
- 2015
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Contaminated sites in low prioritized regions demand remediation technologies that are cost- and energy-effective and locally adapted. Parameters such as the time frame during which bioremediation degradation needs to occur may not be as restraining as in urban environments. This licentiate project aims to explore opportunities and constraints for appropriate soil remediation based on organic by-products in tropical developing countries and sparsely populated areas in industrial countries. Ecological Engineering and the Framework for Strategic Sustainable Development were explored as planning tools to steer bioremediation methods towards sustainability. The use of the five concepts of Ecological Engineering within the Framework for Strategic Sustainable Development can steer bioremediation in low prioritized regions towards sustainability. Pilot-scale and laboratory experiments were undertaken to evaluate the feasibility of such bioremediation methods. Experiments carried out at the experiment station in Chontales, Nicaragua showed some promising results, but also revealed problems associated with the clay rich soils, which are typical for tropical regions. Treatment of diesel contaminated ultisol with 6 mL whey kg-1 dw in a pilot-scale experiment considerably increased the degradation rate of diesel constituents, but no effects on the degradation rates were observed after treatment with compost tea or pyroligneous acid.The soil columns study suggests that despite a favorable particle size distribution for microbial transport, the sandy loam retained a greater fraction of the microorganisms present in the ACT in the top 10 cm than the clay loam, presumably because the lower bulk density and higher SOM in the clay loam aided transport and growth of microorganisms.
|
|
| 7. |
- Jonsson, Anders, 1967-
(författare)
-
Dödsbränder i Sverige : En analys av datakvalitet, orsaker och riskmönster
- 2018
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Risken att omkomma genom brand har mer än halverats under de senaste 60 åren i Sverige. Det är idag mycket ovanligt att små barn omkommer i brand men samma positiva utveckling kan inte ses för de äldre. En åldrande befolkning som i ökande omfattning förväntas bo kvar hemma innebär att denna grupp måste prioriteras i framtida brandskyddsarbete. Sambearbetning av data från tre nationella register visar att rutinmässigt framtagen statistik systematiskt underskattar den verkliga situationen. Män, äldre, ensamboende och ekonomiskt svaga är särskilt riskutsatta grupper att omkomma vid bostadsbränder och förekomst av alkohol bland offren är mycket vanligt. Den i särklass vanligaste brandorsaken är rökning. Trots att varje dödsbrand är unik kan det omfattande materialet beskrivas av relativt få och tydligt avgränsade typer av händelser som var och en måste mötas med relevanta preventiva insatser. Avhandlingen visar att dödsbränder i bostäder i huvudsak är ett socialt problem och att det kommer att krävas uthålliga och breda strategier, bestående både av sociala och tekniska åtgärder för att skydda de mest riskutsatta och sårbara människorna i samhället.
|
|
| 8. |
- Jonsson, Marcus, 1977-, et al.
(författare)
-
In-Hospital Physiotherapy and Physical Recovery 3 Months After Lung Cancer Surgery : A Randomized Controlled Trial
- 2019
-
Ingår i: Integrative Cancer Therapies. - : Sage Publications. - 1534-7354 .- 1552-695X. ; 18
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Lung cancer is the most frequently diagnosed cancer and one of the leading causes of cancer deaths. Surgery is the primary approach for curative treatment. Postoperative complications are common, and physiotherapy is often routinely provided for their prevention and treatment, even though the evidence is limited. The aim of this study was to examine the effect of in-hospital physiotherapy on postoperative physical capacity, physical activity, and lung function among patients undergoing lung cancer surgery.Methods: A total of 107 patients undergoing elective thoracic surgery were included in a single-blinded randomized controlled trial, and randomized to a study group, receiving in-hospital physiotherapy treatment, or a control group, not receiving in-hospital physiotherapy treatment. The patients were assessed preoperatively and 3 months after surgery. The in-hospital physiotherapy treatment consisted of early mobilization, ambulation, breathing exercises, and thoracic range of motion exercises. Physical capacity was assessed with the 6-minute walk test. Level of physical activity was objectively assessed with an accelerometer and subjectively assessed with the International Physical Activity Questionnaire Modified for the Elderly.Results: Physical capacity for the whole sample was significantly decreased 3 months postoperatively compared with preoperative values (P = .047). There were no statistically significant differences between the groups regarding physical capacity, physical activity, spirometric values, or dyspnea. However, patients in the study group increased their level of self-reported physical activity from preoperatively to 3 months postoperatively, while the patients in the control group did not.Conclusions: No difference in physical capacity, physical activity, or lung function was found 3 months postoperatively in lung cancer surgery patients receiving in-hospital physiotherapy compared with control patients.
|
|
| 9. |
|
|
| 10. |
- Jonsson, Marcus, 1977-, et al.
(författare)
-
In-hospital physiotherapy improves physical activity level after lung cancer surgery : a randomized controlled trial
- 2019
-
Ingår i: Physiotherapy. - : Elsevier. - 0031-9406 .- 1873-1465. ; 105:4, s. 434-441
-
Tidskriftsartikel (refereegranskat)abstract
- OBJECTIVES: Patients undergoing lung cancer surgery are routinely offered physiotherapy. Despite its routine use, effects on postoperative physical recovery have yet not been demonstrated. The aim of this study was to investigate whether physiotherapy could improve postoperative in-hospital physical activity level and physical capacity.DESIGN: Single-blind randomized controlled trial.SETTING: Thoracic surgery department at a University Hospital.PARTICIPANTS: Patients undergoing elective thoracic surgery (n=94) for confirmed or suspected lung cancer were assessed during hospital stay.INTERVENTION: Daily physiotherapy, consisting of mobilization, ambulation, shoulder exercises and breathing exercises. The control group received no physiotherapy treatment.OUTCOMES: In-hospital physical activity assessed with the Actigraph GT3X+ accelerometer, six-minute walk test, spirometry and dyspnea scores.RESULTS: The treatment group reached significantly more accelerometer counts (2010 (1508) vs 1629 (1146), mean difference 495 [95% CI 44 to 1109]), and steps per hour (49 (47) vs 37 (34), mean difference 14 [95% CI 3 to 30]), compared to the control group, during the first three postoperative days. No significant differences in six-minute walk test (percent of preoperative 71% vs 79%, P=0.13), spirometry (FEV1 percent of preoperative 69% vs 69%, P=0.83) or dyspnoea (M-MRC 2 vs 2, P=0.74) between the groups were found.CONCLUSIONS: Patients receiving in-hospital physiotherapy showed increased level of physical activity during the first days after lung cancer surgery, compared to an untreated control group. However, no effects on the six-minute walk test or spirometric values were found. The clinical importance of an increased physical activity level during the early postoperative period needs to be further evaluated.CLINICAL TRIAL REGISTRATION NUMBER: NCT01961700.
|
|
| 11. |
|
|
| 12. |
- Jonsson, Marcus, 1977-, et al.
(författare)
-
Physical activity level during the first three days after lung cancer surgery improves with physiotherapy : a randomized controlled trial
- 2018
-
Ingår i: European Respiratory Journal. - : European Respiratory Society. - 0903-1936 .- 1399-3003. ; 52:Suppl. 62
-
Tidskriftsartikel (övrigt vetenskapligt/konstnärligt)abstract
- Objectives: Physical inactivity is common after lung cancer surgery. Patients undergoing lung cancer surgery are routinely offered physiotherapy. Despite its routine use, any effect on postoperative physical activity has not yet been demonstrated. The aim of this study was to investigate whether physiotherapy could improve physical activity during the first days after surgery.Methods: A total of 94 patients undergoing elective surgery for confirmed or suspected lung cancer were consecutevily included and randomized to treatment group (n=50) or control group (n=44). The treatment group received daily physiotherapy, consisting of mobilization and ambulation, shoulder exercises and breathing exercises. The control group received no physiotherapy. Physical activity was assessed with the Actigraph GT3X+ accelerometer.Results: The patients in the treatment group reached significantly more counts (1692 vs 1197, p=0.029) and steps per hour (39 vs 25, p=0.013), during the first three days, compared to the control group.Conclusions: Physical activity during the first three days is increased by physiotherapy treament. The long term effect of in-hospital physiotherapy needs to be further evaluated.
|
|
