Interventional magnetic resonance imaging (MRI) aims at performing minimally invasive percutaneous interventions, such as tumor ablations and biopsies, under MRI guidance. During such interventions, the acquired MR image planes are typically aligned to the surgical instrument (needle) axis and to surrounding anatomical structures of interest in order to efficiently monitor the advancement in real-time of the instrument inside the patient's body. Object tracking inside the MRI is expected to facilitate and accelerate MR-guided interventions by allowing to automatically align the image planes to the surgical instrument. In this PhD thesis, an image-based workflow is proposed and refined for automatic image plane alignment. An automatic tracking workflow was developed, performing detection and tracking of a passive marker directly in clinical real-time images. This tracking workflow is designed for fully automated image plane alignment, with minimization of tracking-dedicated time. Its main drawback is its inherent dependence on the slow clinical MRI update rate. First, the addition of motion estimation and prediction with a Kalman filter was investigated and improved the workflow tracking performance. Second, a complementary optical sensor was used for multi-sensor tracking in order to decouple the tracking update rate from the MR image acquisition rate. Performance of the workflow was evaluated with both computer simulations and experiments using an MR compatible testbed. Results show a high robustness of the multi-sensor tracking approach for dynamic image plane alignment, due to the combination of the individual strengths of each sensor.

This thesis describes how a solution can be built to detect human flu-like symptoms. Flu-like symptoms are important to detect to prevent Covid-19 [6]. As people are returning to work there is a need for a simple way of detecting flu-like symptoms to prevent the spread of Covid-19. Other than a solution, this thesis concluded how human flu-like symptoms can be detected, with cameras specifically. This is to know what symptoms are most likely to work for a prototype. The technique of cameras and thermal cameras made this project possible as well as the technique of a single-board computer. The technique of cloud-based services is also an important part of this project. This project has resulted in a novel prototype using a single-board computer, cameras, and various cloud-based services to detect and inform a person if he or she has a human flu-like symptom.

Liver segmentation is a cumbersome task when done manually, often consuming quality time of radiologists. Use of automation in such clinical task is fundamental and the subject of most modern research. Various computer aided methods have been incorporated for this task, but it has not given optimal results due to the various challenges faced as low-contrast in the images, abnormalities in the tissues, etc. As of present, there has been significant progress in machine learning and artificial intelligence (AI) in the field of medical image processing. Though challenges exist, like image sensitivity due to different scanners used to acquire images, difference in imaging methods used, just to name a few. The following research embodies a convolutional neural network (CNN) was incorporated for this process, specifically a U-net algorithm. Predicted masks are generated on the corresponding test data and the Dice similarity coefficient (DSC) is used as a statistical validation metric for performance evaluation. Three datasets, from different scanners (two1.5 T scanners and one 3.0 T scanner), have been evaluated. The U-net performs well on the given three different datasets, even though there was limited data for training, reaching upto DSC of 0.93 for one of the datasets.

Background: Glioblastoma (GBM) is the most common and malignant primary brain tumor. It is a rapidly progressing tumor that infiltrates the adjacent healthy brain tissue and is difficult to treat. Despite modern treatment including surgical resection followed by radiochemotherapy and adjuvant chemotherapy, the outcome remains poor. The median overall survival is 10-12 months. Neuroimaging is the most important diagnostic tool in the assessment of GBMs and the current imaging standard is contrast-enhanced magnetic resonance imaging (MRI). Positron emission tomography (PET) has been recommended as a complementary imaging modality. PET provides additional information to MRI, in biological behavior and aggressiveness of the tumor. This study aims to investigate if the combination of PET and MRI can improve the diagnostic assessment of these tumors. Patients and methods: In this study, 22 patients fulfilled the inclusion criteria, diagnosed with GBM, and participated in all four 18F-fluorothymidine (FLT)-PET/MR examinations. FLT-PET/MR examinations were performed preoperative (baseline), before the start of the oncological therapy, at two and six weeks into therapy. Optimization of an adaptive thresholding algorithm, a batch processing pipeline, and image feature extraction algorithms were developed and implemented in MATLAB and the analyzing tool imlook4d. Results: There was a significant difference in radiochemotherapy treatment response between long-term and short-term survivors’ tumor volume in MRI (p<0.05), and marginally significant (p<0.10) for maximum standard uptake value (SUVmax), PET tumor volume, and total lesion activity (TLA). Preoperative short-term survivors had on average larger tumor volume, higher SUV, and total lesion activity (TLA). The overall trend seen was that long-term survivors had a better treatment response in both MRI and PET than short-term survivors.  During radiochemotherapy, long-term survivors displayed shrinking MR tumor volume after two weeks, and almost no remaining tumor volume was left after six weeks; the short-term survivors display marginal tumor volume reduction during radiochemotherapy. In PET, long-term survivors mean tumor volumes start to decrease two weeks into radiochemotherapy. Short-term survivors do not show any PET volume reduction two and six weeks into radiochemotherapy. For patients with more or less than 200 days progression-free survival, PET volume and TLA were significantly different, and MR volume only marginally significant, suggesting that PET possibly could have added value. Conclusion: The combination of PET and MRI can be used to predict radiochemotherapy response between two and six weeks, predicting overall survival and progression-free survival using MR and PET volume, SUVmax, and TLA. This study is limited by small sample size and further research with greater number of participants is recommended.

Laser speckle contrast imaging (LSCI) and its extension, multi-exposure laser speckle contrast imaging (MELSCI) are non-invasive techniques to monitor peripheral blood perfusion. One of the main drawbacks of LSCI and MELSCI in clinical use is that the techniques are sensitive to tissue movement. Moreover, the image background contributes to unnecessary data. The aim of this project was to develop and evaluate different methods to detect local motion artifacts and image backgrounds in LSCI and MELSCI. In this project, three different methods were developed: one using statistical analysis and two using machine learning. The method based on classical statistics was developed in MATLAB with a dataset made up of 1797 frames of 256 x 320 images taken from a recording of a hand where the thumb and middle finger were taking turns making small movements while the middle finger was the subject of three different states made by an occlusion cuff (baseline, occlusion, and reperfusion). The main filter that was used in the first method was the Hampel filter. Furthermore, networks for the machine learning method were developed in Python using the same dataset but with 20,000 small patches extracted from the dataset of sizes 3 x 3 to 21 x 21 pixels. The first machine learning method was based on two-dimensional data patches, hence no time dimension was included, while the second machine learning method used three-dimensional data patches where the time dimension was included (from 1s to 10s). The generation of ground truth for the dataset was manually created frame by frame in a ground truth generation graphical user interface (GUI) in MATLAB. To assess the three methods, the Dice coefficient was used. The statistical method resulted in a Dice coefficient of 0.7557. The highest Dice coefficient for the machine learning method with a 2D dataset was 0.2902 (patch size 13 x 13) and the lowest was 0.2372 (patch size 7 x 7). For the machine learning method with 3D datasets, the patch size of 21 x 21 x 4 resulted in the highest Dice coefficient (0.5173), and the 21 x 21 x 40 model had the lowest Dice coefficient (0.1782). Since the two methods based on temporal data proved to be performing best in this project, one conclusion for further development of an improved model is the usage of temporal data in the training of a model. However, one important difference between the statistical method and the three-dimensional machine learning method is that the statistical method does not handle fast perfusion changes as well as the machine learning method and can not detect image background and static tissue. Therefore, the overall most useful method to further develop is the three-dimensional machine learning method.

Microscopic video images of microcirculation have been used in clinical diagnosis for years, and theparameters obtained from images reveal most physiological activities and body organizations.Particularly, the blood flow speed is one of important indexes, which reflects the state ofmicrocirculation and make significant marks in diagnosis.In order to measure capillary blood velocity, a quantity of methods and instruments have beenstudied and developed. Based on the format of measurement, microscopy approaches used widely,can be grouped into two categories. One direct way applies microscopic-imaging technology forvisualization. The other way uses assistant methods such as laser-illumination [1] or labeling RBCswith fluorescein isothiocyanate [2]. In previous study, four methods (Direct Observation Method,Dual-windows Method, Single-window Method, Optical Flow Method) have been studied andanalysed in order to achieve better performance. But still there is a non-negligible deviation inmeasurement within different tries and compared to the data we retrieve from hospital.This study, inspired by previous work, aims to further investigate efficient and reliable algorithms forextracting capillary blood velocity. One possible solution is to implement feature based estimation tocalculate the blood flow speed distribution automatically, point by point along the middle line oftargeting blood vessel. We inherit the idea of generating motion vectors from Optic Flow algorithmwhich has the best accuracy performance in vehicle identification domain. But original optic flowalgorithm makes the system too sophisticated and time consuming. Moreover, its two required basicrules may not stand during the blood flow velocity detection. So a customized feature basedestimation is brought up here and supposed to be a practicable method for analysis not only inaccuracy but also in efficiency. Moreover, this report also introduces picture shifting, red blood cellmotion, and double windows marking to compare and to confirm the results. Previous work will beused as a reference for the assessment of new algorithms. / Mikroskopiska videobilder av mikrosirkulation har använts vid klinisk diagnos i flera år, och parametrarna erhållna från bilder avslöjar de flesta fysiologiska aktiviteter och kroppsorganisationer. Särskilt är blodflödeshastigheten ett av viktiga index, som återspeglar tillståndet för mikrosirkulation och gör betydande märken vid diagnosen.För att mäta kapillärblodshastighet har en mängd metoder och instrument studerats och utvecklats. Baserat på mätformatet kan mikroskopimetoder som används allmänt grupperas i två kategorier. Ett direkt sätt använder mikroskopisk bildteknologi för visualisering. Det andra sättet använder assistentmetoder som laserbelysning [1] eller märkning av RBC med fluoresceinisotiocyanat [2]. I tidigare studier har fyra metoder (Direct Observation Method, Dual-windows Method, Single-Window Method, Optical Flow Method) studerats och analyserats för att uppnå bättre prestanda. Men det finns fortfarande en icke försumbar avvikelse i mätningen inom olika försök och jämfört med de data vi hämtar från sjukhuset.Denna studie, inspirerad av tidigare arbete, syftar till att ytterligare undersöka effektiva och tillförlitliga algoritmer för att extrahera kapillärblodhastighet. En möjlig lösning är att implementera funktionsbaserad uppskattning för att beräkna blodflödeshastighetsfördelningen automatiskt, punkt för punkt längs mittlinjen för riktad blodkärl. Vi ärver idén att generera rörelsesvektorer från Optic Flow-algoritmen som har den bästa noggrannhetsprestanda inom fordonsidentifieringsdomän. Men den ursprungliga optiska flödesalgoritmen gör systemet för sofistikerat och tidskrävande. Dessutom kanske dess två nödvändiga grundregler inte gäller under detektionen av blodflödeshastighet. Så en anpassad funktionsbaserad uppskattning tas upp här och antas vara en genomförbar metod för analys inte bara i noggrannhet utan också i effektivitet. Dessutom introducerar detta papper också bildförskjutning, rörelse av röda blodkroppar och dubbla fönstermarkeringar för att jämföra och bekräfta resultaten. Tidigare arbete kommer att användas som referens förbedömning av nya algoritmer.

Hypothesis testing has long been a formal and standardized process. Hypothesis generation, on the other hand, remains largely informal. This thesis assess whether eXplainable AI (XAI) can aid in the standardization of hypothesis generation through its utilization as a hypothesis generating tool for medical research. We produce XAI heat maps for a Convolutional Neural Network (CNN) trained to classify Microsatellite Instability (MSI) in colon and gastric cancer with four different XAI methods: Guided Backpropagation, VarGrad, Grad-CAM and Sobol Attribution. We then compare these heat maps with pathology annotations in order to look for differences to turn into new hypotheses. Our CNN successfully generates non-random XAI heat maps whilst achieving a validation accuracy of 85% and a validation AUC of 93% – as compared to others who achieve a AUC of 87%. Our results conclude that Guided Backpropagation and VarGrad are better at explaining high-level image features whereas Grad-CAM and Sobol Attribution are better at explaining low-level ones. This makes the two groups of XAI methods good complements to each other. Images of Microsatellite Insta- bility (MSI) with high differentiation are more difficult to analyse regardless of which XAI is used, probably due to exhibiting less regularity. Regardless of this drawback, our assessment is that XAI can be used as a useful hypotheses generating tool for research in medicine. Our results indicate that our CNN utilizes the same features as our basic pathology annotations when classifying MSI – with some additional features of basic pathology missing – features which we successfully are able to generate new hypotheses with.

Diffusion spectrum imaging is a type of diffusion magnetic resonance imaging, capable of capturing very complex tissue structures, but requiring a very large amount of samples in q-space and therefore time.  The purpose of this project was to create and evaluate a new sampling scheme in q-space for diffusion MRI, trying to recreate the ensemble averaged propagator (EAP) with fewer samples without significant loss of quality. The sampling scheme was created by greedily selecting the measurements contributing with the most mutual information. The EAP was then recreated using the sampling scheme and interpolation. The mutual information was approximated using the kernel from a Gaussian process machine learning model.  The project showed limited but promising results on synthetic data, but was highly restricted by the amount of available computational power. Having to resolve to using a lower resolution mesh when calculating the optimal sampling scheme significantly reduced the overall performance.

There are several reasons for using non-Cartesian k-space sampling methods in Magnetic Resonance Imaging (MRI). Such a method is radial sampling, which includes the advantage of continuous coverage of the k-space center which results in higher robustness to motion. On the other hand, radial imaging does have some limitations that must be considered. The method is more sensitive to gradient imperfections, such as eddy currents and gradient delays, resulting in inconsistencies between the nominal and actual gradient waveforms. This leads to distortions in the sampling trajectory, also called trajectory errors, yielding reconstructed images with artifacts caused by the gradient imperfections. The aim of this project was therefore to implement a method that takes these errors into account and perform a correction of the trajectory errors to yield images with reduced artifacts. Various methods have been proposed for correction of the gradient errors, some more effective than others. The method implemented in this project was based on the gradient impulse response function (GIRF) which characterizes the gradient system responses. When GIRF was acquired, the actual gradient waveforms played-out during the imaging measurement could be predicted by first modeling the nominal gradient waveforms and then performing a convolution with the corresponding GIRF for each gradient axis. The imaging experiments involved measurements on two different resolution phantoms and in-vivo measurements to note possible differences in correction performance. The used pulse sequences for imaging were FLASH and bSSFP. The results showed that the applied method using GIRF did reduce the artifacts caused by gradient imperfections in the reconstructed images taken with the FLASH sequence. On the other hand, the results for the bSSFP sequence were not as successful due to incomplete modeling of the gradient waveforms. The conclusion to be drawn is that the GIRF-correction does adequately compensate for the trajectory errors when using a radial sampling trajectory for the FLASH sequence and hence yield images with almost eliminated artifacts. A suggestion for future work would be to further investigate the bSSFP sequence modeling to obtain better bSSFP-images. / Det finns flera anledningar till att använda icke-Kartesiska k-space samplingsmetoder i magnetisk resonanstomografi. En sådan metod är radiell sampling, som har fördelen att kontinuerligt samla in mätdata från mittpunkten av k-space, vilket resulterar i lägre rörelsekänslighet under bildtagningstillfället. Radiell sampling har dock begränsningar som måste tas i beaktande, som gradient imperfektioner och gradientfördröjningar. Dessa leder till förvrängningar i samplingspositioneringen i k-space, även känt som trajektoriefel, vilket ger upphov till artefakter vid bildrekonstruktion. Syftet med projektet är att korrigera för dessa trajektoriefel så att den rekonstruerade bilden innehåller färre artefakter. Olika metoder har föreslagits för korrektion av gradientfel. Metoden som användes i detta projekt baseras på gradient impulsresponsfunktionen (GIRF), som karaktäriserar gradient systemet. För att estimera de verkliga samplingspositionerna i k-space beräknades de förvrängda gradientvågformerna efter varje mätning. Detta gjordes genom att först modellera de nominella gradientvågformerna och därefter utföra en faltning med GIRF. De utförda experimenten under projektets gång bestod av bildtagning av två fantomer och ett antal in-vivo mätningar för att identifiera eventuella skillnader i de rekonstruerade bilderna. Pulssekvenserna som användes under projektet var FLASH och bSSFP. Resultaten visade att GIRF-korrektionen reducerade artefakter orsakade av gradient imperfektioner i de rekonstruerade bilderna tagna med FLASH-sekvensen. Erhållna resultat med bSSFP-sekvensen var å andra sidan inte lika lyckade på grund av inkomplett modellering av gradientvågformerna. Slutsatsen som kan dras är att GIRF-korrektionen kompenserar för trajektoriefel i radiell sampling för FLASH-sekvensen och ger rekonstruerade bilder där artefakterna nästan eliminerats. Ett förslag för framtida arbeten är att vidare undersöka modelleringen av bSSFP-sekvensen för att erhålla bättre bilder.

Medical imaging based on radioactive tracers exposes the patient to radiation. For this reason, a phantom is preferably used for non-clinical studies such as routine quality assurance and research. The aim of this project was to design, build and test a multi-compartment phantom to be used in dynamic SPECT/CT, PET/CT and PET/MRI studies. By treating each compartment as a biological system and plotting activity distribution, desired characteristics of the phantom can be obtained. A software program was created to simulate compartment activity distribution for different input parameters. Such parameters include number of compartments, administered activity, flow rates between compartments and compartment volume. Based on the simulation, the phantom was designed to meet the desired characteristics. Due to the outbreak of the SARS-CoV-2 virus, no phantom could be built nor tested. Consequently, leading the project to create a foundation that facilitates future building of the phantom. / Medicinsk avbildning med radioaktiva spårämnen utsätter patienter för en stråldos. Av detta skäl används företrädesvis en fantom för icke-kliniska studier såsom rutinmässig kvalitetssäkring och forskning. Syftet med detta projekt var att designa, bygga och testa ett flerkammarfantom som ska användas i dynamiska SPECT/CT, PET/CT och PET/MRI studier. Genom att behandla varje kammare som ett biologiskt system och plotta aktivitetsfördelning kan önskade egenskaper hos fantomen erhållas. Ett program skapades för att simulera aktivitetsdistributionen i flerkammarfantomer för olika in parametrar så som antal kammare, administrerad aktivitet, flöden mellan kammare och kammarvolym. Baserat på simuleringen utformades fantomen för att uppfylla de önskade egenskaperna. På grund av utbrottet av SARS-CoV-2 viruset kunde ingen fantom byggas eller testas. Följaktligen leddes projektet till att skapa en grund som underlättar framtida byggande av fantomen.