Diffuse Lung Diseases (DLD), impute to 15% of respiratory practice and are accountable for a large class of disorders, primarily affecting lung parenchyma. As a part of the diagnostic workup by the physician, a chest CT image is often required in addition to a thorough medical history and physical examination. The reliable identification of the features among interstitial lung diseases and the patterns they may take is challenging, particularly given the volume of data on a CT scan that must be processed by the radiologist. It has been shown that even among expert chest radiologists there is significant inter-observer and intra-observer variability. To make an objective quantitative and qualitative assessment of lung disease patterns, an accurate and reliable computer aided diagnostic system is likely to be extremely useful to assist with dealing with data volume for an expert radiologist. There will also be the opportunity to improve sensitivity and specificity in a non-expert radiologist group. Literature suggests that computer based pattern classifiers can discern image abnormalities due to lung diseases such as consolidation, cyst, emphysema, fibrosis, ground glass opacity, honey combing, nodularity, reticulation, scar and tree-in-bud. Researchers have focused on developing algorithms to quantify and analyse the surface changes of the lung, since DLD patterns often manifest as texture differences within the lung parenchyma. Research reported in this thesis has incorporated texture quantification, fractal analysis and scale invariant feature transform methods as complementary feature extraction techniques to improve the classification accuracy, especially in the presence of large number of classes associated with interstitial diseases. Classification of ten lung pathologies and healthy lung regions are validated based on different combination of diseases using leave-one-out and 5-fold cross validation techniques and an Artificial Neural Network (ANN). Classification accuracy based on features selected using scale invariant feature transform method alone generates 99% accuracy for up to four classes and more than 71% for up to eleven classes using an ANN. Classification accuracy is 85% for eleven classes using a combination of scale invariant feature transform, texture and fractal based features. Classification accuracies improve for higher number of classes (> 5) when the combination of above mentioned features are incorporated. Detailed classification accuracies for several DLD features compared to a healthy lung, and combinations of DLD features, such as fibrosis, reticulation, honey combing in comparison with healthy lung are evaluated throughout this thesis. / Thesis / Doctor of Philosophy (PhD)

No description available.

This thesis presents a prototype beacon-based indoor positioning system using IR-based triangulation together with various inertial sensors mounted onto the receiver. By applying a Kalman filter, the mobile receivers can estimate their position by fusing the data received from the two independent measurement systems. Furthermore, the system is aimed to operate and conduct all calculations using microcontrollers. Multiple IR beacons and an AGV were constructed to determine the systems performance. Empirical and practical experiments show that the proposed localisation system is capable centimeter accuracy. However, because of hardware limitation the system has lacking update frequency and range. With the limitations in mind, it can be established that the final sensor-fused solution shows great promise but requires an extended component assessment and more advanced localisation estimations method such as an Extended Kalman Filter or particle filter to increase reliability.

Imbalanced training data is a common problem in machine learning applications. Thisproblem refers to datasets in which the foreground pixels are significantly fewer thanthe background pixels. By training a machine learning model with imbalanced data, theresult is typically a model that classifies all pixels as the background class. A result thatindicates no presence of a specific condition when it is actually present is particularlyundesired in medical imaging applications. This project proposes a sequential system oftwo fully convolutional neural networks to tackle the problem. Semantic segmentation oflung nodules in thoracic computed tomography images has been performed to evaluate theperformance of the system. The imbalanced data problem is present in the training datasetused in this project, where the average percentage of pixels belonging to the foregroundclass is 0.0038 %. The sequential system achieved a sensitivity of 83.1 % representing anincrease of 34 % compared to the single system. The system only missed 16.83% of thenodules but had a Dice score of 21.6 % due to the detection of multiple false positives. Thismethod shows considerable potential to be a solution to the imbalanced data problem withcontinued development.

Radiotherapy plays an increasingly important role in cancer treatment, and medical imaging plays an increasingly important role in radiotherapy. Magnetic resonance imaging (MRI) is poised to be a major component in the development towards more effective radiotherapy treatments with fewer side effects. This thesis attempts to contribute in realizing this potential. Radiotherapy planning requires simulation of radiation transport. The necessary physical properties are typically derived from CT images, but in some cases only MR images are available. In such a case, a crude but common approach is to approximate all tissue properties as equivalent to those of water. In this thesis we propose two methods to improve upon this approximation. The first uses a machine learning algorithm to automatically identify bone tissue in MR. The second, which we refer to as atlas-based regression, can be used to generate a realistic, patient-specific, pseudo-CT directly from anatomical MR images. Atlas-based regression uses deformable registration to estimate a pseudo-CT of a new patient based on a database of aligned MR and CT pairs. Cancerous tissue has a dierent structure from normal tissue. This affects molecular diusion, which can be measured using MRI. The prototypical diusion encoding sequence has recently been challenged with the introduction of more general  waveforms. To take full advantage of their capabilities it is, however, imperative to respect the constraints imposed by the hardware while at the same time maximizing the diffusion encoding strength. In this thesis we formulate this as a constrained optimization problem that is easily adaptable to various hardware constraints.

Trabecular bone structure as well as bone mineral density (BMD) have impact on the biomechanical competence of bone. In osteoporosis-related fractures, there have been shown to exist disconnections in the trabecular network as well as low bone mineral density. Imaging of bone parameters is therefore of importance in detecting osteoporosis. One available imaging device is cone-beam computed tomography (CBCT). This device is often used in pre-operative imaging of dental implants, for which the trabecular network also has great importance. Fourteen or 15 trabecular bone specimens from the radius were imaged for conducting this in vitro project. The imaging data from one dual-energy X-ray absorptiometry (DXA), two multi-slice computed tomography (MSCT), one high-resolution peripheral quantitative computed tomography (HR-pQCT) and four CBCT devices were segmented using an in-house developed code based on homogeneity thresholding. Seven trabecular microarchitecture parameters, as well as two trabecular bone stiffness parameters, were computed from the segmented data. Measurements from micro-computed tomography (micro-CT) data of the same bone specimens were regarded as gold standard. Correlations between MSCT and micro-CT data showed great variations, depending on device, imaging parameters and between the bone parameters. Only the bone-volume fraction (BV/TV) parameter was stable with strong correlations. Regarding both HR-pQCT and CBCT, the correlations to micro-CT were strong for bone structure parameters as well as bone stiffness parameters. The CBCT device 3D Accuitomo showed the strongest correlations, but overestimated BV/TV more than three times compared to micro-CT. The imaging protocol most often used in clinical imaging practice at our clinic demonstrated strong correlations as well as low radiation dose. CBCT data of trabecular bone can be used for analysing trabecular bone properties, like bone microstructure and bone biomechanics, showing strong correlations to the reference method of micro-CT. The results depend on choice of CBCT device as well as segmentation method used. The in-house developed code based on homogeneity thresholding is appropriate for CBCT data. The overestimations of BV/TV must be considered when estimating bone properties in future clinical dental implant and osteoporosis research.

Deep learning methods for medical image segmentation are hindered by the lack of training data. This thesis aims to develop a method that overcomes this problem. Basic U-net trained on XCAT phantom data was tested first. The segmentation results were unsatisfactory even when artificial quantum noise was added. As a workaround, CycleGAN was used to add tissue textures to the XCAT phantom images by analyzing patient CT images. The generated images were used totrain the network. The textures introduced by CycleGAN improved the segmentation, but some errors remained. Basic U-net was replaced with Attention U-net, which further improved the segmentation. More work is needed to fine-tune and thoroughly evaluate the method. The results obtained so far demonstrate the potential of this method for the segmentation of medical images. The proposed algorithms may be used in iterative image reconstruction algorithms in multi-energy computed tomography.

When radiologists examine X-rays, it is crucial that they are aware of the laterality of the examined body part. The laterality refers to which side of the body that is considered, e.g. Left and Right. The consequences of a mistake based on information regarding the incorrect laterality could be disastrous. This thesis aims to address this problem by providing a deep neural network model that classifies X-rays based on their laterality. X-ray images contain markers that are used to indicate the laterality of the image. In this thesis, both a classification model and a detection model have been trained to detect these markers and to identify the laterality. The models have been trained and evaluated on four body parts: knees, feet, hands and shoulders. The images can be divided into three laterality classes: Bilateral, Left and Right. The model proposed in this thesis is a combination of two classification models: one for distinguishing between Bilateral and Unilateral images, and one for classifying Unilateral images as Left or Right. The latter utilizes the confidence of the predictions to categorize some of them as less accurate (Uncertain), which includes images where the marker is not visible or very hard to identify. The model was able to correctly distinguish Bilateral from Unilateral with an accuracy of 100.0 %. For the Unilateral images, 5.00 % were categorized as Uncertain and for the remaining images, 99.99 % of those were classified correctly as Left or Right.

Prostate cancer is one of the most common cancer diagnosis in men. This project aimed to help in characterization and treatment planning of prostate cancer by producing a Gleason grading probability based on apparent diffusion coefficient (ADC). In a study, from which this project received the patient data, the patients were first imaged using magnetic resonance imaging (MRI) in a 3T positron emission tomography MRI (PET/MRI) scanner. The prostates were surgically removed and placed in a patient specific mold. While inside the mold, the prostates were imaged using the same scanner, producing ex-vivo images of the prostates. Lastly the prostates were cut in histopathology slices and Gleason graded by a pathologist. To get correlation between ADC and Gleason grade all images needed to be correctly related to each other. This was done by three image registrations, which was the main part of this project. The histopathology slices were first registered to the ex-vivo images of the prostate, and then to the in-vivo T2-weighted images. The in-vivo T2w images were matched to images depicting the diffusion of water in the prostates, known as ADC-maps. The ADC-values were collected and matched to their possible Gleason grade. Information from 149 images were used, which came from 22 different patients. 3D pixels, known as voxels, with a corresponding Gleason grade annotation measured a lower average ADC-value. These voxels also showed more variation with a larger standard deviation. Furthermore, these voxels measured a larger range of ADC-values compared to voxels without a corresponding Gleason grade, but the probability of a Gleason grade was mainly seen for ADC-values below 1200 mm2/s. Filtering the ADC-map before collecting the information showed less spread in measurements, and larger total probability of Gleason grade annotation for lower ADC-values. To test the validity of the result a movement of the Gleason grade map was used to simulate registration errors. No large impact was observed for small movements but more obvious change for large. The results indicate this method as promising in predicting regions with a probability for Gleason grade of 3 or 4, however it was less accurate in separating the two. Gleason 5 showed very low probability, mainly as a result of the low sample size since only two patients had such tumors. Further research with better optimized filtering is recommended in the future.

The goal of this master thesis is to develop a computerized method for automatic estimation of the mammographic density of mammographic images from 5 different types of mammography units.   Mammographic density is a measurement of the amount of fibroglandular tissue in a breast. This is the single most attributable risk factor for breast cancer; an accurate measurement of the mammographic density can increase the accuracy of cancer prediction in mammography. Today it is commonly estimated through visual inspection by a radiologist, which is subjective and results in inter-reader variation.   The developed method estimates the density as a ratio of #pixels-containing-dense-tissue over #pixels-containing-any-breast-tissue and also according to the BI-RADS density categories. To achieve this, each mammographic image is: corrected for breast thickness and normalized such that some global threshold can separate dense and non-dense tissue. iteratively thresholded until a good threshold is found.  This process is monitored and automatically stopped by a classifier which is trained on sample segmentations using features based on different image intensity characteristics in specified image regions. filtered to remove noise such as blood vessels from the segmentation. Finally, the ratio of dense tissue is calculated and a BI-RADS density class is assigned based on a calibrated scale (after averaging the ratings of both craniocaudal images for each patient). The calibration is based on resulting density ratio estimations of over 1300 training samples against ratings by radiologists of the same images.   The method was tested on craniocaudal images (not included in the training process) acquired with different mammography units of 703 patients which had also been rated by radiologists according to the BI-RADS density classes. The agreement with the radiologist rating in terms of Cohen’s weighted kappa is substantial (0.73). In 68% of the cases the agreement is exact, only in 1.2% of the cases the disagreement is more than 1 class.