No description available.

Image degradation arises from various environmental conditions due to the exis tence of aerosols such as fog, haze, and dust. These phenomena mitigate image vis ibility by creating color distortion, reducing contrast, and fainting object surfaces. Although the end-to-end deep learning approach has made significant progress in the field of homogeneous dehazing, the image quality of these algorithms in the context of non-homogeneous real-world images has not yet been satisfactory. We argue two main reasons that are responsible for the problem: 1) First, due to the unbalanced information processing of the high-level and low-level information in conventional dehazing algorithms, 2) due to lack of trainable data pairs. To ad dress the above two problems, we propose a parallel dual-branch design that aims to balance the processing of high-level and low-level information, and through a method of transfer learning, utilize the small data sets to their full potential. The results from the two parallel branches are aggregated in a simple fusion tail, in which the high-level and low-level information are fused, and the final result is generated. To demonstrate the effectiveness of our proposed method, we present extensive experimental results in the thesis. / Thesis / Master of Applied Science (MASc)

This master thesis is the study of multi-task learning to train a neural network to segment medical images and predict anatomical landmarks. The paper shows the results from experiments using medical landmarks in order to attempt to help the network learn the important organ structures quicker. The results found in this study are inconclusive and rather than showing the efficiency of the multi-task framework for learning, they tell a story of the importance of choosing the tasks and dataset wisely. The study also reflects and depicts the general difficulties and pitfalls of performing a project of this type.

In recent years, pictures from handheld devices such as smartphones have been increasingly utilized as a documentation tool by medical practitioners not trained to take professional photographs. Similarly to the other types of image modalities, the images should be taken in a way to capture the vital information in the region of interest. Nevertheless, image capturing cannot always be done as desired, so images may exhibit different blur types at the region of interest. Having blurry images does not serve medical purposes, therefore, the patients might have to schedule a second appointment several days later to retake the images. A solution to this problem is to create an algorithm which immediately after capturing an image determines if it is medically useful and notifies the user of the result. The algorithm needs to perform the analysis at a reasonable speed, and at best, with a limited number of operations to make the calculations directly in the smartphone device. A large number of medical images must be available to create such an algorithm. Medical images are difficult to acquire, and it is specifically difficult to acquire blurry images since they are usually deleted. The main objective of this thesis is to determine the medical usefulness of images taken with smartphone cameras, using both machine learning and handcrafted algorithms, with a low number of floating point operations and a high performance. Seven different algorithms (one hand-crafted and six machine learned) are created and compared regarding both number of floating point operations and performance. Fast Walsh-Hadamard transforms are the basis of the hand-crafted algorithm. The employed machine learning algorithms are both based on common convolutional neural networks (MobileNetV3 and ResNet50) and on our own designs. The issue with the low number of medical images acquired is solved by training the machine learning models on a synthetic dataset, where the non-medically useful images are generated by applying blur on the medically useful images. These models do, however, undergo evaluation using a real dataset, containing medically useful images as well as non-medically useful images. Our results implicate that a real-time determination of the medical usefulness of images is possible on handheld devices, since our machine learned model DeepLAD-Net reaches the highest accuracy with 42 · 106 floating point operations. In terms of accuracy, MobileNetV3-large is the second best model with31 times as many floating point operations as our best model.

Sammanfattning   Hjärt- och kärlsjukdommar är den ledande dödsorsaken i världen. En av det vanligaste hjärt- och kärlsjukdomarna är åderförkalkning. Sjukdomen kännetecknas av förhårdning samt plackansamling i kärl och bidrar till stroke och hjärtinfarkt. Information om kärlväggens styvhet kan spela en viktig roll vid diagnostiseringen av bland annat åderförkalkning. Skjuvvågselastografi (SWE) är en noninvasiv ultraljudsbaserad metod som idag används för att mäta elasticitet och styvhet av större mjuka vävnader som lever- och bröstvävnad. Dock används inte metoden inom kärlapplikationer, då få genomgående studier har utförts på SWE för kärl. Målet med projektet är att automatisera kvantifieringen av skjuvvågshastigheten för SWE och undersöka hur automatiseringens förmåga och begränsningar beror av automatiseringsinställningar. Med verktyg erhållna från CBH (skolan för kemi, bioteknologi och hälsa) skapades ett MATLAB-program med denna förmåga. Programmet applicerades på två fantommodeller. Automatiseringsinställningarna påverkade automatiseringen av dessa modeller olika, vilket innebar att generella optimala inställningar inte kunde finnas. Optimala inställningar beror på vad automatiseringen skall undersöka. / Medicinsk avbildning

Medicinsk bildanalys är både tidskonsumerade och kräver expertis. I den härrapporten vidareutvecklas en 2.5D version av faltningsnätverket U-Net anpassadför automatiserad njuresegmentering. Faltningsnätverk har tidigare visatliknande prestation som experter. Träningsdata för nätverket anpassades genomatt manuellt segmentera MR-bilder av njurar. 2.5D U-Net nätverket tränades med64 st njursegmenteringar från tidigare arbete. Volymanalys på nätverketssegmenterings förslag av 38.000 patienter visade den mängden segmenteradevoxlar som inte tillhörde njurarna var 0,35 %. Efter tillägg av 56 st av vårasegmenteringar minskade det till 0.11 %, en reduktion av cirka 68 %. Det är enstor förbättring av nätverket och ett viktigt steg mot tillämpning avautomatiserad segmentering. / Medical image analysis is both time consuming and requires expertise. In thisreport, a 2.5D version of the U-net convolution network adapted for automatedkidney segmentation is further developed. Convolution neural networks havepreviously shown expert level performance in image segmentation. Training datafor the network was created by manually segmenting MRI images of kidneys.The 2.5D U-Net network was trained with 64 kidney segmentations fromprevious work. Volume analysis on the network’s kidney segmentation proposalsof 38,000 patients showed that the ammount of segmented voxels that are notpart of the kidneys was 0.35%. After the addition of 56 of our segmentations, itdecreased to just 0.11%, indicating a reduction of about 68%. This is a majorimprovement of the network and an important step towards the development ofpractical applications of automated segmentation.

Nonalcoholic liver steatosis (NALS) is a condition where fat infiltrates the tissue of the liver and accumulates in droplets. While not a dangerous condition on its own, if left for long enough it can develop into conditions which could cause serious and potentially permanent damage to the liver. One of the primary approaches for preventing NALS from progressing is through changes in diet and lifestyle. However, explaining to a patient the impact of such a change can be difficult, which hampers motivation in many instances. Digital twin technology can provide simulations of what will happen to the body after a lifestyle change, but the output data is very abstract and can thus be a challenge to convey properly to a patient. In this project I investigate a digital data visualization approach where a photo of a liver sample is morphed to showcase liver fat droplets shrinking as a result of a changed lifestyle, as simulated by the digital twin. The approach uses a simple image morphing algorithm that pulls pixel intensity values from regions designated by a morph field and composites a newimage from the updated values. By selectively choosing regions of interest to pull pixels towards or away from, with a ramping cutoff in morph field strength, it is possible to designate certain regions in the image to be morphed. The program is capable of generating time series of increasingly morphed images in both greyscale and truecolour, and it can save the time series as an animated .GIF file, with linear interpolation between the morphed images in the time series.

In this thesis, algorithms for image registration and segmentation are developed to locate and identify DU penetrators and associated metal projectile debris on or near the surface at the US DoD firing ranges and proving grounds. The proposed registration algorithm supports fusing the LWIR and visible images. Control points are indentified by area-base detection and followed by eliminating outliers. Associated with bilinear interpolation, the gravity centers of control points are used to estimate the transformation parameters. The segmentation with a statistical detector is developed to improve the fusion result. The power spectrum density is invoked to extract and identify the image properties, and the probability of each pixel classified as target further the decision. The final result is consistent with the true vision and carries distinguished target information. The combination of registration and segmentation approaches can effectively orientate and investigate the target area.

No description available.

No description available.