Comparison between Radiographic (2-dimensional and 3-dimensional) and Histologic Findings of Periapical Lesions Treated with Apical Surgery.

INTRODUCTION The aim of this study was to evaluate the concordance of 2- and 3-dimensional radiography and histopathology in the diagnosis of periapical lesions. METHODS Patients were consecutively enrolled in this study provided that preoperative periapical radiography (PR) and cone-beam computed tomographic imaging of the tooth to be treated with apical surgery were performed. The periapical lesional tissue was histologically analyzed by 2 blinded examiners. The final histologic diagnosis was compared with the radiographic assessments of 4 blinded observers. The initial study material included 62 teeth in the same number of patients. RESULTS Four lesions had to be excluded during processing, resulting in a final number of 58 evaluated cases (31 women and 27 men, mean age = 55 years). The final histologic diagnosis of the periapical lesions included 55 granulomas (94.8%) and 3 cysts (5.2%). Histologic analysis of the tissue samples from the apical lesions exhibited an almost perfect agreement between the 2 experienced investigators with an overall agreement of 94.83% (kappa = 0.8011). Radiographic assessment overestimated cysts by 28.4% (cone-beam computed tomographic imaging) and 20.7% (periapical radiography), respectively. Comparing the correlation of the radiographic diagnosis of 4 observers with the final histologic diagnosis, 2-dimensional (kappa = 0.104) and 3-dimensional imaging (kappa = 0.111) provided only minimum agreement. CONCLUSIONS To establish a final diagnosis of an apical radiolucency, the tissue specimen should be evaluated histologically and specified as a granuloma (with/without epithelium) or a cyst. Analysis of 2-dimensional and 3-dimensional radiographic images alike results only in a tentative diagnosis that should be confirmed with biopsy.

Alveolar bone regeneration in response to local application of calcitriol in vitamin D deficient rats

AIM Vitamin D deficiency is considered to diminish bone regeneration. Yet, raising the serum levels takes months. A topic application of the active vitamin D metabolite, calcitriol, may be an effective approach. Thus, it becomes important to know the effect of vitamin D deficiency and local application on alveolar bone regeneration. MATERIAL AND METHODS Sixty rats were divided into three groups; two vitamin depletion groups and a control group. Identical single defects (2 mm diameter) were created in the maxilla and mandible treated with calcitriol soaked collagen in one deficiency group while in the other two groups not. Histomorphometric analysis and micro CTs were performed after 1 and 3 weeks. Serum levels of 25(OH)D3 and PTH were determined. RESULTS Bone formation rate significantly increased within the observation period in all groups. Bone regeneration was higher in the maxilla than in the mandible. However, bone regeneration was lower in the control group compared to vitamin depletion groups, with no significant effects by local administration of calcitriol (micro CT mandible p = 0.003, maxilla p < 0.001; histomorphometry maxilla p = 0.035, mandible p = 0.18). CONCLUSION Vitamin D deficiency not necessarily impairs bone regeneration in the rat jaw and a single local calcitriol application does not enhance healing.

Virtopsy: postmortem imaging of the human heart in situ using MSCT and MRI.

The rapid further development of computed tomography (CT) and magnetic resonance imaging (MRI) induced the idea to use these techniques for postmortem documentation of forensic findings. Until now, only a few institutes of forensic medicine have acquired experience in postmortem cross-sectional imaging. Protocols, image interpretation and visualization have to be adapted to the postmortem conditions. Especially, postmortem alterations, such as putrefaction and livores, different temperature of the corpse and the loss of the circulation are a challenge for the imaging process and interpretation. Advantages of postmortem imaging are the higher exposure and resolution available in CT when there is no concern for biologic effects of ionizing radiation, and the lack of cardiac motion artifacts during scanning. CT and MRI may become useful tools for postmortem documentation in forensic medicine. In Bern, 80 human corpses underwent postmortem imaging by CT and MRI prior to traditional autopsy until the month of August 2003. Here, we describe the imaging appearance of postmortem alterations--internal livores, putrefaction, postmortem clotting--and distinguish them from the forensic findings of the heart, such as calcification, endocarditis, myocardial infarction, myocardial scarring, injury and other morphological alterations.

Loss of bone strength in HLA-B27 transgenic rats is characterized by a high bone turnover and is mainly osteoclast-driven

Objective Although osteopenia is frequent in spondyloarthritis (SpA), the underlying cellular mechanisms and association with other symptoms are poorly understood. This study aimed to characterize bone loss during disease progression, determine cellular alterations, and assess the contribution of inflammatory bowel disease (IBD) to bone loss in HLA-B27 transgenic rats. Methods Bones of 2-, 6-, and 12-month-old non-transgenic, disease-free HLA-B7 and disease-associated HLA-B27 transgenic rats were examined using peripheral quantitative computed tomography, μCT, and nanoindentation. Cellular characteristics were determined by histomorphometry and ex vivo cultures. The impact of IBD was determined using [21-3 x 283-2]F1 rats, which develop arthritis and spondylitis, but not IBD. Results HLA-B27 transgenic rats continuously lost bone mass with increasing age and had impaired bone material properties, leading to a 3-fold decrease in bone strength at 12 months of age. Bone turnover was increased in HLA-B27 transgenic rats, as evidenced by a 3-fold increase in bone formation and a 6-fold increase in bone resorption parameters. Enhanced osteoclastic markers were associated with a larger number of precursors in the bone marrow and a stronger osteoclastogenic response to RANKL or TNFα. Further, IBD-free [21-3 x 283-2]F1 rats also displayed decreased total and trabecular bone density. Conclusions HLA-B27 transgenic rats lose an increasing amount of bone density and strength with progressing age, which is primarily mediated via increased bone remodeling in favor of bone resorption. Moreover, IBD and bone loss seem to be independent features of SpA in HLA-B27 transgenic rats.

Radiation dose reduction in postoperative computed position control of cochlear implant electrodes in lambs - An experimental study

OBJECTIVE Cochlear implants (CI) are standard treatment for prelingually deafened children and postlingually deafened adults. Computed tomography (CT) is the standard method for postoperative imaging of the electrode position. CT scans accurately reflect electrode depth and position, which is essential prior to use. However, routine CT examinations expose patients to radiation, which is especially problematic in children. We examined whether new CT protocols could reduce radiation doses while preserving diagnostic accuracy. METHODS To investigate whether electrode position can be assessed by low-dose CT protocols, a cadaveric lamb model was used because the inner ear morphology is similar to humans. The scans were performed at various volumetric CT dose-indexes CTDIvol)/kV combinations. For each constant CTDIvol the tube voltage was varied (i.e., 80, 100, 120 and 140kV). This procedure was repeated at different CTDIvol values (21mGy, 11mGy, 5.5mGy, 2.8mGy and 1.8mGy). To keep the CTDIvol constant at different tube voltages, the tube current values were adjusted. Independent evaluations of the images were performed by two experienced and blinded neuroradiologists. The criteria diagnostic usefulness, image quality and artifacts (scaled 1-4) were assessed in 14 cochlear-implanted cadaveric lamb heads with variable tube voltages. RESULTS Results showed that the standard CT dose could be substantially reduced without sacrificing diagnostic accuracy of electrode position. The assessment of the CI electrode position was feasible in almost all cases up to a CTDIvol of 2-3mGy. The number of artifacts did not increase for images within this dose range as compared to higher dosages. The extent of the artifacts caused by the implanted metal-containing CI electrode does not depend on the radiation dose and is not perceptibly influenced by changes in the tube voltage. Summarizing the evaluation of the CI electrode position is possible even at a very low radiation dose. CONCLUSIONS CT imaging of the temporal bone for postoperative electrode position control of the CI is possible with a very low and significantly radiation dose. The tube current-time product and voltage can be reduced by 50% without increasing artifacts. Low-dose postoperative CT scans are sufficient for localizing the CI electrode.

Effect of boundary conditions on yield properties of human femoral trabecular bone

Trabecular bone plays an important mechanical role in bone fractures and implant stability. Homogenized nonlinear finite element (FE) analysis of whole bones can deliver improved fracture risk and implant loosening assessment. Such simulations require the knowledge of mechanical properties such as an appropriate yield behavior and criterion for trabecular bone. Identification of a complete yield surface is extremely difficult experimentally but can be achieved in silico by using micro-FE analysis on cubical trabecular volume elements. Nevertheless, the influence of the boundary conditions (BCs), which are applied to such volume elements, on the obtained yield properties remains unknown. Therefore, this study compared homogenized yield properties along 17 load cases of 126 human femoral trabecular cubic specimens computed with classical kinematic uniform BCs (KUBCs) and a new set of mixed uniform BCs, namely periodicity-compatible mixed uniform BCs (PMUBCs). In stress space, PMUBCs lead to 7–72 % lower yield stresses compared to KUBCs. The yield surfaces obtained with both KUBCs and PMUBCs demonstrate a pressure-sensitive ellipsoidal shape. A volume fraction and fabric-based quadric yield function successfully fitted the yield surfaces of both BCs with a correlation coefficient R2≥0.93. As expected, yield strains show only a weak dependency on bone volume fraction and fabric. The role of the two BCs in homogenized FE analysis of whole bones will need to be investigated and validated with experimental results at the whole bone level in future studies.

Dissecting abdominal aortic aneurysm in Angiotensin II-infused mice: the importance of imaging

INTRODUCTION Since the initial publication in 2000, Angiotensin II-infused mice have become one of the most popular models to study abdominal aortic aneurysm in a pre-clinical setting. We recently used phase contrast X-ray based computed tomography to demonstrate that these animals develop an apparent luminal dilatation and an intramural hematoma, both related to mural ruptures in the tunica media in the vicinity of suprarenal side branches. AIMS The aim of this narrative review was to provide an extensive overview of small animal applicable techniques that have provided relevant insight into the pathogenesis and morphology of dissecting AAA in mice, and to relate findings from these techniques to each other and to our recent PCXTM-based results. Combining insights from recent and consolidated publications we aimed to enhance our understanding of dissecting AAA morphology and anatomy. RESULTS AND CONCLUSION We analyzed in vivo and ex vivo images of aortas obtained from macroscopic anatomy, histology, high-frequency ultrasound, contrast-enhanced micro-CT, micro-MRI and PCXTM. We demonstrate how in almost all publications the aorta has been subdivided into a part in which an intact lumen lies adjacent to a remodeled wall/hematoma, and a part in which elastic lamellae are ruptured and the lumen appears to be dilated. We show how the novel paradigm fits within the existing one, and how 3D images can explain and connect previously published 2D structures. We conclude that PCXTM-based findings are in line with previous results, and all evidence points towards the fact that dissecting AAAs in Angiotensin II-infused mice are actually caused by ruptures of the tunica media in the immediate vicinity of small side branches.

Comparison between magnetic resonance imaging and computed tomography of the lung in patients with cystic fibrosis with regard to clinical, laboratory, and pulmonary functional parameters

OBJECTIVE To evaluate whether magnetic resonance imaging (MRI) is effective as computed tomography (CT) in determining morphologic and functional pulmonary changes in patients with cystic fibrosis (CF) in association with multiple clinical parameters. MATERIALS AND METHODS Institutional review board approval and patient written informed consent were obtained. In this prospective study, 30 patients with CF (17 men and 13 women; mean (SD) age, 30.2 (9.2) years; range, 19-52 years) were included. Chest CT was acquired by unenhanced low-dose technique for clinical purposes. Lung MRI (1.5 T) comprised T2- and T1-weighted sequences before and after the application of 0.1-mmol·kg gadobutrol, also considering lung perfusion imaging. All CT and MR images were visually evaluated by using 2 different scoring systems: the modified Helbich and the Eichinger scores. Signal intensity of the peribronchial walls and detected mucus on T2-weighted images as well as signal enhancement of the peribronchial walls on contrast-enhanced T1-weighted sequences were additionally assessed on MRI. For the clinical evaluation, the pulmonary exacerbation rate, laboratory, and pulmonary functional parameters were determined. RESULTS The overall modified Helbich CT score had a mean (SD) of 15.3 (4.8) (range, 3-21) and median of 16.0 (interquartile range [IQR], 6.3). The overall modified Helbich MR score showed slightly, not significantly, lower values (Wilcoxon rank sum test and Student t test; P > 0.05): mean (SD) of 14.3 (4.7) (range, 3-20) and median of 15.0 (IQR, 7.3). Without assessment of perfusion, the overall Eichinger score resulted in the following values for CT vs MR examinations: mean (SD), 20.3 (7.2) (range, 4-31); and median, 21.0 (IQR, 9.5) vs mean (SD), 19.5 (7.1) (range, 4-33); and median, 20.0 (IQR, 9.0). All differences between CT and MR examinations were not significant (Wilcoxon rank sum tests and Student t tests; P > 0.05). In general, the correlations of the CT scores (overall and different imaging parameters) to the clinical parameters were slightly higher compared to the MRI scores. However, if all additional MRI parameters were integrated into the scoring systems, the correlations reached the values of the CT scores. The overall image quality was significantly higher for the CT examinations compared to the MRI sequences. CONCLUSIONS One major diagnostic benefit of lung MRI in CF is the possible acquisition of several different morphologic and functional imaging features without the use of any radiation exposure. Lung MRI shows reliable associations with CT and clinical parameters, which suggests its implementation in CF for routine diagnosis, which would be particularly important in follow-up imaging over the long term.

Pulmonary lymphangioleiomyomatosis: analysis of disease manifestation by region-based quantification of lung parenchyma

PURPOSE Lymphangioleiomyomatosis (LAM) is characterized by proliferation of smooth muscle tissue that causes bronchial obstruction and secondary cystic destruction of lung parenchyma. The aim of this study was to evaluate the typical distribution of cystic defects in LAM with quantitative volumetric chest computed tomography (CT). MATERIALS AND METHODS CT examinations of 20 patients with confirmed LAM were evaluated with region-based quantification of lung parenchyma. Additionally, 10 consecutive patients were identified who had recently undergone CT imaging of the lung at our institution, in which no pathologies of the lung were found, to serve as a control group. Each lung was divided into three regions (upper, middle and lower thirds) with identical number of slices. In addition, we defined a "peel" and "core" of the lung comprising the 2 cm subpleural space and the remaining inner lung area. Computerized detection of lung volume and relative emphysema was performed with the PULMO 3D software (v3.42, Fraunhofer MEVIS, Bremen, Germany). This software package enables the quantification of emphysematous lung parenchyma by calculating the pixel index, which is defined as the ratio of lung voxels with a density <-950HU to the total number of voxels in the lung. RESULTS Cystic changes accounted for 0.1-39.1% of the total lung volume in patients with LAM. Disease manifestation in the central lung was significantly higher than in peripheral areas (peel median: 15.1%, core median: 20.5%; p=0.001). Lower thirds of lung parenchyma showed significantly less cystic changes than upper and middle lung areas combined (lower third: median 13.4, upper and middle thirds: median 19.0, p=0.001). CONCLUSION The distribution of cystic lesions in LAM is significantly more pronounced in the central lung compared to peripheral areas. There is a significant predominance of cystic changes in apical and intermediate lung zones compared to the lung bases.

Metal Artifact Reduction in Pelvic Computed Tomography With Hip Prostheses: Comparison of Virtual Monoenergetic Extrapolations From Dual-Energy Computed Tomography and an Iterative Metal Artifact Reduction Algorithm in a Phantom Study.

OBJECTIVE The aim of this study was to directly compare metal artifact reduction (MAR) of virtual monoenergetic extrapolations (VMEs) from dual-energy computed tomography (CT) with iterative MAR (iMAR) from single energy in pelvic CT with hip prostheses. MATERIALS AND METHODS A human pelvis phantom with unilateral or bilateral metal inserts of different material (steel and titanium) was scanned with third-generation dual-source CT using single (120 kVp) and dual-energy (100/150 kVp) at similar radiation dose (CT dose index, 7.15 mGy). Three image series for each phantom configuration were reconstructed: uncorrected, VME, and iMAR. Two independent, blinded radiologists assessed image quality quantitatively (noise and attenuation) and subjectively (5-point Likert scale). Intraclass correlation coefficients (ICCs) and Cohen κ were calculated to evaluate interreader agreements. Repeated measures analysis of variance and Friedman test were used to compare quantitative and qualitative image quality. Post hoc testing was performed using a corrected (Bonferroni) P < 0.017. RESULTS Agreements between readers were high for noise (all, ICC ≥ 0.975) and attenuation (all, ICC ≥ 0.986); agreements for qualitative assessment were good to perfect (all, κ ≥ 0.678). Compared with uncorrected images, VME showed significant noise reduction in the phantom with titanium only (P < 0.017), and iMAR showed significantly lower noise in all regions and phantom configurations (all, P < 0.017). In all phantom configurations, deviations of attenuation were smallest in images reconstructed with iMAR. For VME, there was a tendency toward higher subjective image quality in phantoms with titanium compared with uncorrected images, however, without reaching statistical significance (P > 0.017). Subjective image quality was rated significantly higher for images reconstructed with iMAR than for uncorrected images in all phantom configurations (all, P < 0.017). CONCLUSIONS Iterative MAR showed better MAR capabilities than VME in settings with bilateral hip prosthesis or unilateral steel prosthesis. In settings with unilateral hip prosthesis made of titanium, VME and iMAR performed similarly well.

The virtual patient in dental medicine

OBJECTIVES The aim of this Short Communication was to present a workflow for the superimposition of intraoral scan (IOS), cone-beam computed tomography (CBCT), and extraoral face scan (EOS) creating a 3D virtual dental patient. MATERIAL AND METHODS As a proof-of-principle, full arch IOS, preoperative CBCT, and mimic EOS were taken and superimposed to a unique 3D data pool. The connecting link between the different files was to detect existing teeth as constant landmarks in all three data sets. RESULTS This novel application technique successfully demonstrated the feasibility of building a craniofacial virtual model by image fusion of IOS, CBCT, and EOS under 3D static conditions. CONCLUSIONS The presented application is the first approach that realized the fusion of intraoral and facial surfaces combined with skeletal anatomy imaging. This novel 3D superimposition technique allowed the simulation of treatment planning, the exploration of the patients' expectations, and the implementation as an effective communication tool. The next step will be the development of a real-time 4D virtual patient in motion.

Effects of Externally Applied, Cyclical, Low Magnitude Forces on Orthodontic Tooth Movement in Rats

Orthodontic tooth movement requires external orthodontic forces to be converted to cellular signals that result in the coordinated removal of bone on one side of the tooth (compression side) by osteoclasts, and the formation of new bone by osteoblasts on the other side (tension side). The length of orthodontic treatment can take several years, leading to problems of caries, periodontal disease, root resorption, and patient dissatisfaction. It appears that the velocity of tooth movement is largely dependent on the rate of alveolar bone remodeling. Pharmacological approaches to increase the rate of tooth movement are limited due to patient discomfort, severe root resorption, and drug-induced side effects. Recently, externally applied, cyclical, low magnitude forces (CLMF) have been shown to cause an increase in the bone mineral density of long bones, and in the growth of craniofacial structures in a variety of animal models. In addition, CLMF is well tolerated by the patient and produces no known adverse effects. However, its application in orthodontic tooth movement has not been specifically determined. Since factors that increase alveolar bone remodeling enhance the rate of orthodontic tooth movement, we hypothesized that externally applied, cyclical, low magnitude forces (CLMF) will increase the rate of orthodontic tooth movement. In order to test this hypothesis we used an in vivo rat orthodontic tooth movement model. Our specific aims were: Specific Aim 1: To develop an in vivo rat model for tooth movement. We developed a tooth movement model based upon two established rodent models (Ren and Yoshimatsu et al, See Figure 1.). The amount of variation of tooth movement in rats exposed to 25-60 g of mesial force activated viii from the first molar to the incisor for 4 weeks was calculated. Specific Aim 2: To determine the frequency dose response of externally applied, cyclical, low magnitude forces (CLMF) for maximal tooth movement and osteoclast numbers. Our working hypothesis for this aim was that the amount of tooth movement would be dose dependent on the frequency of application of the CLMF. In order to test this working hypothesis, we varied the frequency of the CLMF from 30, 60, 100, and 200 Hz, 0.4N, two times per week, for 10 minutes for 4 weeks, and measured the amount of tooth movement. We also looked at the number of osteoclasts for the different frequencies; we hypothesized an increase in osteoclasts for the dose respnse of different frequencies. Specific Aim 3: To determine the effects of externally applied, cyclical, low magnitude forces (CLMF) on PDL proliferation. Our working hypothesis for this aim was that PDL proliferation would increase with CLMF. In order to test this hypothesis we compared CLMF (30 Hz, 0.4N, two times per week, for 10 minutes for 4 weeks) performed on the left side (experimental side), to the non-CLMF side, on the right (control side). This was an experimental study with 24 rats in total. The experimental group contained fifteen (15) rats in total, and they all received a spring plus a different frequency of CLMF. Three (3) received a spring and CLMF at 30 Hz, 0.4N for 10 minutes. Six (6) received a spring and CLMF at 60 Hz, 0.4N for 10 minutes. Three (3) received a spring and CLMF at 100 Hz, 0.4N for 10 minutes. Three (3) received a spring and CLMF at 200 Hz, 0.4N for 10 minutes. The control group contained six (6) rats, and received only a spring. An additional ix three (3) rats received CLMF (30 Hz, 0.4N, two times per week, for 10 minutes for 4 weeks) only, with no spring, and were used only for histological purposes. Rats were subjected to the application of orthodontic force from their maxillary left first molar to their left central incisor. In addition some of the rats received externally applied, cyclical, low magnitude force (CLMF) on their maxillary left first molar. micro-CT was used to measure the amount of orthodontic tooth movement. The distance between the maxillary first and second molars, at the most mesial point of the second molar and the most distal point of the first molar (1M-2M distance) were used to evaluate the distance of tooth movement. Immunohistochemistry was performed with TRAP staining and BrdU quantification. Externally applied, cyclical, low magnitude forces (CLMF) do appear to have an effect on the rate, while not significant, of orthodontic tooth movement in rats. It appears that lower CLMF decreases the rate of tooth movement, while higher CLMF increases the rate of tooth movement. Future studies with larger sample sizes are needed to clarify this issue. CLMF does not appear to affect the proliferation in PDL cells, and has no effect on the number of osteoclasts.

Reefs shift from net accretion to net erosion along a natural environmental gradient

Coral reefs persist in an accretion-erosion balance and ocean acidification resulting from anthropogenic CO2 emissions threatens to shift this balance in favor of net reef erosion. Corals and calcifying algae, largely responsible for reef accretion, are vulnerable to environmental changes associated with ocean acidification, but the direct effects of lower pH on reef erosion has received less attention, particularly in the context of known drivers of bioerosion and natural variability. This study examines the balance between reef accretion and erosion along a well-characterized natural environmental gradient in Kane'ohe Bay, Hawai'i using experimental blocks of coral skeleton. Comparing before and after micro-computed tomography (µCT) scans to quantify net accretion and erosion, we show that, at the small spatial scale of this study (tens of meters), pH was a better predictor of the accretion-erosion balance than environmental drivers suggested by prior studies, including resource availability, temperature, distance from shore, or depth. In addition, this study highlights the fine-scale variation of pH in coastal systems and the importance of microhabitat variation for reef accretion and erosion processes. We demonstrate significant changes in both the mean and variance of pH on the order of meters, providing a local perspective on global increases in pCO2. Our findings suggest that increases in reef erosion, combined with expected decreases in calcification, will accelerate the shift of coral reefs to an erosion-dominated system in a high-CO2 world. This shift will make reefs increasingly susceptible to storm damage and sea-level rise, threatening the maintenance of the ecosystem services that coral reefs provide.

Optimización de una energía mediante cortes de grafos. Segmentación de imágenes

La segmentación de imágenes puede plantearse como un problema de minimización de una energía discreta. Nos enfrentamos así a una doble cuestión: definir una energía cuyo mínimo proporcione la segmentación buscada y, una vez definida la energía, encontrar un mínimo absoluto de la misma. La primera parte de esta tesis aborda el segundo problema, y la segunda parte, en un contexto más aplicado, el primero. Las técnicas de minimización basadas en cortes de grafos permiten obtener el mínimo de una energía discreta en tiempo polinomial mediante algoritmos de tipo min-cut/max-flow. Sin embargo, estas técnicas solo pueden aplicarse a energías que son representabas por grafos. Un importante reto es estudiar qué energías son representabas así como encontrar un grafo que las represente, lo que equivale a encontrar una función gadget con variables adicionales. En la primera parte de este trabajo se estudian propiedades de las funciones gadgets que permiten acotar superiormente el número de variables adicionales. Además se caracterizan las energías con cuatro variables que son representabas, definiendo gadgets con dos variables adicionales. En la segunda parte, más práctica, se aborda el problema de segmentación de imágenes médicas, base en muchas ocasiones para la diagnosis y el seguimiento de terapias. La segmentación multi-atlas es una potente técnica de segmentación automática de imágenes médicas, con tres aspectos importantes a destacar: el tipo de registro entre los atlas y la imagen objetivo, la selección de atlas y el método de fusión de etiquetas. Este último punto puede formularse como un problema de minimización de una energía. A este respecto introducimos dos nuevas energías representables. La primera, de orden dos, se utiliza en la segmentación en hígado y fondo de imágenes abdominales obtenidas mediante tomografía axial computarizada. La segunda, de orden superior, se utiliza en la segmentación en hipocampos y fondo de imágenes cerebrales obtenidas mediante resonancia magnética. ABSTRACT The image segmentation can be described as the problem of minimizing a discrete energy. We face two problems: first, to define an energy whose minimum provides the desired segmentation and, second, once the energy is defined we must find its global minimum. The first part of this thesis addresses the second problem, and the second part, in a more applied context, the first problem. Minimization techniques based on graph cuts find the minimum of a discrete energy in polynomial time via min-cut/max-flow algorithms. Nevertheless, these techniques can only be applied to graph-representable energies. An important challenge is to study which energies are graph-representable and to construct graphs which represent these energies. This is the same as finding a gadget function with additional variables. In the first part there are studied the properties of gadget functions which allow the number of additional variables to be bounded from above. Moreover, the graph-representable energies with four variables are characterised and gadgets with two additional variables are defined for these. The second part addresses the application of these ideas to medical image segmentation. This is often the first step in computer-assisted diagnosis and monitoring therapy. Multiatlas segmentation is a powerful automatic segmentation technique for medical images, with three important aspects that are highlighted here: the registration between the atlas and the target image, the atlas selection, and the label fusion method. We formulate the label fusion method as a minimization problem and we introduce two new graph-representable energies. The first is a second order energy and it is used for the segmentation of the liver in computed tomography (CT) images. The second energy is a higher order energy and it is used for the segmentation of the hippocampus in magnetic resonance images (MRI).

Image processing methods for human brain connectivity analysis from in-vivo diffusion MRI

The structural connectivity of the brain is considered to encode species-wise and subject-wise patterns that will unlock large areas of understanding of the human brain. Currently, diffusion MRI of the living brain enables to map the microstructure of tissue, allowing to track the pathways of fiber bundles connecting the cortical regions across the brain. These bundles are summarized in a network representation called connectome that is analyzed using graph theory. The extraction of the connectome from diffusion MRI requires a large processing flow including image enhancement, reconstruction, segmentation, registration, diffusion tracking, etc. Although a concerted effort has been devoted to the definition of standard pipelines for the connectome extraction, it is still crucial to define quality assessment protocols of these workflows. The definition of quality control protocols is hindered by the complexity of the pipelines under test and the absolute lack of gold-standards for diffusion MRI data. Here we characterize the impact on structural connectivity workflows of the geometrical deformation typically shown by diffusion MRI data due to the inhomogeneity of magnetic susceptibility across the imaged object. We propose an evaluation framework to compare the existing methodologies to correct for these artifacts including whole-brain realistic phantoms. Additionally, we design and implement an image segmentation and registration method to avoid performing the correction task and to enable processing in the native space of diffusion data. We release PySDCev, an evaluation framework for the quality control of connectivity pipelines, specialized in the study of susceptibility-derived distortions. In this context, we propose Diffantom, a whole-brain phantom that provides a solution to the lack of gold-standard data. The three correction methodologies under comparison performed reasonably, and it is difficult to determine which method is more advisable. We demonstrate that susceptibility-derived correction is necessary to increase the sensitivity of connectivity pipelines, at the cost of specificity. Finally, with the registration and segmentation tool called regseg we demonstrate how the problem of susceptibility-derived distortion can be overcome allowing data to be used in their original coordinates. This is crucial to increase the sensitivity of the whole pipeline without any loss in specificity.