Search

What are the ways that industrial heritage places are transformed, both considering their tangible and intangible values as well as their potential to achieve urban inclusivity, sustainability, circularity as well as public engagement? The present book is one of the outcomes of the inaugural program of NIT Urban Heritage Lab. Following an introduction on the course, its scope and aims, it showcases the four proposals that the participants of the Industrial Heritage for Sustainable Cities Course developed for the transformation of the Unkapanı Flour Mill. The proposals are original, thought-provoking and inspirational, while at the same time based on solid multi-disciplinary, collaborative work. So, they are perfect demonstrations of what NIT Urban Heritage Lab tries to achieve.

The Neolithic way of life was first established in Northwest Anatolia before the middle of the 7th millennium BC. The recently excavated sites of Barcın Höyük and Bahçelievler have yielded archaeological evidence for the earliest Neolithic levels in the region and provide new archaeobotanical datasets. To compare different adaptations to the changes brought on by the Neolithization processes, we studied 348 archaeobotanical samples from Phases VIe and VId1 at Barcın and 63 samples from the contemporaneous levels, Phase 6 and Phase 5, at Bahçelievler. The economic plants include hulled and naked six-row barley, einkorn, emmer, bread/hard wheat, small-sized naked wheat, lentil, bitter vetch, pea, chickpea, flax, hazelnut, bramble, and pistacia. Our analyses show small but significant differences between the sites in the selected economic plant ranges, among the cereals, pulses as well as gathered plants.

Pit clusters have not received systematic attention from archaeologists working in Anatolia and the Near East, in contrast to many other parts of the world. This paper presents a case study of Middle Chalcolithic pits from the prehistoric site of Barcın Höyük in northwestern Anatolia to show how pit clusters, interpreted as underground grain silos, can inform us about ancient food economies, social organization, and inhabited landscapes. It is argued that the silos at Barcın Höyük were used by small family-sized groups to store surplus grain. Dated to the first quarter of the 5th millennium B.C., the silos present evidence for agricultural productivity during a period that has largely eluded archaeological investigation in western Anatolia.

Landscape archaeological research has been undertaken on the mound of Barcin Höyük in northwest Anatolia, Turkey. It is the oldest Neolithic site in the region, making it of particular interest in regard to the spread of farming from the regions of origin (southeast and central Anatolia) to northwest Anatolia. This study shows that the site was founded approximately 8550 a cal. BP on a natural elevation in a generally wet environment, at the edge of a retreating lake. The site was subject to environmental shifts in younger periods, with a phase of arid conditions and erosion at least prior to the Bronze Age, succeeded by a phase of more humid conditions and rising lake levels mostly in the last two millennia, and finally a return to arid conditions and declining lake levels. Considering the Neolithic environmental conditions, the local environment of Barcin Höyük resembled the local environment of Çatalhöyük, a central Anatolian site. This suggests that despite the general differences between the regions, the earliest sedentary population of northwest Anatolia favoured local conditions similar to those of central Anatolia, while younger sites in the aforementioned region diverge to slightly different localities.

The American Heart Association proposed a 17-segment model for the segmentation of the left ventricle together with a mapping from each segment to a supplying coronary artery. This proposal is based on population averages. Several studies have confirmed the inaccuracy of this mapping due to large anatomical variations of the coronary arteries among individuals. Several proposals have been made for a different mapping between the 17 segments and the coronary arteries.

Pulmonary embolism (PE) is a common life-threatening disorder for which an early diagnosis is desirable. We propose a new system for the automatic detection of PE in contrast-enhanced CT images. The system consists of candidate detection, feature computation and classification. Candidate detection focusses on the inclusion of PE - even complete occlusions - and the exclusion of false detections, such as tissue and parenchymal diseases. Feature computation does not only focus on the intensity, shape and size of an embolus, but also on locations and the shape of the pulmonary vascular tree. Several classifiers have been tested and the results show that the performance is optimized by using a bagged tree classifier with two features based on the shape of a blood vessel and the distance to the vessel boundary. The system was trained on 38 CT data sets. Evaluation on 19 other data sets showed that the system generalizes well. The sensitivity of our system on the evaluation data is 63% at 4.9 false positives per data set, which allowed the radiologist to improve the number of detected PE by 22%.

—Pulmonary embolism (PE) is a common life-threatening disorder for which an early diagnosis is desirable. We propose a new system for the automatic detection of PE in contrast-enhanced CT images. The system consists of candidate detection, feature computation and classification. Candidate detection focusses on the inclusion of PE—even complete occlusions—and the exclusion of false detections, such as tissue and parenchymal diseases. Feature computation does not only focus on the intensity, shape and size of an embolus, but also on locations and the shape of the pulmonary vascular tree. Several classifiers have been tested and the results show that the performance is optimized by using a bagged tree classifier with two features based on the shape of a blood vessel and the distance to the vessel boundary. The system was trained on 38 CT data sets. Evaluation on 19 other data sets showed that the system generalizes well. The sensitivity of our system on the evaluation data is 63% at 4.9 false positives per data set, which allowed the radiologist to improve the number of detected PE by 22%.

Purpose: To evaluate the use of lesion features derived from T2-weighted images in computer-aided classification of benign and malignant masses in dynamic contrast-enhanced (DCE) magnetic resonance imaging studies of the breast. Methods and Materials: Fat-suppressed DCE-T1-weighted and non-fat-suppressed T2-weighted images of 50 patients containing 64 masses (28 benign and 36 malignant) were acquired. All masses were segmented in the DCE-T1-weighted images using a semi-automatic method. Twenty-two morphological and kinetic DCE-T1-weighted features proposed in literature were computed for each mass. Additionally, 14 quantitative T2-weighted features were defined and evaluated. These features are mainly based on the intensity characteristics within the lesion or within the enclosing shell of normal tissue. The intensity values were normalized with respect to the mean intensity of different reference regions to account for the qualitative and inhomogeneous nature of MR images. Feature selection was performed to determine the best set consisting of not more than five DCE-T1-weighted features and the best set combining not more than a total of five DCE-T1-weighted and T2-weighted features. Subsequently, a linear classification function was estimated by Fisher discriminant analysis. Results: In a leave-one-out evaluation, including T2-weighted features improved the classification performance from an area-under-the-ROC-curve of 0.94 to 0.99, yet a test for statistical significance failed with p=0.064. The selected T2-weighted features reflect the existence of bright structures in the surrounding tissue (caused by e.g. edema) and dark areas within the lesion. Conclusion: The obtained results encourage the use of quantitative information from T2-weighted images in CAD for breast MRI.

The bull's eye plot is a commonly used schematic for the visualization of quantitative late enhancement cardiac MRI data. It gives an intuitive overview of the viability of the entire left ventricular myocardium in a single diagram. However, common implementations do not provide a continuous transition between slices and provide poor or no information about the exact location and transmurality of non-viable tissue.

The full-width at half-max (FWHM) criterion is often used for both manual and automatic quantification of the vessel diameter in medical images. The FWHM criterion is easy to understand and it can be implemented with low computational cost. However, it is well known that the FWHM criterion can give an over- and underestimation of the vessel diameter. In this paper, we propose a simple and original method to create an unbiased estimation of the vessel diameter based on the FWHM criterion and we compared the robustness to noise of several edge detectors. The quantitative results of our experiments show that the proposed method is more accurate and precise than other (more complex) edge detectors, even for small vessels.

no abstract

Gaussian derivatives are often used as differential operators to analyze the structure in images. In this paper, we will analyze the accuracy and computational cost of the most common implementations for differentiation and interpolation of Gaussian-blurred multi-dimensional data. We show that – for the computation of multiple Gaussian derivatives – the method based on B-splines obtains a higher accuracy than the truncated Gaussian at equal computational cost.

Purpose: To demonstrate that inter-patient differences in the spreading of the contrast agent throughout the blood, described by the arterial input function (AIF), should be considered in the classification of contrast-agent uptake curves in the tissue of interest, e.g., a suspicious lesion in the breast. In the application of heuristic shape modeling (Kuhl et al. 1999, three-time-point (3TP) by Weinstein et al. 1999), the AIF is not extracted from the DCE-MR image series and, therefore, not taken into account.Methods and Materials: A two-compartment model (extended Kety) with fixed pharmacokinetic parameters is used to simulate the tissue-response curves for different AIFs. The shape of these curves is classified into benign, suspicious or malignant by means of the 3TP method.Results: While AIFs are known to differ in a wide range, our simulations indicate that already small changes of the AIF considerably alter the shape of the response curve. These changes may even lead to different curve classifications, although the simulated response curves relate to ’tissue’ with fixed pharmacokinetic properties. Conclusion: The shape of contrast-agent uptake curves expressed by simulated tissue with fixed pharmacokinetic properties can get classified differently in different patients owing to inter-patient variations of the AIF. Evaluation of the influence of variation in AIFs in real patient data is work in progress. Validation of our observation with real patient data might suggest that deconvolution with patient-specific AIFs, as it is done in pharmacokinetic modeling, improves reliability of tissue classification derived from the shape of contrast-agent uptake curves.

DCE-MR breast imaging is becoming an important modality as a problem-solving tool and in screening of high-risk patients. In DCE-MR a T1-weighted time series is recorded during the uptake of a contrast agent (CA: Gd-DTPA) that enables the assessment of voxel-specific dynamic tissue properties. Tumor malignancy often goes hand in hand with high vascularity and high vessel wall permeability because of angiogenesis; i.e. a large number of new vessels is formed at such a rate that their walls are not properly constructed and therefore highly permeable. These 'malignant' tissue characteristics affect the shape of the CA-uptake curve. The goal is to investigate the use of a pharmacokinetic two-compartment model (extended Kety [1]) in the context of dynamic analysis of CA-uptake curves as a contribution to automatic detection and characterization of breast cancer. The equations: ve*(dCe(t)/dt)=Ktrans(Cp(t)-Ce(t)) and Ct(t)=vp*Cp(t)+ve*Ce(t)describe the model. Ktrans = volume transfer constant [min-1], vc = cell fraction [-], ve = extravascular, extracellular fraction [-] (EES), vp = blood plasma fraction [-], vbc = blood cell fraction [-]. Ct / Cp / Ce = tissue / blood plasma / EES concentration CA [mM]. However, the models that are currently dominating the clinical practice are 'heuristic shape models' such as Kuhl's model [2] and the Three-Time-Point (3TP) model [3]. These models classify the shape of the CA-uptake curves into three categories: (1) persistent uptake (benign), (2) plateau uptake in the intermediate to late phase (suspicious), and (3) washout in the late to intermediate phase (malignant). This type of classification is based on clinical experience and statistics supported by biopsy-proven data. One of our aims is to uncover the limitations of these shape models and to investigate if two-compartment modeling can lead to a better, i.e. physiology-based, tissue classification.

Short axis cine cardiac magnetic resonance is used for measuring ventricular blood volumes. Papillary muscles and trabeculae should be excluded from these volumes, despite being included by delineations that allow measurements of wall thickness. We have developed and evaluated three methods that automatically exclude the papillary muscles and trabeculae from the blood volume by classifying the interior grey values of the endocardial contours to estimate corrected areas for summation with Simpson's Rule. Classification is performed using thresholding, hard clipping and error-function clipping, respectively. An evaluation of our methods on 69 clinical data sets showed that our methods provide significantly different results for the end diastolic volume, the end systolic volume, the stroke volume and the ejection fraction, with respect to conventional methods. Furthermore, the difference between our methods was found to be not significant.

This paper discusses two applications of probing dense volumetric data for MR orthopedics and dynamic contrast enhanced MRI mammography. In order not to reduce the context information and to extract the essential part of the data, we apply Profile Flags. A Profile Flag is a 3D glyph for probing and annotating the volumetric data. The first application area deals with visualization of T2 profiles for interactive inspection of knee cartilage and detection of lesions. In the second application, we present the usability the Profile Flags for measuring of time-signal profiles for a set of time-dependent MR volumes. Several extensions of the basic Profile Flag concept are described in detail and discussed. These extensions include selection of a set of profiles based on spatial as well as curve differences, automatic positioning of the Profile Flags, and adaptation for probing of time-varying volumetric data. Additionally, we include the evaluation of the used methods by our medical partners.

Mutual information based nonrigid registration of medical images is a popular approach. The coordinate mapping that relates the two images is found in an iterative optimisation procedure. In every iteration a computationally expensive evaluation of the mutual information's derivative is required. In this work two acceleration strategies are compared. The first technique aims at reducing the number of iterations, and, consequently, the number of derivative evaluations. The second technique reduces the computational costs per iteration by employing stochastic approximations of the derivatives. The performance of both methods is tested on an artificial registration problem, where the ground truth is known, and on a clinical problem involving low-dose CT scans and large deformations. The experiments show that the stochastic approximation approach is superior in terms of speed and robustness. However, more accurate solutions are obtained with the first technique. © Springer-Verlag Berlin Heidelberg 2006.

The quality of direct volume rendering is strongly influenced by the transfer function. The purpose of the transfer function is to highlight aspects of the volume data that are of interest to the user. Material boundaries are often regarded as interesting, since they reveal the shapes of objects and their spatial relations.We use a new transfer function domain that facilitates the selection of material boundaries in the volumetric data sets. Our domain enables to distinguish between both sides of the boundary. By reducing the partial volume effect on the boundary, we are able to easily assign optical properties to objects and their boundaries.

The current surgical intervention criterion for abdominal aortic aneurysm is based on the maximum transverse diameter of the aorta. Recent research advances indicate that a better rupture predictor may be derived from the wall stress, which can be computed with the finite element method. An essential prerequisite for this modelling is an accurate description of the geometry of the aneurysm. We developed an automatic method to derive the dynamic patient-specific aneurysm geometry from non-contrast enhanced MRA balanced turbo field images. The slices of our 2D-scanned volumes are registered onto 3D-scanned volumes to restore spatial coherence. The resulting images are noise-filtered and the enddiastolic volume is segmented with an active objects technique (deformable models). The resulting geometrical model is propagated to the remaining phases using the correlation between grey value profiles on the surface as an external force for the active object. From our segmentations we derived tetrahedral finite element meshes which were used as the input for finite element wall stress simulations.