Lung Pattern Classification for Interstitial Lung Diseases Using a Deep Convolutional Neural Network

Automated tissue characterization is one of the most crucial components of a computer aided diagnosis (CAD) system for interstitial lung diseases (ILDs). Although much research has been conducted in this field, the problem remains challenging. Deep learning techniques have recently achieved impressive results in a variety of computer vision problems, raising expectations that they might be applied in other domains, such as medical image analysis. In this paper, we propose and evaluate a convolutional neural network (CNN), designed for the classification of ILD patterns. The proposed network consists of 5 convolutional layers with 2×2 kernels and LeakyReLU activations, followed by average pooling with size equal to the size of the final feature maps and three dense layers. The last dense layer has 7 outputs, equivalent to the classes considered: healthy, ground glass opacity (GGO), micronodules, consolidation, reticulation, honeycombing and a combination of GGO/reticulation. To train and evaluate the CNN, we used a dataset of 14696 image patches, derived by 120 CT scans from different scanners and hospitals. To the best of our knowledge, this is the first deep CNN designed for the specific problem. A comparative analysis proved the effectiveness of the proposed CNN against previous methods in a challenging dataset. The classification performance (~85.5%) demonstrated the potential of CNNs in analyzing lung patterns. Future work includes, extending the CNN to three-dimensional data provided by CT volume scans and integrating the proposed method into a CAD system that aims to provide differential diagnosis for ILDs as a supportive tool for radiologists.

Correlation between alveolar ventilation and electrical properties of lung parenchyma.

One key problem in modern medical imaging is linking measured data and actual physiological quantities. In this article we derive such a link between the electrical bioimpedance of lung parenchyma, which can be measured by electrical impedance tomography (EIT), and the magnitude of regional ventilation, a key to understanding lung mechanics and developing novel protective ventilation strategies. Two rat-derived three-dimensional alveolar microstructures obtained from synchrotron-based x-ray tomography are each exposed to a constant potential difference for different states of ventilation in a finite element simulation. While the alveolar wall volume remains constant during stretch, the enclosed air volume varies, similar to the lung volume during ventilation. The enclosed air, serving as insulator in the alveolar ensemble, determines the resulting current and accordingly local tissue bioimpedance. From this we can derive a relationship between lung tissue bioimpedance and regional alveolar ventilation. The derived relationship shows a linear dependence between air content and tissue impedance and matches clinical data determined from a ventilated patient at the bedside.

E.A.O. guidelines for the use of diagnostic imaging in implant dentistry 2011. A consensus workshop organized by the European Association for Osseointegration at the Medical University of Warsaw

Diagnostics imaging is an essential component of patient selection and treatment planning in oral rehabilitation by means of osseointegrated implants. In 2002, the EAO produced and published guidelines on the use of diagnostic imaging in implant dentistry. Since that time, there have been significant developments in both the application of cone beam computed tomography as well as in the range of surgical and prosthetic applications that can potentially benefit from its use. However, medical exposure to ionizing radiation must always be justified and result in a net benefit to the patient. The as low a dose as is reasonably achievable principle must also be applied taking into account any alternative techniques that might achieve the same objectives. This paper reports on current EAO recommendations arising from a consensus meeting held at the Medical University of Warsaw (2011) to update these guidelines. Radiological considerations are detailed, including justification and optimization, with a special emphasis on the obligations that arise for those who prescribe or undertake such investigations. The paper pays special attention to clinical indications and radiographic diagnostic considerations as well as to future developments and trends.

Narrow band imaging (NBI) during medical thoracoscopy: first impressions

Background This is the first ever evaluation of narrow band imaging (NBI), an innovative endoscopic imaging procedure, for the visualisation of pleural processes. Methods The pleural cavity was examined in 26 patients with pleural effusions using both white light and narrow band imaging during thoracoscopy under local anaesthesia. Results In the great majority of the patients narrow band imaging depicted the blood vessels more clearly than white light, but failed to reveal any differences in number, shape or size. Only in a single case with pleura thickened by chronic inflammation and metastatic spread of lung cancer did narrow band imaging show vessels that were not detectable under white light. Conclusion It is not yet possible to assess to what extent the evidence provided by NBI is superior to that achieved with white light. Further studies are required, particularly in the early stages of pleural processes.

Incidence, clinical presentation and imaging findings of cavernous malformations of the CNS. A twenty-year experience

Up to 88% of cavernous malformations (CMs) of the central nervous system can become symptomatic and cause long-term disability. The aim of this study was to document the characteristics of CMs in the catchment area of our institution.

Pacemakers and magnetic resonance imaging: Current status and survey in Switzerland

Electromagnetic fields arising from magnetic resonance imaging (MRI) can cause various clinically relevant functional disturbances in patients with cardiac pacemakers. Consequently, an implanted pacemaker is generally considered a contraindication for an MRI scan. With approximately 60 million MRI scans performed worldwide per year, MRI may be indicated for an estimated majority of pacemaker patients during the lifetime of their pacemakers. The availability of MR conditional pacemakers with CE labelling is of particular advantage since they allow the safe use of pacemakers in MRI. In this article the current state of knowledge on pacemakers and MR imaging is discussed. We present the results of a survey conducted among Swiss radiologists to assess current practice in patients with pacemakers.

Intraobserver and interobserver agreement for results of low-field magnetic resonance imaging in dogs with and without clinical signs of disk-associated wobbler syndrome

OBJECTIVE: To determine interobserver and intraobserver agreement for results of low-field magnetic resonance imaging (MRI) in dogs with and without disk-associated wobbler syndrome (DAWS). DESIGN: Validation study. ANIMALS: 21 dogs with and 23 dogs without clinical signs of DAWS. PROCEDURES: For each dog, MRI of the cervical vertebral column was performed. The MRI studies were presented in a randomized sequence to 4 board-certified radiologists blinded to clinical status. Observers assessed degree of disk degeneration, disk-associated and dorsal compression, alterations in intraspinal signal intensity (ISI), vertebral body abnormalities, and new bone formation and categorized each study as originating from a clinically affected or clinically normal dog. Interobserver agreement was calculated for 44 initial measurements for each observer. Intraobserver agreement was calculated for 11 replicate measurements for each observer. RESULTS: There was good interobserver agreement for ratings of disk degeneration and vertebral body abnormalities and moderate interobserver agreement for ratings of disk-associated compression, dorsal compression, alterations in ISI, new bone formation, and suspected clinical status. There was very good intraobserver agreement for ratings of disk degeneration, disk-associated compression, alterations in ISI, vertebral body abnormalities, and suspected clinical status. There was good intraobserver agreement for ratings of dorsal compression and new bone formation. Two of 21 clinically affected dogs were erroneously categorized as clinically normal, and 4 of 23 clinically normal dogs were erroneously categorized as clinically affected. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggested that variability exists among observers with regard to results of MRI in dogs with DAWS and that MRI could lead to false-positive and false-negative assessments.

A survey of MRI-based medical image analysis for brain tumor studies

MRI-based medical image analysis for brain tumor studies is gaining attention in recent times due to an increased need for efficient and objective evaluation of large amounts of data. While the pioneering approaches applying automated methods for the analysis of brain tumor images date back almost two decades, the current methods are becoming more mature and coming closer to routine clinical application. This review aims to provide a comprehensive overview by giving a brief introduction to brain tumors and imaging of brain tumors first. Then, we review the state of the art in segmentation, registration and modeling related to tumor-bearing brain images with a focus on gliomas. The objective in the segmentation is outlining the tumor including its sub-compartments and surrounding tissues, while the main challenge in registration and modeling is the handling of morphological changes caused by the tumor. The qualities of different approaches are discussed with a focus on methods that can be applied on standard clinical imaging protocols. Finally, a critical assessment of the current state is performed and future developments and trends are addressed, giving special attention to recent developments in radiological tumor assessment guidelines.

X-ray phase contrast imaging of metallized and non-metallized biological samples

Imaging of biological samples has been performed with a variety of techniques for example electromagnetic waves, electrons, neutrons, ultrasound and X-rays. Also conventional X-ray imaging represents the basis of medical diagnostic imaging, it remains of limited use in this application because it is based solely on the differential absorption of X-rays by tissues. Coherent and bright photon beams, such as those produced by third-generation synchrotron X-ray sources, provide further information on subtle X-ray phase changes at matter interfaces. This complements conventional X-ray absorption by edge enhancement phenomena. Thus, phase contrast imaging has the potential to improve the detection of structures on images by detecting those structures that are invisible with X-ray absorption imaging. Images of a weakly absorbing nylon fibre were recorded in in-line holography geometry using a high resolution low-noise CCD camera at the ESRF in Grenoble. The method was also applied to improve image contrast for images of biological tissues. This paper presents phase contrast microradiographs of vascular tree casts and images of a housefly. These reveal very fine structures, that remain invisible with conventional absorption contrast only.

Calibration of a portal imaging device for high-precision dosimetry: a Monte Carlo study

Today electronic portal imaging devices (EPID's) are used primarily to verify patient positioning. They have, however, also the potential as 2D-dosimeters and could be used as such for transit dosimetry or dose reconstruction. It has been proven that such devices, especially liquid filled ionization chambers, have a stable dose response relationship which can be described in terms of the physical properties of the EPID and the pulsed linac radiation. For absolute dosimetry however, an accurate method of calibration to an absolute dose is needed. In this work, we concentrate on calibration against dose in a homogeneous water phantom. Using a Monte Carlo model of the detector we calculated dose spread kernels in units of absolute dose per incident energy fluence and compared them to calculated dose spread kernels in water at different depths. The energy of the incident pencil beams varied between 0.5 and 18 MeV. At the depth of dose maximum in water for a 6 MV beam (1.5 cm) and for a 18 MV beam (3.0 cm) we observed large absolute differences between water and detector dose above an incident energy of 4 MeV but only small relative differences in the most frequent energy range of the beam energy spectra. It is shown that for a 6 MV beam the absolute reference dose measured at 1.5 cm water depth differs from the absolute detector dose by 3.8%. At depth 1.2 cm in water, however, the relative dose differences are almost constant between 2 and 6 MeV. The effects of changes in the energy spectrum of the beam on the dose responses in water and in the detector are also investigated. We show that differences larger than 2% can occur for different beam qualities of the incident photon beam behind water slabs of different thicknesses. It is therefore concluded that for high-precision dosimetry such effects have to be taken into account. Nevertheless, the precise information about the dose response of the detector provided in this Monte Carlo study forms the basis of extracting directly the basic radiometric quantities photon fluence and photon energy fluence from the detector's signal using a deconvolution algorithm. The results are therefore promising for future application in absolute transit dosimetry and absolute dose reconstruction.

Magnetic resonance imaging of the hip at 3 Tesla: clinical value in femoroacetabular impingement of the hip and current concepts

Magnetic resonance imaging (MRI) is the most promising noninvasive modality for hip joint evaluation, but it has limitations in diagnosing cartilage lesion and acetabular labrum changes, especially in early stages. This is significant due to superior outcome results of surgery intervention in hip dysplasia or femoroacetabular impingement in patients not exceeding early degeneration. This emphasizes the need for accurate and reproducible methods in evaluating cartilage structure. In this article, we discuss the impact of the most recent technological advance in MRI, namely the advantage of 3-T imaging, on diagnostic imaging of the hip. Limitations of standard imaging techniques are shown with emphasis on femoroacetabular impingement. Clinical imaging examples and biochemical techniques are presented that need to be further evaluated.

Imaging structural abnormalities in the hip joint: instability and impingement as a cause of osteoarthritis

Osteoarthritis is thought to be caused by a combination of intrinsic vulnerabilities of the joint, such as anatomic shape and alignment, and environmental factors, such as body weight, injury, and overuse. It has been postulated that much of osteoarthritis is due to anatomic deformities. Advances in surgical techniques such as the periacetabular osteotomy, safe surgical dislocation of the hip, and hip arthroscopy have provided us with effective and safe tools to correct these anatomical problems. The limiting factor in treatment outcome in many mechanically compromised hips is the degree of cartilage damage which has occurred prior to treatment. In this regard, the role of imaging, utilizing plain radiographs in conjunction with magnetic resonance imaging, is becoming vitally important for the detection of these anatomic deformities and pre-radiographic arthritis. In this article, we will outline the plain radiographic features of hip deformities that can cause instability or impingement. Additionally, we will illustrate the use of MRI imaging to detect subtle anatomic abnormalities, as well as the use of biochemical imaging techniques such as dGEMRIC to guide clinical decision making.

Advanced morphological and biochemical magnetic resonance imaging of cartilage repair procedures in the knee joint at 3 Tesla

Morphological and biochemical magnetic resonance imaging (MRI) is due to high field MR systems, advanced coil technology, and sophisticated sequence protocols capable of visualizing articular cartilage in vivo with high resolution in clinical applicable scan time. Several conventional two-dimensional (2D) and three-dimensional (3D) approaches show changes in cartilage structure. Furthermore newer isotropic 3D sequences show great promise in improving cartilage imaging and additionally in diagnosing surrounding pathologies within the knee joint. Functional MR approaches are additionally able to provide a specific measure of the composition of cartilage. Cartilage physiology and ultra-structure can be determined, changes in cartilage macromolecules can be detected, and cartilage repair tissue can thus be assessed and potentially differentiated. In cartilage defects and following nonsurgical and surgical cartilage repair, morphological MRI provides the basis for diagnosis and follow-up evaluation, whereas biochemical MRI provides a deeper insight into the composition of cartilage and cartilage repair tissue. A combination of both, together with clinical evaluation, may represent a desirable multimodal approach in the future, also available in routine clinical use.

Cartilage T2 assessment at 3-T MR imaging: in vivo differentiation of normal hyaline cartilage from reparative tissue after two cartilage repair procedures--initial experience

PURPOSE: To prospectively compare cartilage T2 values after microfracture therapy (MFX) and matrix-associated autologous chondrocyte transplantation (MACT) repair procedures. MATERIALS AND METHODS: The study had institutional review board approval by the ethics committee of the Medical University of Vienna; informed consent was obtained. Twenty patients who underwent MFX or MACT (10 in each group) were enrolled. For comparability, patients of each group were matched by mean age (MFX, 40.0 years +/- 15.4 [standard deviation]; MACT, 41.0 years +/- 8.9) and postoperative interval (MFX, 28.6 months +/- 5.2; MACT, 27.4 months +/- 13.1). Magnetic resonance (MR) imaging was performed with a 3-T MR imager, and T2 maps were calculated from a multiecho spin-echo measurement. Global, as well as zonal, quantitative T2 values were calculated within the cartilage repair area and within cartilage sites determined to be morphologically normal articular cartilage. Additionally, with consideration of the zonal organization, global regions of interest were subdivided into deep and superficial areas. Differences between cartilage sites and groups were calculated by using a three-way analysis of variance. RESULTS: Quantitative T2 assessment of normal native hyaline cartilage showed similar results for all patients and a significant trend of increasing T2 values from deep to superficial zones (P < .05). In cartilage repair areas after MFX, global mean T2 was significantly reduced (P < .05), whereas after MACT, mean T2 was not reduced (P > or = .05). For zonal variation, repair tissue after MFX showed no significant trend between different depths (P > or = .05), in contrast to repair tissue after MACT, in which a significant increase from deep to superficial zones (P < .05) could be observed. CONCLUSION: Quantitative T2 mapping seems to reflect differences in repair tissues formed after two surgical cartilage repair procedures. (c) RSNA, 2008.