Chia Chi Teng

A medical image archive solution in the cloud

Growing long-term cost of managing an onsite medical imaging archive has been a subject which the health care industry struggles with. Based on the current trend, it is estimated that over 1 billion diagnostic imaging procedures will be performed in the United States during year 2014, generating about 100 Petabytes of data. The high volume of medical images is leading to scalability and maintenance issues with healthcare providers onsite picture archiving and communication system (PACS) and network. Cloud computing promises lower cost, high scalability, availability and disaster recoverability which can be a natural solution some of the problems we faced for long-term medical image archive. A prototype system was implemented to study such as solution on one of the industry leading cloud computing platform, Microsoft Windows Azure. It includes a Digital Imaging and Communications in Medicine (DICOM) server which handles standard store/query/retrieve requests; a DICOM image indexer that parses the metadata and store them in a SQL Azure database; and a web UI for searching and viewing archived images based on patient and image attributes. The comprehensive tools and functionality of Windows Azure made it an ideal platform to develop and deploy this kind of service oriented applications.

Mobile Application Development: Essential New Directions for IT

As mobile devices have become more powerful and pervasive, mobile computing has become more important. As the market share of mobile operating systems steadily grows and more IT applications are developed and deployed on mobile devices, this will become a larger and increasing part of professional IT practice. Consequently IT students should gain experience creating and deploying mobile applications. We designed a project for a junior level operating systems course and asked 35 students to develop an application using one of the following leading mobile device software development kits: Apple iPhone, Microsoft Windows Mobile and Google Android. Students were free to select a platform and define their own applications. While iPhone has approximately 50% of the mobile OS market share, only 8 students selected iPhone as their development platform of choice, 15 selected Android, another 12 selected Windows Mobile. A post-project survey was administered to the students to evaluate the process of choosing a platform, hardware and software features used in their applications, and the overall learning experience. This provided several conclusions about preferences, future applications and learning. The feedback was overwhelmingly positive and will help us improve future learning experiences.

Changes in frontal morphology after single-stage open posterior-middle vault expansion for sagittal craniosynostosis.

Background: There is controversy regarding whether the frontal bossing associated with sagittal synostosis requires direct surgical correction or spontaneously remodels after isolated posterior cranial expansion. The authors retrospectively measured changes in frontal bone morphology in patients with isolated sagittal synostosis 2 years after open posterior and midvault cranial expansion and compared these changes with those occurring in age-comparable healthy control groups. Methods: Forty-three patients age 1 year or younger (mean, 6 months) with sagittal synostosis underwent computed tomography scan digital analysis immediately after and 2 years after posterior–middle cranial vault expansion. Quantitative angular and linear measures were taken along the midsagittal and axial planes to capture both aspects of frontal bossing. The change in values over the 2 years were compared with healthy controls with normal computed tomography scans taken to rule out head trauma. Results: All measures indicative of frontal bossing decreased significantly from the time of posterior–middle vault expansion to 2 years postoperatively. Whereas the majority of patients at time of the operation had frontal bossing measures greater than two standard deviations outside the age-comparable control mean, almost all patients were within two standard deviations of the norm 2 years later. Lateral forehead bossing and anterior cranial growth was greater the older the patient was at the time of the operation, suggesting that the more time that passed before the operation, the more compensatory anterior fossa growth occurred. Central forehead position relative to the anterior cranial base was greatest in the younger patients at the time of operation, suggesting that a central forehead bulge was an early compensatory response to premature sagittal fusion. Conclusions: As a group, patients with sagittal synostosis start to normalize their forehead morphology within 2 years if an isolated posterior operation is performed at 1 year of age or younger, and this occurs by a combination of restriction of growth and reduction relative to patients without synostosis. This protocol decreases the risks of intraoperative positioning, forehead contour deformities, and two-stage operations. CLINICAL QUESTION/LEVEL OF EVIDENCE: Therapeutic, III.

Head and neck lymph node region delineation with image registration

BackgroundThe success of radiation therapy depends critically on accurately delineating the target volume, which is the region of known or suspected disease in a patient. Methods that can compute a contour set defining a target volume on a set of patient images will contribute greatly to the success of radiation therapy and dramatically reduce the workload of radiation oncologists, who currently draw the target by hand on the images using simple computer drawing tools. The most challenging part of this process is to estimate where there is microscopic spread of disease.MethodsGiven a set of reference CT images with gold standard lymph node regions drawn by the experts, we are proposing an image registration based method that could automatically contour the cervical lymph code levels for patients receiving radiation therapy. We are also proposing a method that could help us identify the reference models which could potentially produce the best results.ResultsThe computer generated lymph node regions are evaluated quantitatively and qualitatively.ConclusionsAlthough not conforming to clinical criteria, the results suggest the technique has promise.

Head and neck lymph node region delineation using a hybrid image registration method

The success of radiation therapy depends critically on accurately delineating the target volume, which is the region of known or suspected disease in a patient. Methods that can compute a contour set defining a target volume on a set of patients biomedical images will contribute greatly to the success of radiation therapy and drastically reduce the workload of radiation oncologists, who currently often draw the targets by hand on images using simple computer drawing tools. We are developing methods for automatically selecting and adapting standardized regions of tumor spread based on the location of lymph node regions in a standard or reference case, using image registration techniques. Previously available image registration techniques (deformable transformations computed using mutual information ) appear promising and can be supplemented by utilizing landmark correspondences in the optimization process to come closer to achieving a clinically acceptable match

Managing DICOM image metadata with desktop operating systems native user interface

Picture Archiving and Communication System (PACS) is commonly used in the hospital environment as the tool to manage radiological images which has standardized on the DICOM format. PACS usually consists of dedicated high performance server computers to provide functionalities of acquisition, storage, retrieval, editing (metadata), distribution and presentation. As the personal computers become more powerful, average desktop workstations can process large amount of data with performance comparable to the high cost dedicated systems. Recent desktop operating systems (OS) such as Microsoft Windows Vista have built-in indexing and search capability integrated with the graphic user interface (GUI) to allow fast retrieval and customized presentation for digital media and documents. Integrating DICOM image management into desktop OS GUI can reduce the cost of software acquisition, maintenance, and user training. We developed DicomProp as a Windows plug-in to provide an easy to use alternative for managing DICOM images on desktop computers.

HEAD AND NECK CANCER PATIENT SIMILARITY BASED ON ANATOMICAL STRUCTURAL GEOMETRY

As we develop radiation treatment planning systems for head and neck cancer patients, there is a need to identify reference patients whose anatomical structures share similar features. By finding previously treated patients or prototypical models with the most similar anatomy in the head and neck region, our prototype system can do a better job of identifying lymph node regions based on known regions predefined for similar patients. This can potentially expedite the radiation treatment planning process which currently depends on oncologists manually delineating lymph node regions for each patient slice by slice on their CT scans. Our methodology uses the identification of landmark anatomical structures from the CT scan of the cancer patient and the computation of a Hausdorff distance between these and the corresponding structures of the reference scans. The combination of these structural correspondence features and a set of global shape features provide a distance metric that can rapidly find the most similar reference patient case and thus efficiently and accurately generate the desired lymph node regions for a given test case.

Automatic Segmentation of Neck CT Images

In this era of cross-sectional imaging, it is useful to think of the neck in terms of adjacent anatomical spaces separated by fascial layers extended from the skull base to the thoracic inlet. Although these layers are not usually visible in CT or MR images, their locations can be inferred by knowing their relationships to various anatomical structures that are visible in cross-sectional images. Identifying these anatomical structures in the neck region can facilitate applications such as automated diagnosis by finding abnormality, or computer assisted radiation therapy planning by inferring cervical lymph node regions from these anatomical structures. We are proposing a system that can automatically segment the neck region from CT images by identifying the cranial and caudal bounding slices and then locating and labeling various anatomical structures in the region

Pediatric cranial defect surface analysis for craniosynostosis postoperation CT images

Craniosynostosis is a congenital disease which consists of premature fusion of one or more cranial sutures, resulting in an abnormal head shape. Patients are usually treated by cranial vault expansion surgery to minimize the potential for brain damage. Full thickness cranial defects result from the expansion surgery, with the size directly proportional to the degree of expansion. The growing cranial skeleton has a unique regenerative capacity to heal small defects; however, when this regenerative capacity is exceeded, the defect is classed as one of critical size and requires surgical treatment to restore protection to the underlying brain. Although what constitutes a critical cranial defect is well known in animal models, it is not as clear for pediatric human skulls. The purpose of this study is to investigate a method that can effectively quantify healing of the pediatric cranial defect surface after cranial vault expansion surgery for craniosynostosis.

Automatic identification and tracking of retraction fibers in time-lapse microscopy

Digital image processing in the field of time-lapse microscopy and biological research has been an interesting research area to provide statistical data of cellular dynamics to the field of cell biology. Digital image processing enables rapid and consistent quantification of qualitative observations. The image processing model examined here provides a tool for the biologists to identify structures called retraction fibers (RF) that are formed during Epithelial-Mesenchyme Transition (EMT), an important developmental process which also occurs during cancer metastasis. Quantifying RF formation is an important tool for biologists studying cellular regulation of EMT. The dynamic EMT process is captured using time-lapse microscopy. We use computer vision algorithms to detect and track the RF in image sequences of cells undergoing EMT to generate statistical information such as the number of RF formed during a window, lifetime of the RF, and their geometric dimension. This information can in turn be used by the biologist to quantitatively measure the extent of EMT under different test conditions.