In the past, we have presented an imaging-informatics based eFolder system for managing and analyzing imaging and lesion data of multiple sclerosis (MS) patients, which allows for data storage, data analysis, and data mining in clinical and research settings. The system integrates the patient’s clinical data with imaging studies and a computer-aided detection (CAD) algorithm for quantifying MS lesion volume, lesion contour, locations, and sizes in brain MRI studies. For compliance with IHE integration protocols, long-term storage in PACS, and data query and display in a DICOM compliant clinical setting, CAD results need to be converted into DICOM-Structured Report (SR) format. Open-source dcmtk and customized XML templates are used to convert quantitative MS CAD results from MATLAB to DICOM-SR format. A web-based GUI based on our existing web-accessible DICOM object (WADO) image viewer has been designed to display the CAD results from generated SR files. The GUI is able to parse DICOM-SR files and extract SR document data, then display lesion volume, location, and brain matter volume along with the referenced DICOM imaging study. In addition, the GUI supports lesion contour overlay, which matches a detected MS lesion with its corresponding DICOM-SR data when a user selects either the lesion or the data. The methodology of converting CAD data in native MATLAB format to DICOM-SR and displaying the tabulated DICOM-SR along with the patient’s clinical information, and relevant study images in the GUI will be demonstrated. The developed SR conversion model and GUI support aim to further demonstrate how to incorporate CAD post-processing components in a PACS and imaging informatics-based environment.

We introduce a novel remote volume rendering pipeline for medical visualization targeted for mHealth (mobile health) applications. The necessity of such a pipeline stems from the large size of the medical imaging data produced by current CT and MRI scanners with respect to the complexity of the volumetric rendering algorithms. For example, the resolution of typical CT Angiography (CTA) data easily reaches 512^3 voxels and can exceed 6 gigabytes in size by spanning over the time domain while capturing a beating heart. This explosion in data size makes data transfers to mobile devices challenging, and even when the transfer problem is resolved the rendering performance of the device still remains a bottleneck. To deal with this issue, we propose a thin-client architecture, where the entirety of the data resides on a remote server where the image is rendered and then streamed to the client mobile device. We utilize the display and interaction capabilities of the mobile device, while performing interactive volume rendering on a server capable of handling large datasets. Specifically, upon user interaction the volume is rendered on the server and encoded into an H.264 video stream. H.264 is ubiquitously hardware accelerated, resulting in faster compression and lower power requirements. The choice of low-latency CPU- and GPU-based encoders is particularly important in enabling the interactive nature of our system. We demonstrate a prototype of our framework using various medical datasets on commodity tablet devices.

Most medical images are archived and transmitted using the DICOM format. The DICOM information model combines image pixel data and associated metadata into a single object. It is not possible to access the metadata separately from the pixel data. However, there are important use cases that only need access to metadata, and the DICOM format increases the running time of those use cases. Tag morphing is an example of one such use case. Tag or attribute morphing includes insertion, deletion, or modification of one or more of the metadata attributes in a study. It is typically used for order reconciliation on study acquisition or to localize the Issuer of Patient ID and the Patient ID attributes when data from one Medical Record Number (MRN) domain is transferred to or displayed in a different domain. This work uses the Multi-Series DICOM (MSD) format to reduce the time required for tag morphing. The MSD format separates metadata from pixel data, and at the same time eliminates duplicate attributes. MSD stores studies using two files rather than in many single frame files typical of DICOM. The first file contains the de-duplicated study metadata, and the second contains pixel data and other bulkdata. A set of experiments were performed where metadata updates were applied to a set of DICOM studies stored in both the traditional Single Frame DICOM (SFD) format and the MSD format. The time required to perform the updates was recorded for each format. The results show that tag morphing is, on average, more than eight times faster in MSD format.

Many projects to evaluate or conduct research in medical imaging require the large-scale collection of images (both unprocessed and processed) and associated data. This demand has led us to design and implement a flexible oncology image repository, which prospectively collects images and data from multiple sites throughout the UK. This Oncology Medical Image Database (OMI-DB) has been created to support research involving medical imaging and contains unprocessed and processed medical images, associated annotations and data, and where applicable expert-determined ground truths describing features of interest. The process of collection, annotation and storage is almost fully automated and is extremely adaptable, allowing for quick and easy expansion to disparate imaging sites and situations. Initially the database was developed as part of a large research project in digital mammography (OPTIMAM). Hence the initial focus has been digital mammography; as a result, much of the work described will focus on this field. However, the OMI -DB has been designed to support multiple modalities and is extensible and expandable to store any associated data with full anonymisation. Currently, the majority of associated data is made up of radiological, clinical and pathological annotations extracted from the UK’s National Breast Screening System (NBSS). In addition to the data, software and systems have been created to allow expert radiologists to annotate the images with interesting clinical features and provide descriptors of these features. The data from OMI-DB has been used in several observer studies and more are planned. To date we have collected 34,104 2D mammography images from 2,623 individuals.

The treatment process of tumor patients is supported by different stand-alone ePR and clinical decision support (CDS) systems. We developed a concept for the integration of a specialized ePR for head and neck tumor treatment and a DICOM-RT based CDS system for radiation therapy in order to improve the clinical workflow and therapy outcome. A communication interface for the exchange of information that is only available in the respective other system will be realized. This information can then be used for further assistance and clinical decision support functions. In the first specific scenario radiation therapy related information such as radiation dose or tumor size are transmitted from the CDS to the ePR to extend the information base. This information can then be used for the automatic creation of clinical documents or retrospective clinical trial studies. The second specific use case is the transmission of follow-up information from the ePR to the CDS system. The CDS system uses the current patient’s anatomy and planned radiation dose distribution for the selection of other patients that already received radiation therapy. Afterwards, the patients are grouped according to the therapy outcome so that the physician can compare radiation parameters and therapy results for choosing the best possible therapy for the patient. In conclusion this research project shows that centralized information availability in tumor therapy is important for the improvement of the patient treatment process and the development of sophisticated decision support functions.

This paper presents a novel approach to biomedical image representation for classification by mapping image regions to local concepts and represent images in a weighted entropy based probabilistic feature space. In a heterogeneous collection of medical images, it is possible to identify specific local patches that are perceptually and/or semantically distinguishable. The variation of these patches is effectively modeled as local concepts based on their low-level features as inputs to a multi-class SVM classifier. The probability of occurrence of each concept in an image is measured by spreading and normalizing each region’s class confidence score based on the probabilistic output of the classifier. Furthermore, importance of concepts is measured as Shannon entropy based on pixel values of image patches and used to refine the feature vector to overcome the limitation of the “TF-IDF”- based weighting. In addition, to take the localization information of concepts into consideration, each image each segmented into five overlapping regions and local concept feature vectors are generated from those regions to finally obtain a combined semi-global feature vector. A systematic evaluation of image classification on two biomedical image data sets demonstrates improvement of more than 10% for the proposed feature representation approach compared to the commonly used low level and visual word-based approaches.

Imaging and image-based measurements nowadays play an essential role in controlled clinical trials, but electronic data capture (EDC) systems insufficiently support integration of captured images by mobile devices (e.g. smartphones and tablets). The web application OpenClinica has established as one of the world’s leading EDC systems and is used to collect, manage and store data of clinical trials in electronic case report forms (eCRFs). In this paper, we present a mobile application for instantaneous integration of images into OpenClinica directly during examination on patient’s bed site. The communication between the Android application and OpenClinica is based on the simple object access protocol (SOAP) and representational state transfer (REST) web services for metadata, and secure file transfer protocol (SFTP) for image transfer, respectively. OpenClinica’s web services are used to query context information (e.g. existing studies, events and subjects) and to import data into the eCRF, as well as export of eCRF metadata and structural information. A stable image transfer is ensured and progress information (e.g. remaining time) visualized to the user. The workflow is demonstrated for a European multi-center registry, where patients with calciphylaxis disease are included. Our approach improves the EDC workflow, saves time, and reduces costs. Furthermore, data privacy is enhanced, since storage of private health data on the imaging devices becomes obsolete.

One key problem for continuity of patient care is identification of a proper method to share and exchange patient medical records among multiple hospitals and healthcare providers. This paper focuses in the imaging document component of medical record. The XDS-I (Cross- Enterprise Document Sharing – Image) Profile based on the IHE IT-Infrastructure extends and specializes XDS to support imaging “document” sharing in an affinity domain. We present three studies about image sharing solutions based on IHE XDS-I Profile. The first one is to adopt the IHE XDS-I profile as a technical guide to design image and report sharing mechanisms between hospitals for regional healthcare service in Shanghai. The second study is for collaborating image diagnosis in regional healthcare services. The latter study is to investigate the XDS-I based clearinghouse for patient controlled image sharing in the RSNA Image Sharing Network Project. We conclude that the IHE XDS/XDS-I profiles can be used as the foundation to design medical image document sharing for Various Healthcare Services.

Radiological imaging is fundamental within the healthcare industry and has become routinely adopted for diagnosis, disease monitoring and treatment planning. Over the past two decades both diagnostic and therapeutic imaging have undergone a rapid growth, the ability to be able to harness this large influx of medical images can provide an essential resource for research and training. Traditionally, the systematic collection of medical images for research from heterogeneous sites has not been commonplace within the NHS and is fraught with challenges including; data acquisition, storage, secure transfer and correct anonymisation. Here, we describe a semi-automated system, which comprehensively oversees the collection of both unprocessed and processed medical images from acquisition to a centralised database. The provision of unprocessed images within our repository enables a multitude of potential research possibilities that utilise the images. Furthermore, we have developed systems and software to integrate these data with their associated clinical data and annotations providing a centralised dataset for research. Currently we regularly collect digital mammography images from two sites and partially collect from a further three, with efforts to expand into other modalities and sites currently ongoing. At present we have collected 34,014 2D images from 2623 individuals. In this paper we describe our medical image collection system for research and discuss the wide spectrum of challenges faced during the design and implementation of such systems.

Strongly evolving imaging technologies such as magnetic resonance imaging (MRI) nowadays provide a multitude of new complementary techniques for the analysis of cardiovascular tissue properties, function, and hemodynamics. The purpose of the presented work is to provide a research tool, which enables a quick validation of newly developed imaging techniques and supports the co-development of clinically usable analysis tools, which allow an integration with existing complementary examination methods. The concepts combined to this end consist of an integration with the open source research PACS OsiriX, an advanced heuristic DICOM classification and preprocessing as well as an integrative data model, which accumulates patient-specific image data, results and the data relations. Specific processing and analysis plugins can easily be integrated in such a way that they use the data integration and visualization infrastructure as well as results from other existing plugins. The presented example applications, such as the evaluation of slice orientations for cardiac function quantification or the integrated analysis of different types of image data for diagnosis of myocarditis show that the provided tool can be successfully used for a multitude of research applications in cardiovascular imaging.

Pain is a common complication after spinal cord injury with prevalence estimates ranging 77% to 81%, which highly affects a patient’s lifestyle and well-being. In the current clinical setting paper-based forms are used to classify pain correctly, however, the accuracy of diagnoses and optimal management of pain largely depend on the expert reviewer, which in many cases is not possible because of very few experts in this field. The need for a clinical decision support system that can be used by expert and non-expert clinicians has been cited in literature, but such a system has not been developed. We have designed and developed a stand-alone tool for correctly classifying pain type in spinal cord injury (SCI) patients, using Bayesian decision theory. Various machine learning simulation methods are used to verify the algorithm using a pilot study data set, which consists of 48 patients data set. The data set consists of the paper-based forms, collected at Long Beach VA clinic with pain classification done by expert in the field. Using the WEKA as the machine learning tool we have tested on the 48 patient dataset that the hypothesis that attributes collected on the forms and the pain location marked by patients have very significant impact on the pain type classification. This tool will be integrated with an imaging informatics system to support a clinical study that will test the effectiveness of using Proton Beam radiotherapy for treating spinal cord injury (SCI) related neuropathic pain as an alternative to invasive surgical lesioning.

Wearable technology defines a new class of smart devices that are accessories or clothing equipped with computational power and sensors, like Google Glass. In this work, we propose a novel concept for supporting everyday clinical pathways with wearable technology. In contrast to most prior work, we are not focusing on the omnipresent screen to display patient information or images, but are trying to maintain existing workflows. To achieve this, our system supports clinical staff as a documenting observer, only intervening adequately if problems are detected. Using the example of medication preparation and administration, a task known to be prone to errors, we demonstrate the full potential of the new devices. Patient and medication identifier are captured with the built-in camera, and the information is send to a transaction server. The server communicates with the hospital information system to obtain patient records and medication information. The system then analyses the new medication for possible side-effects and interactions with already administered drugs. The result is sent to the device while encapsulating all sensitive information respecting data security and privacy. The user only sees a traffic light style encoded feedback to avoid distraction. The server can reduce documentation efforts and reports in real-time on possible problems during medication preparation or administration. In conclusion, we designed a secure system around three basic principles with many applications in everyday clinical work: (i) interaction and distraction is kept as low as possible; (ii) no patient data is displayed; and (iii) device is pure observer, not part of the workflow. By reducing errors and documentation burden, our approach has the capability to boost clinical care.

Clinical trials usually have a demand to collect, track and analyze multimedia data according to the workflow. Currently, the clinical trial data management requirements are normally addressed with custom-built systems. Challenges occur in the workflow design within different trials. The traditional pre-defined custom-built system is usually limited to a specific clinical trial and normally requires time-consuming and resource-intensive software development. To provide a solution, we present a user customizable imaging informatics-based intelligent workflow engine system for managing stroke rehabilitation clinical trials with intelligent workflow. The intelligent workflow engine provides flexibility in building and tailoring the workflow in various stages of clinical trials. By providing a solution to tailor and automate the workflow, the system will save time and reduce errors for clinical trials. Although our system is designed for clinical trials for rehabilitation, it may be extended to other imaging based clinical trials as well.

ApoliopoproteinE Ɛ4 (ApoE-Ɛ4) polymorphism is the most well known genetic risk factor for developing Alzheimers Disease. The exact mechanism through which ApoE 4 increases AD risk is not fully known, but may be related to decreased clearance and increased oligomerization of Aβ. By making measurements of white matter integrity via diffusion MR and correlating the metrics in a voxel-based statistical analysis with ApoE-Ɛ4 genotype (whilst controlling for vascular risk factor, gender, cognitive status and age) we are able to identify changes in white matter associated with carrying an ApoE Ɛ4 allele. We found potentially significant regions (P_uncorrected < 0:05) near the hippocampus and the posterior cingulum that were independent of voxels that correlated with age or clinical dementia rating (CDR) status suggesting that ApoE may affect cognitive decline via a pathway in dependent of normal aging and acute insults that can be measured by CDR and Framingham Coronary Risk Score (FCRS).

Most white matter related neurological disease exhibit a large number of White Matter Hyperintensities (WMHs) on FLAIR MRI images. However, these lesions are not well understood. At the same time, Diffusion MRI has been gaining popularity as a powerful method of characterizing White Matter (WM) integrity. This work aims to study the behavior of the diffusion signal within the WMH voxels. The goal is to develop hybrid MR metrics that leverage information from multiple MR acquisitions to solve clinical problems. In our case, we are trying to address the WMH penumbra (as defined by Maillard et al 20112) where WMH delineates a foci that is more widespread than the actual damage area presumably due to acute inflammation. Our results show that diffusion MR metrics may be able to better delineate tissue that is inflamed versus scar tissue but may be less specific to lesions than FLAIR. Therefore, a hybrid metric that encodes information from both FLAIR and Diffusion MR may yield new and novel imaging information about the progression of white matter disease progression. We hope that this work also demonstrates how future PACS systems could have image fusion capabilities that would be able to leverage information from multiple imaging series to yield new and novel imaging contrast.

Radiation treatment planning (RTP) of the stereotactic body radiotherapy (SBRT) was more complex compared with conventional radiotherapy because of using a number of beam directions. We reported that similar planning cases could be helpful for determination of beam directions for treatment planners, who have less experiences of SBRT. The aim of this study was to develop a framework of searching for usable similar cases to an unplanned case in a RTP database based on a local image descriptor. This proposed framework consists of two steps searching and rearrangement. In the first step, the RTP database was searched for 10 cases most similar to object cases based on the shape similarity of two-dimensional lung region at the isocenter plane. In the second step, the 5 most similar cases were selected by using geometric features related to the location, size and shape of the planning target volume, lung and spinal cord. In the third step, the selected 5 cases were rearranged by use of the Euclidean distance of a local image descriptor, which is a similarity index based on the magnitudes and orientations of image gradients within a region of interest around an isocenter. It was assumed that the local image descriptor represents the information around lung tumors related to treatment planning. The cases, which were selected as cases most similar to test cases by the proposed method, were more resemble in terms of the tumor location than those selected by a conventional method. For evaluation of the proposed method, we applied a similar-cases-based beam arrangement method developed in the previous study to the similar cases selected by the proposed method based on a linear registration. The proposed method has the potential to suggest the superior beam-arrangements from the treatment point of view.

We have built a decision support system that provides recommendations for customizing radiation therapy treatment plans, based on patient models generated from a database of retrospective planning data. This database consists of relevant metadata and information derived from the following DICOM objects - CT images, RT Structure Set, RT Dose and RT Plan. The usefulness and accuracy of such patient models partly depends on the sample size of the learning data set. Our current goal is to increase this sample size by expanding our decision support system into a collaborative framework to include contributions from multiple collaborators. Potential collaborators are often reluctant to upload even anonymized patient files to repositories outside their local organizational network in order to avoid any conflicts with HIPAA Privacy and Security Rules. We have circumvented this problem by developing a tool that can parse DICOM files on the client’s side and extract de-identified numeric and text data from DICOM RT headers for uploading to a centralized system. As a result, the DICOM files containing PHI remain local to the client side. This is a novel workflow that results in adding only relevant yet valuable data from DICOM files to the centralized decision support knowledge base in such a way that the DICOM files never leave the contributor’s local workstation in a cloud-based environment. Such a workflow serves to encourage clinicians to contribute data for research endeavors by ensuring protection of electronic patient data.

“Imaging signs” are a critical part of radiology’s language. They not only are important for conveying diagnosis, but may also aid in indexing radiology literature and retrieving relevant cases and images. Here we report our work towards representing and categorizing imaging signs of abdominal abnormalities in figures in the radiology literature. Given a region-of-interest (ROI) from a figure, our goal was to assign a correct imaging sign label to that ROI from the following seven: accordion, comb, ring, sandwich, small bowel feces, target, or whirl. As training and test data, we created our own “gold standard” dataset of regions containing imaging signs. We computed 2997 feature attributes to represent imaging sign characteristics for each ROI in training and test sets. Following feature selection they were reduced to 70 attributes and were input to a Support Vector Machine classifier. We applied image-enhancement methods to compensate for variable quality of the images in radiology articles. In particular we developed a method for automatic detection and removal of pointers/markers (arrows, arrowheads, and asterisk symbols) on the images. These pointers/markers are valuable for approximately locating ROIs; however, they degrade the classification because they are often (partially) included in the training ROIs. On a test set of 283 ROIs, our method achieved an overall accuracy of 70% in labeling the seven signs, which we believe is a promising result for using imaging signs to search/retrieve radiology literature. This work is also potentially valuable for the creation of a visual ontology of biomedical imaging entities.

The new standard for radiology reporting templates being developed through the Integrating the Healthcare Enterprise (IHE) and DICOM organizations defines the storage and exchange of reporting templates as Hypertext Markup Language version 5 (HTML5) documents. The use of HTML5 enables the incorporation of "dynamic HTML," in which documents can be altered in response to their content. HTML5 documents can employ JavaScript, the HTML Document Object Model (DOM), and external web services to create intelligent reporting templates. Several reporting templates were created to demonstrate the use of scripts to perform in-template calculations and decision support. For example, a template for adrenal CT was created to compute contrast washout percentage from input values of precontrast, dynamic postcontrast, and delayed adrenal nodule attenuation values; the washout value can used to classify an adrenal nodule as a benign cortical adenoma. Dynamic templates were developed to compute volumes and apply diagnostic criteria, such as those for determination of internal carotid artery stenosis. Although reporting systems need not use a web browser to render the templates or their contents, the use of JavaScript creates innumerable opportunities to construct highly sophisticated HTML5 reporting templates. This report demonstrates the ability to incorporate dynamic content to enhance the use of radiology reporting templates.

Radiology and imaging informatics education have rapidly evolved over the past few decades. With the increasing recognition that future growth and maintenance of radiology practices will rely heavily on radiologists with fundamentally sound informatics skills, the onus falls on radiology residency programs to properly implement and execute an informatics curriculum. In addition, the American Board of Radiology may choose to include even more informatics on the new board examinations. However, the resources available for didactic teaching and guidance most especially at the introductory level are widespread and varied. Given the breadth of informatics, a centralized web-based interface designed to serve as an adjunct to standardized informatics curriculums as well as a stand-alone for other interested audiences is desirable. We present the development of a curriculum using PearlTrees, an existing web-interface based on the concept of a visual interest graph that allows users to collect, organize, and share any URL they find online as well as to upload photos and other documents. For our purpose, the group of “pearls” includes informatics concepts linked by appropriate hierarchal relationships. The curriculum was developed using a combination of our institution’s current informatics fellowship curriculum, the Practical Imaging Informatics textbook¹ and other useful online resources. After development of the initial interface and curriculum has been publicized, we anticipate that involvement by the informatics community will help promote collaborations and foster mentorships at all career levels.

With the impetus towards personalized and evidence-based medicine, the need for a framework to analyze/interpret quantitative measurements (blood work, toxicology, etc.) with qualitative descriptions (specialist reports after reading images, bio-medical knowledgebase, etc.) to predict diagnostic risks is fast emerging. Addressing this need, we pose and answer the following questions: (i) How can we jointly analyze and explore measurement data in context with qualitative domain knowledge? (ii) How can we search and hypothesize patterns (not known apriori) from such multi-structure data? (iii) How can we build predictive models by integrating weakly-associated multi-relational multi-structure data? We propose a framework towards answering these questions. We describe a software solution that leverages hardware for scalable in-memory analytics and applies next-generation semantic query tools on medical data.

With computing capability and display size growing, the mobile device has been used as a tool to help clinicians view patient information and medical images anywhere and anytime. It is uneasy and time-consuming for transferring medical images with large data size from picture archiving and communication system to mobile client, since the wireless network is unstable and limited by bandwidth. Besides, limited by computing capability, memory and power endurance, it is hard to provide a satisfactory quality of experience for radiologists to handle some complex post-processing of medical images on the mobile device, such as real-time direct interactive three-dimensional visualization. In this work, remote rendering technology is employed to implement the post-processing of medical images instead of local rendering, and a service protocol is developed to standardize the communication between the render server and mobile client. In order to make mobile devices with different platforms be able to access post-processing of medical images, the Extensible Markup Language is taken to describe this protocol, which contains four main parts: user authentication, medical image query/ retrieval, 2D post-processing (e.g. window leveling, pixel values obtained) and 3D post-processing (e.g. maximum intensity projection, multi-planar reconstruction, curved planar reformation and direct volume rendering). And then an instance is implemented to verify the protocol. This instance can support the mobile device access post-processing of medical image services on the render server via a client application or on the web page.

We have developed an external storage system by using secret sharing scheme and tokenization for regional medical cooperation, PHR service and information preservation. The use of mobile devices such as smart phones and tablets will be accelerated for a PHR service, and the confidential medical information is exposed to the risk of damage and intercept. We verified the transfer rate of the sending and receiving of data to and from the external storage system that connected it with PACS by the Internet this time. External storage systems are the data centers that exist in Okinawa, in Osaka, in Sapporo and in Tokyo by using secret sharing scheme. PACS continuously transmitted 382 CT images to the external data centers. Total capacity of the CT images is about 200MB. The total time that had been required to transmit was about 250 seconds. Because the preservation method to use secret sharing scheme is applied, security is strong. But, it also takes the information transfer time of this system too much. Therefore, DICOM data is masked to the header information part because it is made to anonymity in our method. The DICOM data made anonymous is preserved in the data base in the hospital. Header information including individual information is divided into two or more tallies by secret sharing scheme, and preserved at two or more external data centers. The token to relate the DICOM data anonymity made to header information preserved outside is strictly preserved in the token server. The capacity of header information that contains patient's individual information is only about 2% of the entire DICOM data. This total time that had been required to transmit was about 5 seconds. Other, common solutions that can protect computer communication networks from attacks are classified as cryptographic techniques or authentication techniques. Individual number IC card is connected with electronic certification authority of web medical image conference system. Individual number IC card is given only to the person to whom the authority to operate web medical image conference system was given.

MR imaging has been used to perform imaging guided high-intensity focused ultrasound (HIFU) and meanwhile can also be used precisely to measure tissue temperature in theory. But in practice, the temperature environment and target are complex. Therefore, it is difficult to measure targeted temperature just by simply using the theory of numerical calculation based on MR image information. In this presentation, we presented new MR temperature measurement, based on imaging informatics, to measure the targeted tissue temperature in MR imaging guided HIFU therapeutic procedure. By heating up the water phantom experiments under HIFU, the new algorithm gives a satisfactory result compared with existing algorithm. Based on experimental data, we can see the accuracy increase 37.5% from 0.4048℃ up to 0.2530℃ when we choose new algorithms.

With Synchrotron Radiation light source, there was a lot of imaging methods being developed to perform biomedical and medical imaging researches such as X-ray absorption imaging, phase-contrast imaging and micro-CT imaging. In this presentation, we present an approach to transform a various kinds of SR images into proper DICOM images so that to use a rich of medical processing display software to process and display SR biomedical and medical images. The new generated SR DICOM images can be transferred, stored, processed and displayed by using most of commercial medical imaging software.

One of the key challenges in three-dimensional (3D) medical imaging is to enable the fast turn-around time, which is often required for interactive or real-time response. This inevitably requires not only high computational power but also high memory bandwidth due to the massive amount of data that need to be processed. For this purpose, we previously developed a software platform for high-performance 3D medical image processing, called HPC 3D-MIP platform, which employs increasingly available and affordable commodity computing systems such as the multicore, cluster, and cloud computing systems. To achieve scalable high-performance computing, the platform employed size-adaptive, distributable block volumes as a core data structure for efficient parallelization of a wide range of 3D-MIP algorithms, supported task scheduling for efficient load distribution and balancing, and consisted of a layered parallel software libraries that allow image processing applications to share the common functionalities. We evaluated the performance of the HPC 3D-MIP platform by applying it to computationally intensive processes in virtual colonoscopy. Experimental results showed a 12-fold performance improvement on a workstation with 12-core CPUs over the original sequential implementation of the processes, indicating the efficiency of the platform. Analysis of performance scalability based on the Amdahl’s law for symmetric multicore chips showed the potential of a high performance scalability of the HPC 3DMIP platform when a larger number of cores is available.

The development of bigger and faster computers has not yet provided the computing power for medical image processing required nowadays. This is the result of several factors, including: i) the increasing number of qualified medical image users requiring sophisticated tools; ii) the demand for more performance and quality of results; iii) researchers are addressing problems that were previously considered extremely difficult to achieve; iv) medical images are produced with higher resolution and on a larger number. These factors lead to the need of exploring computing techniques that can boost the computational power of Healthcare Institutions while maintaining a relative low cost. Parallel computing is one of the approaches that can help solving this problem. Parallel computing can be achieved using multi-core processors, multiple processors, Graphical Processing Units (GPU), clusters or Grids. In order to gain the maximum benefit of parallel computing it is necessary to write specific programs for each environment or divide the data in smaller subsets. In this article we evaluate the performance of the two parallel computing tools when dealing with a medical image processing application. We compared the performance of the EELA-2 (E-science grid facility for Europe and Latin- America) grid infrastructure with a small Cluster (3 nodes x 8 cores = 24 cores) and a regular PC (Intel i3 – 2 cores). As expected the grid had a better performance for a large number of processes, the cluster for a small to medium number of processes and the PC for few processes.

Many US combat personnel have sustained nervous tissue trauma during service, which often causes Neuropathic pain as a side effect and is difficult to manage. However in select patients, synapse lesioning can provide significant pain control. Our goal is to determine the effectiveness of using Proton Beam radiotherapy for treating spinal cord injury (SCI) related neuropathic pain as an alternative to invasive surgical lesioning. The project is a joint collaboration of USC, Spinal Cord Institute VA Healthcare System, Long Beach, and Loma Linda University. This is first system of its kind that supports integration and standardization of imaging informatics data in DICOM format; clinical evaluation forms outcomes data and treatment planning data from the Treatment planning station (TPS) utilized to administer the proton therapy in DICOM-RT format. It also supports evaluation of SCI subjects for recruitment into the clinical study, which includes the development, and integration of digital forms and tools for automatic evaluation and classification of SCI pain. Last year, we presented the concept for the patient recruitment module based on the principle of Bayesian decision theory. This year we are presenting the fully developed patient recruitment module and its integration to other modules. In addition, the DICOM module for integrating DICOM and DICOM-RT-ION data is also developed and integrated. This allows researchers to upload animal/patient study data into the system. The patient recruitment module has been tested using 25 retrospective patient data and DICOM data module is tested using 5 sets of animal data.