Single cell transcriptomics reveals early photoreceptor states, cell-specific transcript isoforms, and cancer-predisposing features.

Human cone photoreceptors differ from rods and serve as the retinoblastoma cell-of-origin. Here, we used deep full-length single-cell RNA-sequencing to distinguish post-mitotic cone and rod developmental states and cone-specific features that contribute to retinoblastomagenesis. The analyses revealed early post-mitotic cone- and rod-directed populations characterized by higher THRB or NRL regulon activities, an immature photoreceptor precursor population with concurrent cone and rod gene and regulon expression, and distinct early and late cone and rod maturation states distinguished by maturation-associated declines in RAX regulon activity. Unexpectedly, both L/M cone and rod precursors co-expressed NRL and THRB RNAs, yet they differentially expressed functionally antagonistic NRL isoforms and prematurely terminated THRB transcripts. Early L/M cone precursors exhibited successive expression of lncRNAs along with MYCN, which composed the seventh most L/M-cone-specific regulon, and SYK, which contributed to the early cone precursors' proliferative response to RB1 loss. These findings reveal previously unrecognized photoreceptor precursor states and a role for early cone-precursor-intrinsic SYK expression in retinoblastoma initiation.

funcLAB/G-service-oriented architecture for standards-based analysis of functional magnetic resonance imaging in HealthGrids.

Functional MRI is successfully being used in clinical and research applications including preoperative planning, language mapping, and outcome monitoring. However, clinical use of fMRI is less widespread due to its complexity of imaging, image workflow, post-processing, and lack of algorithmic standards hindering result comparability. As a consequence, wide-spread adoption of fMRI as clinical tool is low contributing to the uncertainty of community physicians how to integrate fMRI into practice. In addition, training of physicians with fMRI is in its infancy and requires clinical and technical understanding. Therefore, many institutions which perform fMRI have a team of basic researchers and physicians to perform fMRI as a routine imaging tool. In order to provide fMRI as an advanced diagnostic tool to the benefit of a larger patient population, image acquisition and image post-processing must be streamlined, standardized, and available at any institution which does not have these resources available. Here we describe a software architecture, the functional imaging laboratory (funcLAB/G), which addresses (i) standardized image processing using Statistical Parametric Mapping and (ii) its extension to secure sharing and availability for the community using standards-based Grid technology (Globus Toolkit). funcLAB/G carries the potential to overcome the limitations of fMRI in clinical use and thus makes standardized fMRI available to the broader healthcare enterprise utilizing the Internet and HealthGrid Web Services technology.

Globus MEDICUS - federation of DICOM medical imaging devices into healthcare Grids.

The Digital Imaging and Communications in Medicine (DICOM) standard defines Radiology medical device interoperability and image data exchange between modalities, image databases - Picture Archiving and Communication Systems (PACS) - and image review end-points. However the scope of DICOM and PACS technology is currently limited to the trusted and static environment of the hospital. In order to meet the demand for ad-hoc tele-radiology and image guided medical procedures within the global healthcare enterprise, a new technology must provide mobility, security, flexible scale of operations, and rapid responsiveness for DICOM medical devices and subsequently medical image data. Grid technology, an informatics approach to securely federate independently operated computing, storage, and data management resources at the global scale over public networks, meets these core requirements. Here we present an approach to federate DICOM and PACS devices for large-scale medical image workflows within a global healthcare enterprise. The Globus MEDICUS (Medical Imaging and Computing for Unified Information Sharing) project uses the standards-based Globus Toolkit Grid infrastructure to vertically integrate a new service for DICOM devices - the DICOM Grid Interface Service (DGIS). This new service translates between DICOM and Grid operations and thus transparently extends DICOM to Globus based Grid infrastructure. This Grid image workflow paradigm has been designed to provide not only solutions for global image communication, but fault-tolerance and disaster recovery using Grid data replication technology. Actual use-case of 40 MEDICUS Grid connected international hospitals of the Childerns Oncology Group and the Neuroblastoma Cancer Foundation and further clinical applications are discussed. The open-source Globus MEDICU http://dev.globus.org/wiki/Incubator/MEDICUS.

Somatosensory lateralization in the newborn brain.

Since the onset and early postnatal development of hemispheric lateralization in the human brain are unknown, we studied cortical activation induced by passive extension and flexion of the hand in neonates using functional magnetic resonance imaging (fMRI). In contrast to that seen in older age groups, somatosensory areas in the pre- and postcentral gyri of the neonate showed no significant hemispheric lateralization at term. Instead, our findings from independent left- and right-hand experiments suggest the presence of an emerging trend of contralateral lateralization of the somatosensory system at around term.

Functional MRI in neonates using neonatal head coil and MR compatible incubator.

Structural and functional magnetic resonance imaging of the newborn brain is a complex and challenging task. Term and preterm neonates require a controlled microenvironment and close monitoring during the MRI study to maintain respiratory and cardiovascular functions, body temperature, and fluid and electrolyte homeostasis. In addition, to minimize motion artifacts, most neonates also need to be sedated, which carries the risk of respiratory depression compromising the neonate's ability to maintain appropriate ventilation and oxygenation during the procedure. Finally, because of their small head size, the use of the standard MR head coils results in suboptimal picture quality in the neonate. Thus, these limitations affect our ability to obtain both high quality structural and functional MRI studies. To overcome these difficulties, we have utilized an MR compatible incubator with a built-in radiofrequency head coil optimized for the neonatal brain volume. In this study we demonstrate that functional MRI and high-resolution structural MRI of the newborn brain can be achieved with this novel design. The use of this equipment offers potential for studying the development of the preterm and term neonatal brain and obtaining state-of-the-art, high-resolution structural and functional imaging in this most vulnerable patient population.

PACS-based functional magnetic resonance imaging.

The picture archiving and communication system (PACS) technology reaches its 10th anniversary. Retrospectively no one could foresee the impact the PACS would have to the health care enterprise, but it is common consent today, that PACS is the key technology crucial to daily clinical image operations and especially to image related basic and clinical research. During the past 10 years the PACS has been matured from a research and developmental stage into commercial products which are provided by all major modality and health care equipment vendors. The PACS, originally implemented in the Radiology Department, needs to grow and has already carried well beyond departmental limits conquering all image relevant areas inside the hospital. During the past 10 years a dramatic development in imaging techniques especially within MRI emerged. Advanced 3D- and 4D-MR imaging techniques result in much more images and more complex data objects than ever before which need to be implemented into the existing PACS. These new imaging techniques require intensive post-processing apart from the imaging modality which need to be integrated into the image workflow and the PACS implementation. Along with these new imaging techniques new clinical applications, e.g. stroke detection, and research applications, e.g. study of heart and brain function, in Neurology and Cardiology require changes to the traditional PACS concept. Therefore inter-disciplinary image distribution will become the high-water mark for the next 10 years in the PACS endeavor. This paper focuses on one new advanced imaging technique, functional magnetic resonance imaging (fMRI), and discusses how fMRI data is defined, what fMRI requires in terms of clinical and research applications and how to implement fMRI in the existing PACS.