To maximize the contrast and clarity of coregistered PET (Positron Emission Tomography) and MR (Magnetic Resonance) brain scans, the
application of certain color maps have been tested. Twelve of Matlab’s predefined color maps were applied on a PET-MR overlay to
discover which are the easiest to read, yet are informative. The experiments are carried out in three phases, or constructs. These
constructs have been executed on Matlab by functions within a custom-built library, which allows for the overlaying and resizing of brain
scans. The first and second constructs consist of gray scale PET over colored MR and gray scale MR over colored PET, respectively.
These first two constructs determine which type of scan, PET or MR, should be colored in an overlay, as well as the type of color map that
should be applied. The third construct performs a fusion between colored PET and colored MR scans. A clinical trial consisting of twentyfive
high school students has been completed. Students answered questions based on images generated by the Matlab software. Results
confirm that the second construct, gray scale MR over colored PET, was preferred over the first. Next, students compared different color
maps to identify which one allows for the best distinction between PET and MR scans in their overlay. Participants ranked the color maps,
summer, copper, and cool, as first, second, and third, based on the contrast and clarity of the scans. Future studies will examine
responses from brain imaging professionals and test custom-designed color maps on PET-MR overlays.
Shiladitya Dutta and Carl Taswell, 2018,
SPARQL-Based Search Engine and Agent for Finding Brain Literature and Converting References to NPDS Metadata Records
presented December 2018 as Abstract B277 at the
11th International Conference on Brain Informatics
in Arlington, Texas.
We describe CoVaSEA (Concept-Validating Search Engine Agent): an automated web crawler/query engine that is
interoperable with the Nexus-PORTAL-DOORS System. The Nexus-PORTAL-DOORS System (NPDS) is a data management system that
organizes repositories of lexical metadata (in PORTAL servers) and semantic representations (in DOORS servers) of resources. Due to
the purpose built hybridized nature of NPDS, it is well placed to perform a variety of data analysis tasks. However, many of these tasks
require records of semantic descriptions which are labor intensive to create and maintain due to the substantial and rapidly increasing
quantities of brain related literature available on the open web. To remedy this, we created CoVaSEA with the intention of providing an
automated method for users to navigate and expand the semantic records of brain literature in the NPDS directories. To this end,
CoVaSEA integrates multiple features which benefit NPDS including: (A) An implementation of SPARQL query based search to allow
retrieval and manipulation of RDF descriptions, (B) Targeted web-crawling for relevant articles from external biomedical literature
databases to broaden NPDS records, and (C) Translation of free-form text into RDF triples to derive the semantic portrayals of lexical data.
CoVaSEA consists of three principal components: the web-crawler, the lexical to semantic converter, and the SPARQL query engine. The
web crawler retrieves articles along with their basic metadata (title, abstract, author(s), etc.) from several of biomedical literature databases
via REST API. However, in order to capture a full semantic description of the data in each article, key RDF triples which describe the
abstract are constructed. First, each of the unique nouns in the passage are registered via coreference resolution and pronomial
anaphora. Then the sentences are parsed into constituency tree format so that the subject(s), verb(s), and object(s) can be extracted.
Once the SVO triples are extracted, they are transformed into valid RDF by assigning unique resource identifiers (URI) to each part of the
triples. This is accomplished by using various databases (i.e. MeSH) for terminology and select named entities, word sense
disambiguation for standard words, and literals for any other sections. These triples are stored via the Scribe API in either a DOORS
directory or a localized triplestore where they can be retrieved via the SPARQL query engine. In order to create a more conducive user
experience, the query engine supports the capability to construct SPARQL queries from expressions in conjunctive normal form, thus
circumventing the need to know SPARQL syntax. With the distinct advantage that the system is automated, CoVaSEA presents the
capability to search “externally” to furnish large numbers of brain-related literature descriptions on a regular basis and search “internally” to
provide a method of retrieving those descriptions, thus laying the groundwork for a variety of future NPDS applications for which semantic
metadata stores of brain literature are a functional necessity.
Measuring the merits of a scholarly article
only by how often other articles or social media posts
cite it creates a perverse incentive for authors to avoid
citing potential rivals. To uphold established standards of
scholarship, institutions should also consider one or more
metrics of how appropriately an article cites relevant prior
work. This paper describes the general characteristics of
the FAIR Attribution to Indexed Reports (FAIR) family of
metrics, which we have designed for this purpose. We
formulate five FAIR metrics suitable for use with primary research
articles. Two measure adherence to best practices:
number of correctly attributed background statements
and number of genuinely original claims. Three measure
specific deviations from best practices: number of misattributed
background statements, number of background
statements with missing references, and number of claims
falsely indicated as original. We conclude with a discussion
of plans to implement a web application for calculating
metric values of scholarly works described by records in
Nexus-PORTAL-DOORS System (NPDS) servers.
Measuring the merits of scholarly research articles only by citation counts and how often other
research articles or social media messages cite a particular publication creates a perverse incentive
for some authors to refrain from citing potential rivals. This dilemma has developed despite
the historical publishing standard expected in peer review for citing and discussing related prior
work. To encourage and support a countervailing incentive, research organizations should also
consider metrics for how well and appropriately a scholarly article cites relevant prior work in
the spirit of the classic phrase and metaphor standing on the shoulders of giants. We present a
proposal for a family of such article-level metrics called the FAIR metrics and described as the
FAIR Attribution to Indexed Reports or the FAIR Acknowledgment of Information Records.
Objective: To assess the psychological impact of disclosing
a positive or negative amyloid brain scan result to symptomatic
individuals with mild cognitive impairment (MCI) or mild
Alzheimer’s disease (AD).
Design: Prospective longitudinal cohort study.
Setting: Florey Institute of Neuroscience & Mental Health,
University of Melbourne, Australia.
Participants: A total of 133 individuals aged 50–85 with
MCI or mild AD enrolled in the study with data collected
between October 2014 and June 2016.
Interventions: Disclosure of amyloid imaging results to
Measurements: Positron emission tomography (PET) brain
amyloid imaging with [18F]-NAV4694; psychometric scales
including the Center for Epidemiologic Studies Depression
(CES-D) scale, Geriatric Depression Scale (GDS), Hospital
Anxiety and Depression Scales (HADS-A and HADS-D) and
State-Trait Anxiety Inventory (STAI) performed before and
after disclosure of amyloid imaging results.
Results: We did not observe any worsening of psychological
health with a panel of psychometric scales assessed on
individuals to whom amyloid brain scan results were disclosed.
Conclusions: We consider it safe, without apparent risk of
harm to patients, to disclose amyloid imaging results to patients
who have no prior history of neuropsychiatric illness.
The Nexus-PORTAL-DOORS System (NPDS) has
been designed with the Hierarchically Distributed Mobile
Metadata (HDMM) architectural style to provide
an infrastructure system for managing both lexical
and semantic metadata about both virtual and
physical entities. We describe here how compatibility
between version 0.9 of the NPDS schema, the new
NPDS-interfacing ontologies, and the domain-specific
concept-validating hypothesis-exploring ontologies allows
NPDS to bootstrap the semantic web onto the
more developed lexical web. We then describe how
this system will serve as the foundation of a planned
platform for automated meta-analysis.
Even though online databases make it easier than
ever to access the biomedical and scientific literature about
dementia, accelerating growth in the size of these databases
has made it more difficult for humans to gather and analyze
manually all articles relevant to any given topic. We document
a Nexus-PORTAL-DOORS System (NPDS) Concept-Validating
Search Engine Agent that can populate Nexus diristries with
concept-validated metadata records for citations of journal
articles found in literature databases.
Does the clinical status of patients with either
Alzheimer’s disease or mild cognitive impairment when compared
with the normal healthy status of control subjects have
an effect on the co-registration accuracy of the participants
PET and MRI brain scans? An initial evaluation reveals that
a statistically significant difference may exist in co-registration
accuracy with some popular algorithms for the different groups
of participants’ brain scans. These differences suggest that
investigators should use appropriate caution when reviewing
fusion studies of co-registered PET and MRI brain scans.
The ability to view medical images as 3D objects,
which can be explored interactively, has now become possible
due to the advent of rapidly emerging virtual reality (VR)
technologies. In the past, VR has been used as an educational
tool for learning anatomy, a visualization tool for assisting
surgery, and a therapeutic tool for rehabilitating patients with
motor disorders. However, these older systems were either
expensive to build or difficult to acquire and use. Exploiting the
arrival of new consumer devices such as the Oculus Rift that are
now affordable, we have developed a software application called
BrainWatch for VR ready computers to enable 3D visualization
and interactive exploration of DICOM data sets focusing on
PET and MRI brain scans. BrainWatch software provides a
unique set of 3 approaches for interacting with the virtual
object which we have named the observatory scenario with
an external camera, the planetarium scenario with an internal
camera, and the voyager scenario with a mobile camera. A live
interactive demonstration of BrainWatch VR with the Oculus
Rift CV1 will be available for conference attendees to experience
at EMBC 2017.
The Nexus-PORTAL-DOORS System (NPDS) has been designed
with the Hierarchically Distributed Mobile Metadata (HDMM)
architectural style to provide an infrastructure system for managing both
lexical and semantic metadata about both virtual and physical entities.
We describe version 0.8 of NPDS, including the separation of concerns
between the original Problem-Oriented Registry of Tags And Labels
(PORTAL) registries and the Domain Ontology Oriented Resource System
(DOORS) directories, the combined registry and directory functionality
of Nexus diristries, and the RESTful read-only web service API through
which resource representation metadata records can be retrieved from
these NPDS servers. We also introduce Scribe registrars with a
corresponding RESTful read-write web service API for management of
metadata records by both software agents accessing the web services directly
and human users accessing them indirectly via web applications.
In routine clinical imaging, PET and MR images often undergo co-registration,
however, methods for co-registration may vary.
The significance of differences between methods has not been previously determined.
Registration accuracy was calculated both qualitatively and quantitatively using different metrics.
Both the quantitative metrics and subsequent visual inspection confirm that there exists a significant difference
between different registration methods. Because a difference does exist across co-registration methods, clinicians
and researchers must take appropriate care when choosing what method to use for PET-MR co-registration.
In a world of rapidly emerging commercial Virtual Reality (VR) technologies, such as Oculus Rift,
the ability to view medical images as an interactive 3D object, which can be virtually entered, becomes a possibility.
In the past, virtual reality has been used as an educational tool, for therapy in motor disorders, and also as a
visualization tool for surgery, however, many of these systems have been a combination of incredibly costly and
often difficult to come by. Using these improved and more readily available technologies, we have created an
application for use with a VR enabled computer and Oculus Rift to allow the 3D visualization of DICOM datasets,
specifically MRI and PET brain scans.
The PORTAL-DOORS system (PDS) has been designed
as a resource metadata management system intended to
support applications such as automated searches of online resources
and meta-analyses of published literature. PDS comprises
a network of Problem Oriented Registry of Tags and Labels
(PORTAL) lexical registries and Domain Ontology Oriented Resource
System (DOORS) semantic directories. Here we introduce
a PDS-compliant concept-validating registry and hypothesisexploring
ontology that organizes focal-onset dementias including
Sensory-Onset, Language-Onset and Motor-ONset (SOLOMON)
dementias with novel classifying and relating concepts. This
approach facilitates semantic search of resources and exploration
of hypotheses related to neurodegeneration. SOLOMON
interoperates with other PDS registries and ontologies including
BrainWatch, ManRay and GeneScene.
The PORTAL-DOORS system (PDS) has been designed as a resource metadata management system
intended to support applications such as automated searches of online resources and meta-analyses of
published literature. We present a methodological approach with a PDS-compliant concept-validating registry
and hypothesis-exploring ontology that organizes focal-onset dementias including
Sensory-Onset, Language-Onset and Motor-ONset(SOLOMON) dementias with novel classifying and relating concepts.
This approach facilitates semantic search of resources and exploration of hypotheses related to neurodegeneration.
SOLOMON interoperates with other PDS registries and ontologies including BrainWatch, ManRayand GeneScene.
Clinical telegaming integrates telecare and
videogaming to enable a more convenient and enjoyable experience
for patients when providers diagnose, monitor, and treat
a variety of health problems via web-enabled telecommunications.
In recent years, clinical telegaming systems have been
applied to physical therapy and rehabilitation, evaluation of
mental health, and prevention and management of obesity and
diabetes. Parkinson’s disease (PD) is suitable for development
of new clinical telegaming applications because PD patients are
known to experience motor symptoms that can be improved
by physical therapy. Recent research suggests that sensory
processing deficits may also play an important role in these
motor impairments because successful motor function requires
multisensory integration. In this paper, we describe a new
web-enabled software system that uses clinical telegaming to
evaluate and improve multisensory integration ability in users.
This software has the potential to be used in diagnostic and
therapeutic telegaming for PD patients.
Alzheimer disease is the cause of up to one-third of cases of primary
progressive aphasia or corticobasal syndrome. The primary objective
of this study was to determine the accuracy of 18F-FDG PET metabolic
imaging for the detection of Alzheimer disease in patients with
primary progressive aphasia or corticobasal syndrome. Methods: A
cohort of patients (n = 94), including those with an expert clinical
diagnosis of logopenic (n = 19), nonfluent (n = 16), or semantic
(n = 13) variants of primary progressive aphasia, corticobasal syndrome
(n = 14), or Alzheimer disease (n = 24), underwent 18F-FDG
metabolic and 11C-labeled Pittsburgh compound B (11C-PiB) amyloid
PET brain imaging. 18F-FDG PET scans interpreted with Neurostat
and 3D-SSP displays were classified as revealing Alzheimer disease
or “other” by interpreters who were unaware of the clinical assessments
and 11C-PiB PET results. 11C-PiB PET imaging was considered
to be the diagnostic reference standard, with a threshold standardized
uptake value ratio of 1.5 being indicative of Alzheimer disease
pathology. To address possible bias from subgroup selection for the
Alzheimer disease binary classifier, we calculated both conventional
and balanced accuracies. Results: Diagnoses of Alzheimer disease
based on 18F-FDG PET resulted in 84% accuracy (both conventional
and balanced). In comparison, diagnoses based on clinical assessments
resulted in 65% conventional accuracy and 67% balanced
accuracy. Conclusion: Brain 18F-FDG PET scans interpreted with
Neurostat and 3D-SSP displays accurately detected Alzheimer disease
in patients with primary progressive aphasia or corticobasal syndrome
as focal-onset dementias. In such diagnostically challenging cohorts,
18F-FDG PET imaging can provide more accurate diagnoses, enabling
more appropriate therapy.