medRxiv (Cold Spring Harbor Laboratory),
Journal Year:
2023,
Volume and Issue:
unknown
Published: Aug. 9, 2023
Abstract
Background
Large
Language
Models
(LLMs)
like
GPT-4
demonstrate
potential
applications
in
diverse
areas,
including
healthcare
and
patient
education.
This
study
evaluates
GPT-4’s
competency
against
osteoarthritis
(OA)
treatment
guidelines
from
the
United
States
China
assesses
its
ability
diagnosing
treating
orthopedic
diseases.
Methods
Data
sources
included
OA
management
examination
case
questions.
Queries
were
directed
to
based
on
these
resources,
responses
compared
with
established
cases.
The
accuracy
completeness
of
evaluated
using
Likert
scales,
while
inquiries
stratified
into
four
tiers
correctness
completeness.
Results
exhibited
strong
performance
providing
accurate
complete
recommendations
both
American
Chinese
guidelines,
high
scale
scores
for
It
demonstrated
proficiency
handling
clinical
cases,
making
diagnoses,
suggesting
appropriate
tests,
proposing
plans.
Few
errors
noted
specific
complex
Conclusions
exhibits
as
an
auxiliary
tool
practice
education,
demonstrating
interpreting
analyzing
Further
validation
capabilities
real-world
scenarios
is
needed.
Retos,
Journal Year:
2024,
Volume and Issue:
54, P. 312 - 319
Published: Feb. 28, 2024
El
análisis
bibliométrico
de
la
inteligencia
artificial
(IA)
en
el
deporte
revela
una
creciente
tendencia
investigación
y
aplicación
esta
tecnología
este
fenómeno
social.
En
última
década,
se
ha
observado
un
aumento
significativo
número
publicaciones
científicas
relacionadas
con
deporte,
lo
que
indica
gran
interés
tema.
objetivo
fue
analizar
bibliométricamente
los
elementos
deporte.
La
metodología
utilizada
hermenéutica
tres
componentes
fundamentales
Autores,
Revistas
Aportes
(ARA)
propuesta
por
autores
para
revisión
bibliométrica.
Se
analizaron
1002
artículos
científicos
pertenecientes
a
las
bases
datos
Scopus
(825),
Science
Direct
(172)
Mendeley
(5).
Como
criterios
inclusión
tomaron
dos:
todos
debían
ser
científicos,
idioma
español
e
inglés.
Los
principales
resultados
parten
identificación
autores,
revistas
aportes
potencia
IA
teniendo
cuenta
nuevas
metodologías
tendencias
anterior.
conclusión,
define
como
herramienta
corrige
errores,
ayuda
toma
decisiones,
estrategias
entrenamiento
deportivo
competencia,
prevenir
lesiones
deportivas,
estudiar
contrarios
potenciar
escenarios
deportivos
alta
calidad.
Palabras
clave:
Inteligencia
artificial,
deportivo,
moderno,
bibliométrico,
ARA.
Abstract.
The
bibliometric
analysis
of
intelligence
(AI)
in
sports
reveals
growing
trend
the
research
and
application
this
technology
social
phenomenon.
In
last
decade,
there
has
been
significant
increase
number
scientific
publications
related
to
sports,
indicating
great
interest
topic.
objective
was
bibliometrically
analyze
elements
sports.
methodology
used
hermeneutics
three
fundamental
components
Authors,
Journals
Contributions
proposed
by
authors
for
review.
articles
belonging
(5)
databases
were
analyzed.
Two
criteria
taken
as
inclusion
research:
all
had
be
articles,
Spanish
English.
main
results
are
based
on
identification
authors,
journals
contributions
that
enhance
AI
sport,
taking
into
account
new
methodologies
trends
above.
conclusion,
is
defined
tool
corrects
errors,
helps
decision-making,
enhances
training
competition
strategies,
prevent
injuries,
study
opponents
scenarios.
high
quality
Keywords:
Artificial
intelligence,
training,
modern
analysis,
ARA
methodology.
PLoS ONE,
Journal Year:
2024,
Volume and Issue:
19(10), P. e0310028 - e0310028
Published: Oct. 9, 2024
ChatGPT,
a
general
artificial
intelligence,
has
been
recognized
as
powerful
tool
in
scientific
writing
and
programming
but
its
use
medical
is
largely
overlooked.
The
accessibility,
rapid
response
time
comprehensive
training
database
might
enable
ChatGPT
to
serve
diagnostic
augmentation
certain
clinical
settings.
process
neurology
often
challenging
complex.
In
time-sensitive
scenarios,
evaluation
decisions
are
needed,
while
other
cases
clinicians
faced
with
rare
disorders
atypical
disease
manifestations.
Due
these
factors,
the
accuracy
suboptimal.
Here
we
evaluated
whether
can
be
utilized
valuable
innovative
various
neurological
We
used
synthetic
data
generated
by
experts
represent
descriptive
anamneses
of
patients
known
neurology-related
diseases,
then
probability
for
an
appropriate
diagnosis
made
was
measured.
To
give
clarity
AI-determined
diagnosis,
all
have
cross-validated
doctors
well.
found
that
ChatGPT-determined
(ranging
from
68.5%
±
3.28%
83.83%
2.73%)
reach
(81.66%
2.02%),
furthermore,
it
surpasses
if
examiner
doctor
(57.15%
2.64%).
Our
results
showcase
efficacy
intelligence
like
medicine.
future,
AI-based
supporting
tools
useful
amendments
practice
help
improve
neurology.
Geoheritage,
Journal Year:
2024,
Volume and Issue:
16(4)
Published: Oct. 15, 2024
Abstract
Generative
artificial
intelligence
systems
based
on
the
GPT
model
have
shown
groundbreaking
capabilities
in
producing
scientific
texts,
revolutionising
how
geoscientists
research
and
teach.
However,
there
is
a
noticeable
absence
of
generative
AI
tailored
specifically
for
geoscience,
efficacy
technology
has
not
been
examined
within
Earth
Science
realm.
To
fill
these
gaps,
we
propose
new
system
(GeologyOracle)
built
upon
GPT-4o
trained
data.
It
designed
to
simulate
conversation
with
geoscientist,
having
analyse
geologic
datasets,
suggest
geoscience
hypotheses,
explain
Earth-Science
concepts,
interpret
geosites.
evaluate
performance
GeologyOracle,
was
tested
152
questions
provided
evaluated
by
panel
academics
scale
0–10.
Performance
analysis
revealed
that
79.6%
answers
scored
equal
or
above
passing
mark
5,
tendency
deliver
high-quality
responses
(mean:
6.5;
median
=
7;
interquartile
range:
5–8).
The
results
demonstrate
GeologyOracle
effective
performing
complex
tasks,
such
as
identifying
rocks,
fossils,
minerals,
interpreting
outcrop,
core
data,
quantitative
datasets.
exhibited
noteworthy
ability
interpretation
internationally-acknowledged
geosites
situated
geoparks
nature
reserves.
comparable
specialists,
suggesting
its
potential
an
assistant
various
fields,
including
structural
geology,
palaeontology,
geomorphology,
sedimentology,
economic
geology.
While
does
render
obsolete,
it
offers
significant
accelerating
discovery,
automating
research,
assisting
educators,
students,
geotourists.
medRxiv (Cold Spring Harbor Laboratory),
Journal Year:
2023,
Volume and Issue:
unknown
Published: Aug. 9, 2023
Abstract
Background
Large
Language
Models
(LLMs)
like
GPT-4
demonstrate
potential
applications
in
diverse
areas,
including
healthcare
and
patient
education.
This
study
evaluates
GPT-4’s
competency
against
osteoarthritis
(OA)
treatment
guidelines
from
the
United
States
China
assesses
its
ability
diagnosing
treating
orthopedic
diseases.
Methods
Data
sources
included
OA
management
examination
case
questions.
Queries
were
directed
to
based
on
these
resources,
responses
compared
with
established
cases.
The
accuracy
completeness
of
evaluated
using
Likert
scales,
while
inquiries
stratified
into
four
tiers
correctness
completeness.
Results
exhibited
strong
performance
providing
accurate
complete
recommendations
both
American
Chinese
guidelines,
high
scale
scores
for
It
demonstrated
proficiency
handling
clinical
cases,
making
diagnoses,
suggesting
appropriate
tests,
proposing
plans.
Few
errors
noted
specific
complex
Conclusions
exhibits
as
an
auxiliary
tool
practice
education,
demonstrating
interpreting
analyzing
Further
validation
capabilities
real-world
scenarios
is
needed.
