Advanced

Role of ventilation scintigraphy in diagnosis of acute pulmonary embolism: an evaluation using artificial neural networks.

Evander, Eva LU ; Holst, Holger; Järund, Andreas; Ohlsson, Mattias LU ; Wollmer, Per LU ; Åström, Karl LU and Edenbrandt, Lars LU (2003) In European Journal of Nuclear Medicine and Molecular Imaging 30(7). p.961-965
Abstract
The purpose of this study was to assess the

value of the ventilation study in the diagnosis of acute

pulmonary embolism using a new automated method.

Either perfusion scintigrams alone or two different combinations

of ventilation/perfusion scintigrams were used

as the only source of information regarding pulmonary

embolism. A completely automated method based on

computerised image processing and artificial neural networks

was used for the interpretation. Three artificial

neural networks were trained for the diagnosis of pulmonary

embolism. Each network was trained with 18 automatically

obtained features. Three different sets of... (More)
The purpose of this study was to assess the

value of the ventilation study in the diagnosis of acute

pulmonary embolism using a new automated method.

Either perfusion scintigrams alone or two different combinations

of ventilation/perfusion scintigrams were used

as the only source of information regarding pulmonary

embolism. A completely automated method based on

computerised image processing and artificial neural networks

was used for the interpretation. Three artificial

neural networks were trained for the diagnosis of pulmonary

embolism. Each network was trained with 18 automatically

obtained features. Three different sets of features

originating from three sets of scintigrams were

used. One network was trained using features obtained

from each set of perfusion scintigrams, including six

projections. The second network was trained using features

from each set of (joint) ventilation and perfusion

studies in six projections. A third network was trained

using features from the perfusion study in six projections

combined with a single ventilation image from the posterior

view. A total of 1,087 scintigrams from patients with

suspected pulmonary embolism were used for network

training. The test group consisted of 102 patients who

had undergone both scintigraphy and pulmonary angiography.

Performances in the test group were measured as

area under the receiver operation characteristic curve.

The performance of the neural network in interpreting

perfusion scintigrams alone was 0.79 (95% confidence

limits 0.71–0.86). When one ventilation image (posterior

view) was added to the perfusion study, the performance

was 0.84 (0.77–0.90). This increase was statistically significant

(P=0.022). The performance increased to 0.87

(0.81–0.93) when all perfusion and ventilation images

were used, and the increase in performance from 0.79 to

0.87 was also statistically significant (P=0.016). The automated

method presented here for the interpretation of

lung scintigrams shows a significant increase in performance

when one or all ventilation images are added to

the six perfusion images. Thus, the ventilation study has

a significant role in the diagnosis of acute lung embolism. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
European Journal of Nuclear Medicine and Molecular Imaging
volume
30
issue
7
pages
961 - 965
publisher
Springer
external identifiers
  • wos:000184644800004
  • pmid:12748832
  • scopus:0042192115
ISSN
1619-7070
DOI
10.1007/s00259-003-1182-5
language
English
LU publication?
yes
id
1e802720-4449-4afa-a920-104b224b4e09 (old id 114036)
alternative location
http://www.ncbi.nlm.nih.gov/sites/entrez?cmd=Retrieve&db=PubMed&list_uids=12748832&dopt=Abstract
date added to LUP
2007-07-17 13:51:41
date last changed
2018-01-07 06:18:06
@article{1e802720-4449-4afa-a920-104b224b4e09,
  abstract     = {The purpose of this study was to assess the<br/><br>
value of the ventilation study in the diagnosis of acute<br/><br>
pulmonary embolism using a new automated method.<br/><br>
Either perfusion scintigrams alone or two different combinations<br/><br>
of ventilation/perfusion scintigrams were used<br/><br>
as the only source of information regarding pulmonary<br/><br>
embolism. A completely automated method based on<br/><br>
computerised image processing and artificial neural networks<br/><br>
was used for the interpretation. Three artificial<br/><br>
neural networks were trained for the diagnosis of pulmonary<br/><br>
embolism. Each network was trained with 18 automatically<br/><br>
obtained features. Three different sets of features<br/><br>
originating from three sets of scintigrams were<br/><br>
used. One network was trained using features obtained<br/><br>
from each set of perfusion scintigrams, including six<br/><br>
projections. The second network was trained using features<br/><br>
from each set of (joint) ventilation and perfusion<br/><br>
studies in six projections. A third network was trained<br/><br>
using features from the perfusion study in six projections<br/><br>
combined with a single ventilation image from the posterior<br/><br>
view. A total of 1,087 scintigrams from patients with<br/><br>
suspected pulmonary embolism were used for network<br/><br>
training. The test group consisted of 102 patients who<br/><br>
had undergone both scintigraphy and pulmonary angiography.<br/><br>
Performances in the test group were measured as<br/><br>
area under the receiver operation characteristic curve.<br/><br>
The performance of the neural network in interpreting<br/><br>
perfusion scintigrams alone was 0.79 (95% confidence<br/><br>
limits 0.71–0.86). When one ventilation image (posterior<br/><br>
view) was added to the perfusion study, the performance<br/><br>
was 0.84 (0.77–0.90). This increase was statistically significant<br/><br>
(P=0.022). The performance increased to 0.87<br/><br>
(0.81–0.93) when all perfusion and ventilation images<br/><br>
were used, and the increase in performance from 0.79 to<br/><br>
0.87 was also statistically significant (P=0.016). The automated<br/><br>
method presented here for the interpretation of<br/><br>
lung scintigrams shows a significant increase in performance<br/><br>
when one or all ventilation images are added to<br/><br>
the six perfusion images. Thus, the ventilation study has<br/><br>
a significant role in the diagnosis of acute lung embolism.},
  author       = {Evander, Eva and Holst, Holger and Järund, Andreas and Ohlsson, Mattias and Wollmer, Per and Åström, Karl and Edenbrandt, Lars},
  issn         = {1619-7070},
  language     = {eng},
  number       = {7},
  pages        = {961--965},
  publisher    = {Springer},
  series       = {European Journal of Nuclear Medicine and Molecular Imaging},
  title        = {Role of ventilation scintigraphy in diagnosis of acute pulmonary embolism: an evaluation using artificial neural networks.},
  url          = {http://dx.doi.org/10.1007/s00259-003-1182-5},
  volume       = {30},
  year         = {2003},
}