| Title | Role of ventilation scintigraphy in diagnosis of acute pulmonary embolism: an evaluation using artificial neural networks. |
| Authors | Eva Evander, Holger Holst, Andreas Järund, Mattias Ohlsson, Per Wollmer, Karl Åström, Lars Edenbrandt |
| Alternative Location | http://www.ncbi.nlm.nih.gov..., Restricted Access |
| Alternative Location | http://dx.doi.org/10.1007/s..., Restricted Access |
| Publication | European journal of nuclear medicine and molecular imaging |
| Year | 2003 |
| Volume | 30 |
| Issue | 7 |
| Pages | 961 - 965 |
| Document type | Article |
| Status | Published |
| Quality controlled | Yes |
| Language | eng |
| Publisher | Springer-Verlag GmbH |
| Abstract English | The purpose of this study was to assess the<br> value of the ventilation study in the diagnosis of acute<br> pulmonary embolism using a new automated method.<br> Either perfusion scintigrams alone or two different combinations<br> of ventilation/perfusion scintigrams were used<br> as the only source of information regarding pulmonary<br> embolism. A completely automated method based on<br> computerised image processing and artificial neural networks<br> was used for the interpretation. Three artificial<br> neural networks were trained for the diagnosis of pulmonary<br> embolism. Each network was trained with 18 automatically<br> obtained features. Three different sets of features<br> originating from three sets of scintigrams were<br> used. One network was trained using features obtained<br> from each set of perfusion scintigrams, including six<br> projections. The second network was trained using features<br> from each set of (joint) ventilation and perfusion<br> studies in six projections. A third network was trained<br> using features from the perfusion study in six projections<br> combined with a single ventilation image from the posterior<br> view. A total of 1,087 scintigrams from patients with<br> suspected pulmonary embolism were used for network<br> training. The test group consisted of 102 patients who<br> had undergone both scintigraphy and pulmonary angiography.<br> Performances in the test group were measured as<br> area under the receiver operation characteristic curve.<br> The performance of the neural network in interpreting<br> perfusion scintigrams alone was 0.79 (95% confidence<br> limits 0.71–0.86). When one ventilation image (posterior<br> view) was added to the perfusion study, the performance<br> was 0.84 (0.77–0.90). This increase was statistically significant<br> (P=0.022). The performance increased to 0.87<br> (0.81–0.93) when all perfusion and ventilation images<br> were used, and the increase in performance from 0.79 to<br> 0.87 was also statistically significant (P=0.016). The automated<br> method presented here for the interpretation of<br> lung scintigrams shows a significant increase in performance<br> when one or all ventilation images are added to<br> the six perfusion images. Thus, the ventilation study has<br> a significant role in the diagnosis of acute lung embolism. |
| ISBN/ISSN/Other | ISSN: 1619-7070 |
Questions: webmaster
Last update: 2013-04-11
Centre for Mathematical Sciences, Box 118, SE-22100, Lund. Telefon: +46 46-222 00 00 (vx)