Résumé
Objectifs
La précision diagnostique de l’échoendoscopie dans le bilan d’extension ganglionnaire des cancers digestifs pourrait être améliorée par l’analyse paramétrique de la texture des images. Nous avons cherché à évaluer les résultats de cette méthode dans une étude prospective.
Méthodes
Cent cinq ganglions péri-digestifs ont été analysés chez 80 malades. Le diagnostic final était obtenu par l’histologie de la pièce opératoire ou par le suivi des patients. Il s’agissait de 41 ganglions inflammatoires bénins et de 64 ganglions métastatiques. Les images étaient obtenues durant une échoendoscopie conventionnelle, numérisées puis transférées sur une station de travail pour l’analyse paramétrique. Deux régions d’intérêt (RI) étaient sélectionnées pour chaque ganglion (ganglion entier et portion périphérique). 21 paramètres étaient étudiés pour chaque RI (2 paramètres de 1er ordre et 19 de second ordre). Les paramètres étaient étudiés par analyse uni- puis multivariée pour définir les paramètres discriminant les ganglions inflammatoires des ganglions métastatiques.
Principaux résultats
Sept paramètres, tous de second ordre, pouvaient distinguer les ganglions inflammatoires et métastatiques de façon indépendante. La sensibilité de l’interprétation par l’endoscopiste (qui possédait les informations cliniques et morphologiques pertinentes) était de 83% et la spécificité de 90%. La sensibilité et la spécificité de l’analyse de texture étaient de 56% et 60%, respectivement.
Conclusions
Ces résultats suggèrent que l’analyse paramétrique de la texture des images ganglionnaires d’échoendoscopie n’améliore pas les performances de l’interprétation visuelle d’un opérateur entraîné disposant des éléments habituels du dossier clinique.
Summary
Background
The diagnostic accuracy of visual interpretation of peri-digestive lymph nodes (malignant vs inflammatory) by endoscopic ultrasonography (EUS) does not exceed 70–80%.
Aims
To assess the ability of ultrasound image texture analysis to improve diagnostic results and to provide objective diagnostic criteria.
Methods
105 mediastinal, intra-abdominal and peri-rectal lymph nodes were analyzed, from 80 patients. The final diagnosis was obtained from histology of the resected specimen or from patient’s history and follow-up. There were 41 inflammatory lymph nodes and 64 malignant lymph nodes. Images were captured during EUS and computerized on a PC, then transferred on workstation for analysis. 2 regions of interest (ROI) were selected for each lymph node: one was the whole node and the other the peripheral rim of the node. 21 parameters were studied for each ROI (2 parameters of the 1st order, 14 from co-occurrence matrices, and 5 from range length matrices). Parameters were analyzed by logistic regression, then by multivariate analysis for discriminating patients in 2 groups: inflammatory vs malignant lymph nodes.
Results
7 parameters were able to discriminate groups by pairs. The sensitivity of the endoscopist’s visual analysis (including knowledge of patient’s history) was 83%, and specificity was 90%. After texture analysis, se was 56%, and sp was 60%.
Conclusion
These results show that texture analysis of lymph node images does not improve the ability of EUS to discriminate between inflammatory and malignant lymph nodes.
This is a preview of subscription content, log in via an institution to check access.
Références
AKAHOSHI K., MISAWA T., FUJISHIMA H.et al. — Preoperative evaluation of gastric cancer by endoscopic ultrasound.Gut, 1991,32, 479–482.
AKAHOSKI K., MISAWA T., FUJISHMA H., CHIJIIWA Y., NAWATA H. — Regional lymph node metastasis in gastric cancer: Evaluation with endoscopic US.Radiology, 1992,182, 559–564.
RODER J.D., BUSCH R., STEIN H.J., FINK U., SIEXERT J.R. — Ratio of invaded to removed lymph nodes as a predictor of survival in squamous cell carcinoma of the esophagus.Br. J. Surg., 1994,81, 410–413.
BOSSET J.F., GIGNOUX M., TRIBOULET J.P., TIRET E., MANTION G., ELIAS D.et al. — Chemotherapy followed by surgery compared with surgery alone in squamous-cell cancer of the esophagus.N. Engl. J. Med., 1997,337, 161–167.
PALAZZO L., ROSEAU G., GAYET B., VILGRAIN V., BELGHITI J., FÉKÉTÉ F., PAOLAGGI J.A. — Endoscopic ultrasonography in the diagnosis and staging of pancreatic adenocarcinoma. Results of a prospective study with comparison to ultrasonography and CT-scan.Endoscopy, 1993,25, 143–150.
HIELE M., DE LEYN P., SCHURMANS P., LERUT A., HUYS S., GEBOES K.et al. — Relation between endoscopic ultrasound findings and outcome of patients with tumors of the oesophagus and oesophago-gastric junction.Gastrointest. Endosc., 1997,45, 381–386.
AMOUYAL G., ROUSEAU G., PALAZZO L., AMOUYAL P., PAOLAGGI J.A. — Détection des adénopathies métastiques par échoendoscopie.La Gazette Médicale, 1990,97, 14–18.
TIO T.L., TYTGAT G.N.J. — Endoscopic ultrasonography in analyzing periintestinal lymph node abnormality.Scand. J. Gastroenterol., 1986,21 (suppl. 123), 158–163.
AIBE T., ITO T., YOSHIDA T. — Endoscopic ultrasonography of lymph nodes surrounding the upper GI tract.Scand. J. Gastroenterol., 1986,21 (suppl 123), 164–169.
ZIEGLER K., SANFT C., ZEITZ M., RIWCKEN E.O. — Preoperative staging of gastrointestinal tumors by endosonography.Surg. Endosc., 1990,4, 154–160.
CATALANO M.F., ALCOCER E., CHAK A., NGUYEN C.C., RAIJMAN I., GEENEN J.E.et al. — Evaluation of metastatic celiac axis lymph nodes in patients with esophageal carcinoma: accuracy of EUS.Gastrointest. Endosc., 1999,50, 352–356.
BERGER, G., LAUGIER P. — Caractérisation tissulaire: échographie quantitative.J. Med. Nucl. Biophy., 1991,15(2), 161–172.
HARALICK R. — Statistical and structural approaches to texture.IEEE proceedings, 1979,67(5), 786–804.
HARALICK R., SHANMUGAN K., DINSTEINN I. — Textural features for image classification.IEEE transactions on systems, man, and cybernetics, 1973,36, 610–621.
ABE C., KAHN C., DOI K., KATSURAGAWA S. — Computer-aided detection of diffuse liver disease in ultrasound images.Intest. Radiol., 1992,27, 71–77.
LERSKI R, SMITH M., MORLEY P., BARNETT E., MILLS P., WATKINSON G., MacSWEEN R. — Discriminant analysis of ultrasonic texture data in diffuse alcoholic liver disease 1. Fatty liver and cirrhosis.Ultrasonic Imaging, 1981,3, 164–172.
INSAMA M., WAGNER R., GARRA B., MOMENAN R., SHAWKER T. — Pattern recognition methods for optimizing multivariate tissue signatures in diagnostic ultrasound.Ultrasonic Imaging, 1986,8, 165–180.
LAUGIER P., BERGER G., GIAT P., BERGER G., BONNIN-FAYET O. LAVAL-JEANTET M. — Ultrasound attenuation imaging in the os calcis: an improved method.Ultrasonic Imaging, 1994,16, 65–76.
SHUM N., ZUNA I., FEIN M. WEISSER G. — Effects of adaptive filtering on the diagnostic ultrasoundin vivo tissue characterization.Eur. J. Ultrasound, 1995,2, 57–64.
ESGIAR A.N., NAGUIB R.N., SHARIF B.S., BENNETT M.K., MURRAY A. — Microscopic image analysis for quantitative measurement and feature identification of normal and cancerous colonic mucosa.IEEE Trans. Inf. Technol. Biomed., 1998,2(3), 197–203.
HERLIDOU S., ROLLAND Y., BANSARD J.Y.et al. — Comparison of automated and visual texture analysis in MRI: characterization of normal and diseased skeletal muscle.Magn. Reson. Imaging, 1999,17(9), 1393–1397.
MAZZONE AM, URBANI MP, PICANO E.et al. —In vivo ultrasonic parametric imaging of carotid atherosclerotic plaque by videodensitometric technique.Angiology, 1995,46(8), 663–672.
Author information
Authors and Affiliations
About this article
Cite this article
Haddad, C., Prat, F., Lefebvre, F. et al. Caractérisation des ganglions lymphatiques péridigestifs par échoendoscopie: intérêt de l’imagerie paramétrique et analyse quantitative. Acta Endosc 33, 127–137 (2003). https://doi.org/10.1007/BF03023682
Issue Date:
DOI: https://doi.org/10.1007/BF03023682