Encadrement: Pr. Laurent Heutte* (directeur), Assistant Pr. Caroline Petitjean*, Pr. Luc Thiberville**
"Cette thèse porte sur la classification automatique d'images, appliquée aux images acquises par alvéoscopie, une nouvelle technique d'imagerie du poumon profond. L'objectif est la conception et le développement d'un système d'aide au diagnostic permettant d'aider le praticien à analyser ces images jamais vues auparavant.
Nous avons élaboré, au travers de deux contributions, des méthodes performantes, génériques et robustes permettant de classer de façon satisfaisante les images de patients sains et pathologiques. Nous avons proposé un premier système complet de classification basé à la fois sur une caractérisation locale et riche du contenu des images, une approche de classification par méthodes d'ensemble d'arbres aléatoires et un mécanisme de pilotage du rejet de décision, fournissant à l'expert médical un moyen de renforcer la fiabilité du système.
Face à la complexité des images alvéoscopiques et la difficulté de caractériser les cas pathologiques, contrairement aux cas sains, nous nous sommes orientés vers la classification one-class qui permet d'apprendre à partir des seules données des cas sains. Nous avons alors proposé une approche one-class tirant partie des mécanismes de combinaison et d'injection d'aléatoire des méthodes d'ensemble d'arbres de décision pour répondre aux difficultés rencontrées dans les approches standards, notamment la malédiction de la dimension. Les résultats obtenus montrent que notre méthode est performante, robuste à la dimension, compétitive et même meilleure comparée aux méthodes de l'état de l'art sur une grande variété de bases publiques. Elle s'est notamment avérée pertinente pour notre problématique médicale."
Mots-clefs : Alvéoscopie; aide au diagnostic médical; classification automatique; extraction de caractéristiques; méthodes d'ensemble; arbre de décision; injection d'aléatoire; forêts aléatoires; one-class; out-of-class; synthèse de données; malédiction de la dimension
"This thesis deals with automated image classification, applied to images acquired with alveoscopy, a new imaging technique of the distal lung. The aim is to propose and develop a computer aided-diagnosis system, so as to help the clinician analyze these images never seen before. Our contributions lie in the development of effective, robust and generic methods to classify images of healthy and pathological patients. Our first classification system is based on a rich and local characterization of the images, an ensemble of random trees approach for classification and a rejection mechanism, providing the medical expert with tools to enhance the reliability of the system. Due to the complexity of alveoscopy images and to the lack of expertize on the pathological cases (unlike healthy cases), we adopt the one-class learning paradigm which allows to learn a classifier from healthy data only. We propose a one-class approach taking advantage of combining and randomization mechanisms of ensemble methods to respond to common issues such as the curse of dimensionality. Our method is shown to be effective, robust to the dimension, competitive and even better than state-of-the-art methods on various public datasets. It has proved to be particularly relevant to our medical problem."
Keywords: Alveoscopy; computer aided-diagnosis; automatic classification; feature extraction; ensemble methods; decision tree; randomization; random forests; one-class; out-of-class; data synthesis; curse of dimensionality
Téléchargements (downloads) :
Liste de publications
Brevets
> Closed-loop blood glucose control systems and methods (EP3438986C0)
Automated insulin delivery system for blood glucose control based on physiological model.
> System and method for improving rescue carbohydrate recommendations in diabetes management (US20260157702)
Control device for determining an amount of carbohydrate (RC) recommendation.
> Computerized system for communicating treatment information to a user (EP4243031)
System comprising a dose determination module adapted to determine a possible drug dosage based on said data, a disturbance module adapted to determine whether said data or the possible drug dosage triggers at least a disturbance threshold from a plurality of disturbance thresholds and a notification module.
Publications
2009
- Saint-Réquier, Aurélien, Benoît Lelandais, Caroline Petitjean, Chesner Désir, Laurent Heutte, Mathieu Salaün, and Luc Thiberville. "Characterization of endomicroscopic images of the distal lung for computer-aided diagnosis." In International Conference on Intelligent Computing, pp. 994-1003. Berlin, Heidelberg: Springer Berlin Heidelberg, 2009.
- Petitjean, Caroline, Chesner Désir, Mathieu Salaün, Luc Thiberville, and Laurent Heutte. "Automatic classification of in vivo distal lung images for computer-aided diagnosis." In Medical Image Understanding and Analysis, p. 26. 2009.
2010
- Désir, Chesner, Caroline Petitjean, Laurent Heutte, and Luc Thiberville. "Using a priori knowledge to classify in vivo images of the lung." In International Conference on Intelligent Computing, pp. 207-212. Berlin, Heidelberg: Springer Berlin Heidelberg, 2010.
- Lelandais, Benoît, Chesner Désir, Caroline Petitjean, Laurent Heutte, and Luc Thiberville. "Classification d'images alvéoscopiques par extra-trees." In RFIA, p. xxxx. 2010.
2011
- Petitjean, Caroline, Chesner Désir, Luc Thiberville, G. Lettier, and Laurent Heutte. "Manual vs automatic classification of endomicroscopic images of the distal lung." In Proceedings of International Congress and exhibition on Computer Assisted Radiology and Surgery, p. xxx. 2011.
2012
- Désir, Chesner, Simon Bernard, Caroline Petitjean, and Laurent Heutte. "A random forest based approach for one class classification in medical imaging." In International workshop on machine learning in medical imaging, pp. 250-257. Berlin, Heidelberg: Springer Berlin Heidelberg, 2012.
- Désir, Chesner, Simon Bernard, Caroline Petitjean, and Laurent Heutte. "A new random forest method for one-class classification." In Joint IAPR International Workshops on Statistical Techniques in Pattern Recognition (SPR) and Structural and Syntactic Pattern Recognition (SSPR), pp. 282-290. Berlin, Heidelberg: Springer Berlin Heidelberg, 2012.
- Desir, Chesner, Caroline Petitjean, Laurent Heutte, Mathieu Salaun, and Luc Thiberville. "Classification of endomicroscopic images of the lung based on random subwindows and extra-trees." IEEE transactions on biomedical engineering 59, no. 9 (2012): 2677-2683.
- Désir, Chesner, Caroline Petitjean, Laurent Heutte, Luc Thiberville, and Mathieu Salaün. "An SVM-based distal lung image classification using texture descriptors." Computerized Medical Imaging and Graphics 36, no. 4 (2012): 264-270.
- Hébert, David, Chesner Désir, Caroline Petitjean, Laurent Heutte, and Luc Thiberville. "Detection of pathological condition in distal lung images." In 2012 9th IEEE International Symposium on Biomedical Imaging (ISBI), pp. 1603-1606. IEEE, 2012.
2013
- Désir, Chesner, Simon Bernard, Caroline Petitjean, and Laurent Heutte. "One class random forests." Pattern Recognition 46, no. 12 (2013): 3490-3506.
- Désir, Chesner. "Classification Automatique d'Images, Application à l'Imagerie du Poumon Profond.(Automatic Image Classification, Application to Distal Lung Images)." PhD diss., University of Rouen, France, 2013.
2015
- Gerbelot, Rémi, Anne Koenig, Cedric Goyer, Jérôme Willemin, Chesner Desir, Jean Porcherot, H. Sylla Kane, Régis Guillemaud, Jean Christian Borel, and Pierre Jallon. "A wireless patch for sleep respiratory disorders applications." In 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 2279-2282. IEEE, 2015.
- Desir, Chesner, Maeva Doron, Jean-Christian Borel, Corinne Loïodice, Matthieu Lesgoirres, Régis Guillemaud, and Jean-Louis Pepin. "Le seuil optimal pour détecter un micro-éveil par l’écrasement de l’onde de pouls est différent entre sujets sains et patients SAOS." Médecine du Sommeil 12, no. 1 (2015): 23.
- Heutte, Laurent, Caroline Petitjean, and Chesner Désir. "Pruning trees in random forests for minimizing non detection in medical imaging." In Handbook of Pattern Recognition and Computer Vision, pp. 89-107. 2016.
2016
- Chesner, Desir, Laurent Gerfault, Jean-Christian Borel, Corine Loiodice, Matthieu Lesgoirres, Remi Gerbelot, and Jean-Louis Pepin. "Système expert supervisé pour la détection de micro-éveils autonomiques." Médecine du Sommeil 13, no. 1 (2016): 29-30.
2020
- Villeneuve, Emma, Sylvain Lachal, Chesner Desir, Pierre-Yves Benhamou, Sylvia Franc, Guillaume Charpentier, Erik Huneker, and Maeva Doron. "Increasing the safety of unannounced meal detection for artificial pancreas closed-loop with patient’s hourly meal schedule." In 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), pp. 5093-5096. IEEE, 2020.
2022
- Vambergue, Anne, Madleen Lemaitre, Chesner Desir, Thibault Le Roux Mallouf, Yoursa Tourki, Sylvia Franc, Sandrine Lablanche, Damien Subtil, Guillaume Charpentier, and Pierre Yves Benhamou. "Management of pregnancy in a patient with type 1 diabetes and hypoglycemia unawareness by DBL-hu closed-loop insulin delivery system." Diabetes technology & therapeutics 24, no. 9 (2022): 672-674.
2023
- Gauthier, Pierre, Chesner Desir, Maud Plombas, Eloïse Joffray, Pierre‐Yves Benhamou, and Anne‐Laure Borel. "Impact of sleep and physical activity habits on real‐life glycaemic variability in patients with type 2 diabetes." Journal of Sleep Research 32, no. 3 (2023): e13799.
2024
- Romero, Hector, Alice Adenis, Sylvain Lachal, Chesner Desir, Erik Huneker, and Guillaume Charpentier. "DBLG1 System's efficacy with unannounced meals in real-life situations." in diabetes technology & therapeutics, vol. 26, pp. a94-a95. 140 huguenot street, 3rd fl, new rochelle, ny 10801 usa: mary ann liebert, inc, 2024.
- Hocine, Lebad, Phanie Madrolle, Alice Adenis, Chesner Desir, Erik Huneker, Pierre-Yves Benhamou, and Guillaume Charpentier. "Relationship between daily insulin need, weight and gender." in diabetes technology & therapeutics, vol. 26, pp. a54-a55. 140 huguenot street, 3rd fl, new rochelle, ny 10801 usa: mary ann liebert, inc, 2024.
2025
- Lebbad, Hocine, Chesner Desir, Aurélien Vesin, Alice Adenis, Guillaume Charpentier, and Pierre-Yves Benhamou. "Insulin requirements according to sex and weight in a population of 9036 adult persons with type 1 diabetes using closed-loop insulin delivery." Journal of diabetes science and technology 19, no. 5 (2025): 1257-1263.
- Romero-Ugalde, H.M., Adenis, A., Gauthier, P., Desir, C., Le Roux-Mallouf, T., Huneker, E. and Benhamou, P.Y., 2025. 921-P: Impact of Heart Rate and Step Count on Deep Learning Glucose Prediction Models during Physical Activity. Diabetes, 74(Supplement_1), pp.921-P.
- Romero-Ugalde, H.M., Adenis, A., Cartier-Michaud, T., Gauthier, P., Desir, C., Le Roux-Mallouf, T., Huneker, E. and Benhamou, P.Y., 2025. 920-P: The Relationship between Glycemia and Insulin Sensitivity Factor. Diabetes, 74(Supplement_1), pp.920-P.
- Gauthier, P., Cartier-Michaud, T., Adenis, A., Romeo-Ugalde, H., Lachal, S., Le Roux-Mallouf, T., Desir, C., Huneker, E. and Benhamou, P.Y., 2025, March. New insulin sensitivity model to improve glycemic control. in diabetes technology & therapeutics (vol. 27, pp. e53-e54). 140 huguenot street, 3rd fl, new rochelle, ny 10801 usa: mary ann liebert, inc.
2026
- Gauthier, P., Adenis, A., Lachal, S., Gimenez, P., Desir, C., Le Roux-Mallouf, T., Huneker, E. and Benhamou, P.Y., 2026. 1014-or: heart-rate monitoring for early detection of physical activity–induced hypoglycemia in type 1 diabetes. diabetes, 75(supplement_1), pp.1014-or.
- Adenis, A., Lachal, S., Gauthier, P., Gimenez, P., Desir, C., Le Roux-Mallouf, T., Huneker, E. and Benhamou, P.Y., 2026. 1861-p: performance of DBLG1 system for aerobic, anaerobic, and mixed type physical activities. diabetes, 75(supplement_1), pp.1861-p.
- Adenis, A., Lachal, S., Gauthier, P., Gimenez, P., Desir, C., Le Roux-Mallouf, T., Huneker, E. and Benhamou, P.Y., 2026. 1064-or: independent impact of physical activity characteristics and initial metabolic state on glycemic outcomes in the DBLG1 system: a linear mixed-model analysis. diabetes, 75(supplement_1), pp.1064-or.
- Adenis, A., Lachal, S., Gauthier, P., Gimenez, P., Desir, C., Le Roux-Mallouf, T., Huneker, E. and Benhamou, P.Y., 2026. 1860-p: impact of physical activity on the night outcomes with the DBLG1 system. diabetes, 75(supplement_1), pp.1860-p.
Formations
Formation en management:
- S'approprier le rôle de manager
- Pratiquer un management participatif
- Savoir manager à distance
- Gérer les situations difficiles
- Identifier et développer son leadership
MISC:
- Oeuvres mathématiques du Prof. Jacques Harthong [Archives] ou [Web]
- Le bilan de l'intelligence, Paul Valéry, Ed. Allia, http://goo.gl/eEfhS
- L'équation du nénuphar, Albert Jacquard, Ed. Le Livre de Poche, http://goo.gl/yqjC2
- My Stroke of Insight, Jill Bolte Taylor, Plume, 224 pages [Amazon]
