Les tumeurs des voies aérodigestives sont très diverses, tant sur le plan de leur localisation, que sur le plan histologique. Aussi, cette grande hétérogénéité pose particulièrement des problèmes lors du diagnostic histologique mais aussi pour la mise en place du traitement le plus adapté. Ainsi, le réseau du REFCOR (réseau d’expertise français sur les cancers ORL rares) a été créé pour mieux appréhender ces différentes problématiques en proposant une approche épidémiologique, diagnostique ainsi que des collaborations de recherche. Ce réseau est dédié à toutes les tumeurs malignes primitives des glandes salivaires, de l’oreille, des fosses nasales et des sinus et à toutes les tumeurs malignes des VADS autres que les carcinomes épidermoïdes conventionnels. Le réseau REFCORpath est constitué de pathologistes experts, qui proposent sur un site dédié avec lames numérisées un deuxième avis, voire un troisième.Aerodigestive tract tumors are very diverse, either in terms of location, or histologically. Also, this heterogeneity poses particular problems for the histological diagnosis but also for the establishment of the most appropriate treatment. Thus, the network REFCOR (réseau d’expertise français sur les cancers ORL rares/French expert network on rare ENT cancers) was created to better understand these issues, by proposing an epidemiological and diagnostic approach with research collaborations. This network is dedicated to all primary malignant tumors of the salivary glands, ear, nasal cavity and sinuses and all head and neck malignancies other than conventional squamous cell carcinoma. The REFCORpath network consists of expert pathologists and offers, through a network of scanned images, a second opinion or even a third.

Introduction. - To assess interobserver variability for biological target volume (BTV) delineation and to compare the reproductibility of different semiautomatic segmentation methods in pretreatment 18-fluorodeoxyglucose positron emission tomography (PET/CT) of head and neck squamous cell carcinoma (HNSCC).Patients and methods. - Patients with histologically proved HNSCC referred to the nuclear medicine service in Brest for pretreatment PET/CT were prospectively included from February 2009 to June 2010. Three nuclear medicine physicians (two specialized in oncology) delineated manually and independently BTV on each primary tumor. Four semiautomatic segmentation methods have been studied; three using a fixed threshold and one applying an adaptive threshold based on the signal-to-background ratio (Daisne). The variability between kappa observers and/or methods has been assessed. The concordance between the various BTV intersections and unions has been also assessed.Results. - Thirty patients (29M; 1F) were included. The primary site location was oropharynx in six patients, oral cavity in 10 patients, hypopharynx in five patients and larynx in nine patients. A statistically significant global interobserver variability (P = 0.01) was showed, but without statistically difference between the two experienced oncologists (P = 0.15). The maximal concordance of the two experienced observers with the semiautomatic methods was found for the Daisne method (CI = 61.5%; kappa = 0.68), expressing a good agreement according to the Landis and Koch criteria, better than with the segmentation method using a fixed threshold with 40% of maximal signal intensity (CI = 52.1%; kappa = 0.53).Conclusion. - Our results suggest the feasibility of achieving HNSCC BTV delineation by PET/CT using semiautomatic methods, in particular those which apply an adaptative threshold but under the supervision of an experienced operator. (C) 2012 Published by Elsevier Masson SAS.

Introduction: Estimate operator-related variability of delineation of biological target volume (BTV) and compare reproducibility of some semiautomatic methods of segmentation in fludeoxyglucose positron emission tomography (FDG-PET) during the care of head and neck squamous cell carcinoma (HNSCC).

Current state of the art algorithms for functional volume segmentation in PET images for diagnosis, patients follow-up or radiotherapy treatment planning consist of adaptive threshold approaches. We have developed an unsupervised Bayesian segmentation algorithm for tumors in PET, namely the FLAB (for Fuzzy Locally Adaptive Bayesian) algorithm, that was previously validated on simulated images and then successfully extended and applied to inhomogeneous and non spherical real clinical images of lung tumors. In this study, we investigate the robustness of this approach in comparison to fixed threshold based approaches at 42% and 50% of the maximum value, as well as another automatic segmentation algorithm (the fuzzy C-means clustering). For this investigation, a series of IEC phantom acquisitions were performed on different PET/CT scanners (Philips Gemini, Siemens Biograph and GE Discovery LS) and reconstruction algorithms (RAMLA, OSEM). Various acquisition parameters were considered in each case, like the size of the voxels in the reconstructed image, the contrast ratio and the acquisition duration (hence the intensity of the noise). The purpose of this study was to study the robustness of each approach with respect to scanner model, reconstruction algorithm, and various acquisition parameters. For all the scanner types and reconstruction algorithms considered in this preliminary study the FLAB algorithm demonstrated globally higher accuracy in delineation of the spheres. In addition, the FLAB results showed lower variability with respect to changes in the acquisition parameters and/or the scanner model and associated reconstruction algorithm, therefore demonstrating its higher robustness compared to the other semi-automatic or automatic approaches considered. Future studies will investigate the results of the same approaches with GE DST and Time of flight scanners, as well as the inclusion of the adaptive thresholding approaches.

In this work we determine conditions to produce cell samples for imaging with detection of the modification of the magnetic field by maghemite (Fe2O3) nanoparticles acting as a high sensitivity magnetic bio-sensor based on the giant magneto-impedance (GMI) effect. Mat Ly Lu cells are grown for 24h with various maghemite nanoparticles concentrations (from 0 to 6mg/ml). The percentage of viable cells is determined by counting labeled cells with trypan blue under an optical microscope. The quantity of nanoparticles internalized into the cells is evaluated by X-ray fluorescence analysis and expressed in iron moles per cell. The GMI bio-sensor was tested with the various samples. We observed that the best sensitivity of the GMI bio-sensor was obtained at a frequency of 1MHz. To confirm these results in the presence of cell samples, four measurement frequencies were pre-selected (from 1 to 100MHz) and tested. Cell growth conditions compatible with an acceptable percentage of cell viability for various concentrations of nanoparticles were also determined. These experiments allow us to conclude that cell growth with 0.1mg/ml of nanoparticles for 24h shows modifications of the magnetic field detectable optimally at 1MHz frequency.

Les mouvements respiratoires réduisent les performances quantitatives et qualitatives de l’imagerie d’émission. Leur impact est devenu maintenant évident avec l’utilisation des machines hybrides, avec lesquelles des différences entre les conditions respiratoires d’acquisition des images anatomiques et des images fonctionnelles peuvent induire des artéfacts significatifs. Actuellement la solution proposée pour prendre en compte ces effets est de synchroniser les acquisitions sur la respiration. Alors que de telles acquisitions peuvent effectivement réduire l’impact des mouvements respiratoires, le bénéfice est limité car seulement une partie des données disponibles est réellement utilisée pour reconstruire chaque image corrigée. Les approches proposées pour corriger les différences liées à la respiration dans les images synchronisées sont de deux types, selon qu’elles sont appliquées aux images ou aux données brutes. Les méthodes appliquées aux images utilisent des algorithmes de recalage qui réalignent les images synchronisées pour ensuite les sommer, alors que les méthodes appliquées aux données brutes incorporent, avant ou pendant la reconstruction de celles-ci, des transformations modélisant les différences dues à la respiration entre les images. Les travaux déjà réalisés dans ce dernier domaine ont montré que les modèles locaux de transformation non rigide donnent alors de meilleurs résultats que les transformations affines pour prendre en compte les mouvements respiratoires survenus entre les images d’une série synchronisée sur la respiration. De plus, en incorporant une transformation lors du processus de reconstruction, on obtient un contraste dans les images finales qui est meilleur que celui qui est obtenu avec une méthode de correction appliquée aux images.

Évaluer la faisabilité de la tomoscintigraphie myocardique dix minutes après effort par comparaison à l'acquisition classique à 60 minutes. Cent quatre patients adressés pour analyse de la perfusion myocardique ont été inclus avec deux acquisitions posteffort à dix et 60 minutes après injection de sestamibi-Tc99m. La perfusion myocardique a été analysée visuellement en insu par deux médecins experts. Six patients ont été exclus secondairement en raison d'une importante activité digestive : un patient à dix minutes, trois au deux temps, deux au repos. Deux groupes sont définis : groupe 1(G1) = scintigrammes normaux à dix et 60 minutes (n = 53) ; groupe 2 (G2) = scintigrammes anormaux à dix et/ou 60 minutes (n = 45). La fraction d'éjection ventriculaire gauche (FEVG) et les volumes (VTD, VTS) ont été calculés en utilisant le logiciel QGS du Cedars-Sinai. Scores perfusionnels : G1 dix minutes = 0,4 versus G1 60 minutes = 0,4 (p = NS) ; G2 dix minutes = 10,2 versus G2 60 minutes = 10,1 (p = NS) ; FEVG G1 dix minutes = 71 ± 11 % versus G1 60 minutes = 68 ± 10 % (p = 4E-04) ; FEVG G2 dix minutes = 56 ± 15 % versus G2 60 minutes = 53 ± 15 % (p = 0,002) ; VTD G1 dix minutes = 72 ± 20 ml versus G1 60 minutes = 76 ± 19 (p = 0,002) ; VTD G2 dix minutes = 98 ± 46 ml versus G2 60 minutes = 105 ± 38 (p = 0,08) ; VTS G1 dix minutes = 22 ± 12 ml versus G1 60 minutes = 25 ± 12 (p = 9E-04) ; VTS G2 dix minutes = 47 ± 35 ml versus G2 60 minutes = 53 ± 36 (p = 3E-04). Cette étude montre que les acquisitions dix minutes après injection de sestamibi-Tc99m sont réalisables en routine avec une qualité d'image et une analyse de la perfusion myocardique identiques à la méthode classique. Our aim was to assess the feasibility of early acquisition (10 min) gated single photon emission computed tomography (SPECT) by comparison to conventional imaging at one hour. One hundred and four patients referred for exercise test and SPECT were included. Sequential imaging was started 10 min (SPECT 10) and 60 min (SPECT 60) after injection of the radiotracer (Tc-99m sestamibi). Stress myocardial perfusion was visually analyzed from 10 to 60 min stress by two experienced nuclear-cardiologists. Six patients were further excluded, because of high digestive accumulation: one patient at 10 min, three at 10 and 60 min, two at rest. The participants were classified as follows: group G1 (normal SPECT 10 and 60, n = 53), group G2 (abnormal SPECT 10 and/or SPECT 60, n = 45). The left ventricle ejection fraction (EF) and volumes (end-systolic and end-diastolic volumes, ESV, EDV) were calculated with the Cedars-Sinai program. Quality imaging was the same at SPECT 10 min and SPECT 60 min. Perfusion scores: G1 10 min = 0,4 versus G1 60 min = 0,4 (p = NS); G2 10 min = 10,2 versus G2 60 min = 10,1 (p = NS); EFG1 10 min = 71 ± 11% versus EFG1 60 min = 68 ± 10% (p = 4E-04); EFG2 10 min = 56 ± 15% versus EFG2 60 min = 53 ± 15% (p = 0,002); EDV G1 10 min = 72 ± 20 ml versus EDV G1 60 min = 76 ± 19 (p = 0,002); EDV G2 10 min = 98 ± 46 ml versus EDV G2 60 min = 105 ± 38 (p = 0,08); ESVG1 10 min = 22 ± 12 m versus ESV G1 60 min = 25 ± 12 (p = 9E-04); ESVG2 10 min = 47 ± 35 ml versus ESV G2 60 min = 53 ± 36 (p = 3E-04). This study suggests that early gated Sestamibi SPECT after stress provides same perfusion analysis than classical late imaging.

A method is proposed to synchronize positron emission tomography (PET) list-mode data with an externally recorded respiratory signal in the absence of a master clock. When the respiratory signal reaches a user-defined threshold, a trigger mark is stored in the list-mode file. After the acquisition, synchronization is achieved when the stored trigger marks are superimposed on the respiratory curve to form a horizontal line over time at the user-defined threshold. Synchronization was possible and unequivocal for ten out of ten clinical studies. The list-mode acquisition actually started approximately 40 and 4 s after acquisition initiation at the user interface of the Philips Gemini and the GE DLS PET-CT systems, respectively.

A newly developed simulation toolkit, GATE (Geant4 Application for Tomographic Emission), was used to develop a Monte Carlo simulation of a fully three-dimensional (3D) clinical PET scanner. The Philips Allegro/GEMINI PET systems were simulated in order to (a) allow a detailed study of the parameters affecting the system's performance under various imaging conditions, (b) study the optimization and quantitative accuracy of emission acquisition protocols for dynamic and static imaging, and (c) further validate the potential of GATE for the simulation of clinical PET systems. A model of the detection system and its geometry was developed. The accuracy of the developed detection model was tested through the comparison of simulated and measured results obtained with the Allegro/GEMINI systems for a number of NEMA NU2-2001 performance protocols including spatial resolution, sensitivity and scatter fraction. In addition, an approximate model of the system's dead time at the level of detected single events and coincidences was developed in an attempt to simulate the count rate related performance characteristics of the scanner. The developed dead-time model was assessed under different imaging conditions using the count rate loss and noise equivalent count rates performance protocols of standard and modified NEMA NU2-2001 (whole body imaging conditions) and NEMA NU2-1994 (brain imaging conditions) comparing simulated with experimental measurements obtained with the Allegro/GEMINI PET systems. Finally, a reconstructed image quality protocol was used to assess the overall performance of the developed model. An agreement of <3% was obtained in scatter fraction, with a difference between 4% and 10% in the true and random coincidence count rates respectively, throughout a range of activity concentrations and under various imaging conditions, resulting in <8% differences between simulated and measured noise equivalent count rates performance. Finally, the image quality validation study revealed a good agreement in signal-to-noise ratio and contrast recovery coefficients for a number of different volume spheres and two different (clinical level based) tumour-to-background ratios. In conclusion, these results support the accurate modelling of the Philips Allegro/GEMINI PET systems using GATE in combination with a dead-time model for the signal flow description, which leads to an agreement of <10% in coincidence count rates under different imaging conditions and clinically relevant activity concentration levels.

Objectif : Etudier l'impact de la methode de reconstruction tomographique sur l'analyse de la viabilite myocardique. Materiels et methodes : Une etude preliminaire sur fantome thoracique a permis de definir des parametres de reconstruction soit analytique par retroprojection filtree (FBP) (filtre Butterworth, frequence de coupure a 0,4 cycles/pixel, ordre 5) soit iterative par OSEM (10 iterations, 8 subsets pour le 9 9 m Technetium ; 5 iterations, 4 subsets pour le 2 0 1 Thallium) et d'en etudier la resolution spatiale. Ces parametres ont ensuite ete utilises pour reconstruire des tomoscintigraphies de repos de 95 patients (79 hommes, 16 femmes, 64 ans ′ 11,5, 51 au 9 9 m Tc-MIBI, 44 au 2 0 1 Thallium) suivis au CHU de Brest de mars 2003 a decembre 2004. Les discordances de quantification observees par l'analyse QPS® entre les deux methodes avec un seuil de viabilite choisi a 50% ont ete analysees. Un p≤0,05 etait considere comme significatif. Resultats: La resolution spatiale etait globalement meilleure avec la methode iterative. La quantification obtenue par QPS® etait sensiblement plus elevee avec FBP. Les discordances observees entre les deux methodes de reconstruction sont non negligeables avec des resultats statistiquement plus eleves en FBP (10 - 6 ≤p≤0,033). Les discordances affectaient plus la paroi inferieure. Le taux de discordances observees n'etait pas modifie par l'indice de masse corporelle. Conclusion: OSEM semble plus performante que FBP en terme de resolution spatiale. Elle pourra presenter un interet certain dans l'analyse de viabilite myocardique. Nous obtenons des resultats quantitativement plus faibles selon OSEM par rapport a FBP et supposons que le seuil optimal de OSEM devrait etre plus bas.

Visvikis, D.a; Turzo, A.a; Barret, O.b; Fryer, T.D.b; Carson, K.c; Bizais, Y.a; Jarritt, P.; Cheze Le Rest, C.a Author Information

U650 INSERM, Laboratoire du Traitement de l'Information Medicale, University Hospital Medical School, Brest, France Abstracts of the spring 2005 meeting of the British Nuclear Medicine Society Manchester International Convention Centre, UK; 14–16 March, 2005