The purpose of this project was to develop a delivery system optimized for laser interstitial thermotherapy of small tumors of the breast. The proposed approach is to combine laser interstitial thermotherapy with stereotactic imaging for fiber guidance and treatment monitoring. The goals of the dissertation were to design a fiber insertion system for cylindrical diffusing tip optical fibers and to derive optimal laser parameters for coagulation of 1 cm tumor plus a surrounding 1 cm thick rim of healthy tissue. A fiber insertion system compatible with a high resolution stereotactic digital X-ray biopsy system was designed to guide the fiber into the tumor site in similar fashion to the insertion of the biopsy needle. An optical-thermal model consisting of a radiation model, a thermal model, and a coagulation model was developed and validated using ex-vivo porcine tissue. A single integrating sphere optical property measurement system and an inverse Monte Carlo algorithm were developed to measure the optical properties of ex-vivo porcine tissue at 830, 940, and 980 nm. An experimental method was developed to determine the parameters of the Arrhenius model (frequency factor (A) and activation energy (Ea)). The optical-thermal model was validated by comparing the predicted temperature and coagulation to results of laser irradiation experiments at 830, 940, and 980 nm. Using published values of the optical properties of the breast, the model predicts that a 3 cm coagulation size can be produced without vaporization in 10 min with 10.4 W at 980 and 940 nm and 13.2 W at 830 nm. The same outcome can be achieved in 20 min with 4.5 W at 980 and 940 nm and 6.1 W at 830 nm.

L’impact laser obtenu avec les systèmes actuels de photocoagulation laser de la rétine n’est pas maitrisé. L’enjeu est d'obtenir un confinement 3D de l'impact laser, c’est-à-dire de contrôler son positionnement et son extension, pour éviter de porter atteinte aux couches saines de la rétine. Un tel confinement pourrait être réalisé si l’on disposait d’un système laser plus ouvert et d’un dispositif permettant de corriger en temps réel les aberrations de l'œil. L’optique adaptative permet une telle correction ; cette technique est utilisée depuis une vingtaine d’années dans le domaine du diagnostic (imagerie). Cependant, sa mise en œuvre pour des applications thérapeutiques, telles que la photocoagulation laser, nécessite encore d’en améliorer la robustesse : alors qu’en imagerie il suffit de reprendre une image si la précédente est mauvaise, le confinement de l’impact laser doit être assuré en permanence durant toute la procédure chirurgicale. Le but de la thèse est de guider le développement ultérieur d’un système d’optique adaptative médical visant à assister un système de photocoagulation laser de la rétine. Le dimensionnement d’un tel système doit s’appuyer sur une solide connaissance des aberrations oculaires. Ainsi, une étude hautement résolue temporellement et spatialement des aberrations dynamiques de l’œil de grande ampleur a été réalisée et de cette étude, des conclusions sur le dimensionnement d’un système d’optique adaptative pour l’œil ont été tirées. Par ailleurs, un banc de test d’optique adaptative haute cadence a été conçu et intégré pour se confronter aux problèmes pratiques que pose la mise en place d’un système d’optique adaptative pour l’œil. / The laser lesion delivered by current retinal photocoagulation laser systems is not well controlled. The issue is to get a better confinement of the laser lesion that is to control the size of the laser lesion as well as the laser lesion positioning in the retina, in order to prevent any retinal damage. Such a confinement could be reached if the laser system featured a higher numerical aperture and was associated with a real-time correction of the aberrations of the eye. Adaptive optics gives access to such a correction; this technique has been used for the past twenty years for diagnosis (retinal imaging). However, further work is still to be achieved to improve the robustness of current adaptive optics systems before implementing adaptive optics in therapeutic systems, in particular in retinal laser photocoagulation systems. Indeed, unlike imaging where the acquisition can be repeated as many times as necessary, the confinement of the laser lesion must be maintained over time during the whole laser treatment. In this thesis, we provide guidance for the future design of an adaptive optics system for retinal photocoagulation. Such a design has to rely on a thorough knowledge of the ocular aberrations to correct. Thus, a highly temporally as well as spatially resolved aberrometry study on a large population was performed and conclusions on the design of an adaptive optics system for the eye were drawn from this study. Besides, a test bench was designed and set up to face the practical problems coming with the implementation of an adaptive optics system for the eye.