The Control And Optimization Of Light Dose In Photodynamic Therapy For Port Wine Stains

Photodynamic Therapy(PDT), characterized with high selectivity, fading without scar, stable and reliable efficiency, safe and simple operation, and other features, has gradually been a significant clinical treatment for Port Wine Stains(PWS). However, there are still several problems in actual clinical procedures which affect the outcomes of treatment, such as the dependency upon the experiences and skills of doctor, the irradiation conducted by manual operation, and the difficulties in implementing objective, precise and intelligent disease evaluation and light dosage control. Therefore, objective evaluating the lesion and controlling the irradiation objectively and quantitatively may be the only way to realize the individualized PWS therapy. Based on the computer stereovision system and image processing techniques, this thesis built up a set of three dimensional(3D) digital PDT control system consisted of camera and projector to verify the proposed program.Firstly, the 3D digital PDT control system is calibrated to identify the three-dimensional space referenced with the camera, and 3D digital model of a human’s head model as a PWS patient is reconstructed based on the Gray-coded structured light. The positions of lesions and the projector referenced with the camera can be calculated, and 3D cloud points of lesions with the accuracy of 1mm are extracted and acted as 3D digital quantitative model for the optimal control of light dosage and clinical evaluation.Then, the projector can illuminate the lesions automatically according to the calibration parameters of the visual system, which not only achieves a better match between the illumination region and lesions area but also relieve the control for curved lesions. In addition, a weighed optimization method for irradiance based on normal vectors of the lesions was proposed, which can solve the reasonable distribution of the irradiance received by lesions in a single therapy, and is beneficial to improve the treatment efficacy and assist doctors in treatment planning.Finally, considering the heterogeneity of PWS lesions, the super-pixel clustering algorithm based on color similarity and spatial proximity is applied to finely divide the lesions, and the lesions images of seven patients selected from the Phase II clinical study for Hemoporfin are used to verify the correctness. With the color bar tied on the skin of patients, the severity of lesions is graded locally and subtly, and the therapeutic effect in different courses is evaluated and compared, which provides an approach for the dose-effect evaluation.The proposed system provided a new approach for the lesion localization, irradiation control and lesions assessment, and laid the foundation of future engineering practice for the individual and accurate treatment in photodynamic therapy for PWS.