FM - Teses de Doutoramento
Permanent URI for this collection
Browse
Browsing FM - Teses de Doutoramento by advisor "Almeida, José Crespo Mendes de"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- 3D digital breast cancer surgery with augmented realityPublication . Gouveia, Pedro; Almeida, José Crespo Mendes de; Campos, Maria João de Viseu Botelho Cardoso Aires deBreast conservative surgery (BCS) combined with radiotherapy has become the treatment of choice for the majority of early breast cancer patients. With an identical overall survival rate when compared to mastectomy, BCS is usually associated with a better cosmetic outcome. BCS aims to achieve optimal long-term local control, thus performing excisions with free margins. However, the wider the excision, and thus the possibility of attaining free margins, the higher the risk of poor cosmetic outcomes and breast deformity. If a positive margin is obtained, a re-excision may be needed, potentially impacting breast shape and cosmesis. At least 30% of cases still result in fair/poor aesthetic outcomes. These fair/poor results depend on several factors that have been thoroughly studied. Among these, breast volume, tumor size and location are the most important with a major impact on aesthetic outcome in multivariate analysis. But a single, easy and reproducible method to estimate breast volume is not available that could ultimately help in surgical planning for selecting the optimal BCS technique. Moreover, tumor size and location interpretation rely on the treating physician’s cognitive spatial interpretation of radiology images. New tools and methodologies are needed to improve breast cancer surgical planning and oncological outcomes. In this work, a new quantitative and an inexpensive method to accurately calculate breast volume using a three-dimensional (3D) technology was developed and validated with a depth sensor low-cost surface device (Microsoft Kinect®). Standard methods like breast MRI and mastectomy specimen volumes were used as ground truth, but differences between all methods need further developments to reach clinical applicability. Imaging fusion technology was attempted to harmonize anatomic limits and to improve volume delimitation between breast MRI and 3D surface scan methods. Furthermore, a methodology for breast MRI to 3D surface scan fusion was investigated. A patient-specific digital breast model integrating the real breast torso and the tumor location was created and validated with breast MRI to 3D surface scan fusion algorithm in 16 breast cancer patients. This protocol was used to quantify breast shape differences between different modalities and measure the target registration error of several MRI/3D scan fusion algorithm variants. The fusion of single breasts with a free form deformation model of pose transformation had acceptable registration errors and accurate tumor locations. The performance of the fusion algorithm was not affected by breast volume. Breast spatial interpretation and visualization by the treating physician can thus be augmented with a digital 3D breast model that integrates radiological images, allowing immediate image interpretation of breast cancer lesions. Using previous data, a clinical surgical use case was designed and attempted after successfully production of a 3D digital breast with breast MRI to 3D surface scan fusion. An experimental test with a digital non-invasive method for intraoperative breast cancer localization using augmented reality to guide breast conservative surgery was attempted. A successful overlap of the previous standard pre-operative localization technique (carbon tattooing) and tumor visualization inside the patient’s breast with augmented reality was obtained during surgery. Along the pathway and through computer vision as artificial intelligence, various methods were investigated to produce a unified view of a patient-specific 3D digital breast model, with tumor included. A clinical use case was developed to create an innovative methodology and a proof of concept for a digital non-invasive medical device for intraoperative breast cancer localization.