University of Patras
Σύστημα προεγχειρητικού σχεδιασμού χειρουργικής ογκολογίας ήπατος προσανατολισμένο για χρήση από χειρουργούς
Abstract
dc:descriptionThe liver is an organ where benign or malignant tumors can develop, either primary or, more frequently, metastatic. For a significant number of liver tumors, hepatectomy is the treatment of choice. The primary goal of all oncological surgical interventions on the liver is to remove the tumor with negative resection margins (R0 resection) while leaving behind a functional liver remnant (Future Liver Remnant - FLR) of sufficient size to prevent postoperative liver failure. Modern liver surgery leverages the capabilities provided by computer technology to gather as much preoperative information as possible for planning the intervention and to use this data intraoperatively to optimize and enhance the safety of the procedure. Detailed preoperative study is key to the proper selection of patients. The segmentation of anatomical structures produces personalized 3D models that can be modified, managed, and displayed in various ways, making surgical analysis easier by allowing rapid volumetric assessment and simulation of hepatectomies. It has been shown that liver segmentation from medical images is the superior method for preoperative planning and intraoperative guidance in liver surgery. However, liver segmentation is a time-consuming and complex process. A modern liver surgeon would greatly benefit from an application that allows them to use medical imaging CDs/DVDs brought by their patients to easily, quickly, and reliably perform preoperative planning on their personal computer. Taking the above into account, we designed and developed liver segmentation algorithms as well as a comprehensive system for preoperative planning and intraoperative guidance tailored to the philosophy of liver surgeons. This system is called "Pre-Operative Mapping and Evaluation of Tumors for Hepatic surgery and Enhanced Intra-operative Assistance" (PROMETHEIA®). The development and evaluation of our software took place in a MATLAB® environment (The MathWorks, Inc., Natick, Massachusetts, U.S.A) on a 2010 home-use laptop computer (Intel® CoreTMDuo, 2.2 GHz, NVIDIA® Graphics, 4GB Memory, Microsoft Windows® 7 64bits).PROMETHEIA was used for the segmentation of anatomical and pathological structures, the preoperative planning of hepatectomies, and intraoperative guidance. We developed specialized tools and algorithms enabling hybrid semi-automatic segmentation and hybrid interactive segmentation. Additionally, we created specialized tools for perioperative risk assessment. We evaluated the capability to perform computer-assisted oncological liver surgeries using PROMETHEIA in prospective clinical studies. The application enables comprehensive and complete preoperative planning, including segmentation of all structures of interest, volumetric analysis, virtual hepatectomy, and risk assessment. It also facilitates its full use in the operating room for intraoperative guidance and risk assessment, with the ability to add new data and easily reevaluate the surgical plan. The segmentation algorithms can be used by surgeons in daily clinical practice with ease and speed, as their parameterization is simple and their use is guided, providing repeatability and reliability. In our studies, surgical analysis in PROMETHEIA was effectively performed in all cases by specialized surgeons within a time frame that permits its use in daily clinical practice, delivering reliable volumetric results and topographical assessments, thus enabling safe surgical planning, perioperative risk assessment, and intraoperative guidance. Radiologists must collaborate with surgeons to provide their expertise and experience in analyzing imaging studies. Our system bridges the gap between overly detailed computer-assisted surgical analysis performed by specialized personnel with expensive software and hardware and the busy liver surgeons with limited knowledge of medical image processing techniques who seek quick, reliable, and cost-effective results while having sufficient expertise to evaluate them appropriately. Through our study, we were able to highlight techniques and tools that are useful for adequate segmentation of medical images in liver-related interventions. It appears that only a few tools are needed for volumetric analysis, but more specialized tools are required for a complete perioperative risk analysis. The research conducted within the framework of this dissertation contributes to the advancement of preoperative planning in oncological liver surgery in daily clinical practice, aiming for faster decision-making, shorter patient management times at a low cost, safer and more effective hepatectomies, and ultimately, the optimization of liver tumor management, thereby striving for patient cure. In conclusion, our application proves to be accurate and reliable in the preoperative planning of oncological liver surgery procedures while effectively enhancing perception and guidance from medical images during operations. However, the design results must always be carefully evaluated by specialized liver surgeons.
Degree
thesis:*- Grantor dc:publisher
- University of Patras
- Year dc:date
- 2016
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Zygomalas, Apollon
- Ζυγομαλάς, Απόλλων
Subjects
dc:subject × 26- Χειρουργική
- Ήπαρ
- Υπολογιστές
- Ογκολογία
- Προεγχειρητικός σχεδιασμός
- Ογκομέτρηση
- Κατάτμηση εικόνων
- Liver
- Surgery
- Oncology
- Preoperative
- Image segmentation
- Volumetry
- Computer aid
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμες Μηχανικού και Τεχνολογία
- Υπολογιστές, Υλικό (hardware) και Αρχιτεκτονική
- Επιστήμες Υγείας
- Ιατρική και Επιστήμες Υγείας
- Ιατρική Πληροφορική
- Electrical Engineering, Electronic Engineering, Information Engineering
- Engineering and Technology
- Computer science, Hardware and Architecture
- Health Sciences
- Medical and Health Sciences
- Medical Informatics
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/57976