PhD Defense by Marco Esposito
- Speaker: Marco Esposito
- Date: Wednesday, November 13, 2019
- Time: 10:00
- Location: FMI-Building, Room 00.12.019
Seamless Intraoperative Robotic Imaging
Medical imaging has assumed a major role in modern medicine. The ability to precisely ascertain the cause and origin of the patient's condition before starting surgical treatment has removed a large part of the guesswork and imprecision that were previously involved. But also during the surgery, the adoption of imaging technologies has helped reducing the invasivity of the procedures and improving the short- and long-term clinical outcome.
However, the introduction of more and more tools into the operating room has also contributed to increase the complexity of modern surgery. Today, becoming a surgeon also involves learning to use the available technologies effectively and cope with their idiosyncrasies, while keeping focused on the actual objective: healing the patient. An optimal trade-off in this regard can be challenging to achieve, and the clinical outcome can be negatively affected by the cognitive burden and stress posed on the physician.
Such scenarios can occur when the surgeon needs combined anatomical and functional information to perform a surgery with success. Most intraoperative imaging modalities take the form of hand-held devices, each providing its own output image according to its position in space. The physician must hold both devices at the same time in order to scan the same region of space, and mentally orient and merge their output in order to identify structures of interest. One or both devices must then be put aside before carrying out the actual task, based on the acquired knowledge. We explore how collaborative robotic imaging can simplify this picture, by holding one of the devices for the physician and working in concert with them while producing one single, real-time multimodal image which is ready for visualization and can be provided at all times.
An additional obstacle for medical personnel can be represented by fickle equipment that poses many constraints on its handling. Freehand imaging techniques are subject to the limitations of currently available tracking technologies, such as line-of-sight requirements or the degradation of accuracy across the workspace or due to environmental factors. The operator is then required to prevent these problematics by constraining their own actions while pursuing the clinical objective. Failure to do so can compromise the system's performance or force to repeat the image acquisition altogether. We investigate the possibility to mitigate these issues by employing mathematical tools to partially overcome the precision limitations of electromagnetic tracking and make freehand 3D ultrasound more robust and reliable to use.
| WebEventForm |
| Title: || PhD? Defense by Marco Esposito |
| Date: || 13 November 2019 |
| Location: || FMI-Building, Room 00.12.019 |
| Abstract: || In this dissertation, we explore innovative techniques for intraoperative medical imaging. First, we augment ultrasound-guided needle biopsy with functional information provided by a collaborative robotic arm holding a gamma camera. We evaluate its effectivness for sentinel lymph node biopsy in the axilla. Second, we propose a novel Riemannian regularization method for decreasing the jitter present in the output of electromagnetic tracking systems, which are standard for freehand 3D ultrasound. |
| Imageurl: || |
| Type: || News |
| Videourl: || |
| Conferencelink: || |