Prototype of video endoscopic capsule with 3-d imaging capabilities.

Area of Science:

• Medical Devices
• Computer Vision
• Gastroenterology

Background:

• Wireless video capsules are used for gastroenterological examinations, but current systems require post-examination analysis.
• Integrating vision algorithms into video capsules could enable direct, in situ diagnosis.
• Implementing 3-D imaging techniques is a crucial step towards in situ diagnosis.

Purpose of the Study:

• To develop an embedded active vision system (Cyclope project) for wireless video capsules capable of real-time 3-D and texture data acquisition.
• To address the challenges of size, power consumption, and processing limitations inherent in video capsule technology.
• To create a wireless multispectral vision sensor for real-time 3-D scene reconstruction transmission.

Main Methods:

• Hardware and software development of a wireless multispectral vision sensor.
• Implementation of a Field-Programmable Gate Array (FPGA)-based prototype for proof of concept.
• Laboratory, in vitro, and in vivo (pig model) experiments to evaluate the 3-D vision system's performance.

Main Results:

• Successful development of a wireless multispectral vision sensor prototype.
• Demonstration of real-time 3-D reconstruction capabilities.
• Validation of the system's performance through various experimental settings.

Conclusions:

• The Cyclope project successfully demonstrates the feasibility of an embedded 3-D vision system for wireless video capsules.
• The developed system has the potential to significantly enhance gastroenterological diagnostics by enabling real-time in situ analysis.
• A roadmap for integrating this 3-D vision technology into future wireless capsules is established.