Improvements in Image Compression and RF Transmission for Wireless Capsule Endoscopy

Abstract

Endoscopes and colonoscopes are used by many doctors and physicians for diagnoses of pathologies of the gastrointestinal (GI) track. These methods are painful and dangerous for the life of the patients, due to these drawbacks: a wireless capsule endoscopy (WCE) is used. The reason is that diagnosing the small intestine is impossible those the WCE improves the overall performance of endoscopy and colonoscopy. The use of these systems provide benefits in the diagnosis of pathologies of the GI such us, cancer, blooding, lymphoma or Cronh’s diseases. WCE has an image sensor that captures images. The amount of data produced by the image sensor is huge. Also, there is the need of wireless transmission of this data. In order to power these systems by a single battery, the system should use compression of this data and an extra low power transmitter with high data rate transmission. The energy of this battery is currently not sufficient to keep the system working for more than 8 hours continuously with conventional compression algorithms and wireless transmitters. For these reasons, mainly, the use of a novel design that can compress and send efficiently this data is needed. In this research, an entirely improved WCE system is proposed. From the perspective of camera lenses that can be used, through camera sensor and compression algorithm, to wireless transmission system. Hundreds of WCE images were used to identify the unique characteristics of them. Statistical analysis was made and finally achieved the best of the performance for the compression algorithm. Additionally, camera lenses where selected to provide a new perspective of view. Traditionally, WCE uses simple lenses to capture the GI track. A 360º degree lenses is proposed to be used. The reason is the fact that the use of them gives the advantage to map the entire GI track and make a virtual reconstruction of the GI outside human body. With this method, the physician/doctor could be able to walk-through in the entire GI track to make a diagnosis. A compression algorithm designed and implemented. Firstly, evaluation of proposed schemes, of the realized by the help of Matlab software. The compression algorithm that was suitable for a medical application like WCE, were implemented in hardware. The implementation in hardware is made with help of an FPGA development board using Hardware Description Language (HDL). Finally, the proposed compression algorithm was designed in Cadence tool-set software in 16nm FinFet technology. The proposed compression algorithm is designed to be able to connect directly to most common commercially available camera sensors. Furthermore, the output of the compression algorithm is not parallel, but the output data is in serial mode. This feature, makes the proposed system to be connected directly to any RF transmitter that is available now. The proposed system except from the compression performance, offers a compact system that can work as it is, without any modifications with most common used camera sensors and RF transmitters. Apart from the compression algorithm, the design and simulation of a low energy wireless transmitter was made. Digital modulation, On-Off Keying (OOK) was used. The selection of the modulation and the working carrier frequency was based in the body characteristics and the performance in the manner of biterror-rate (BER) and low energy usage. A low power design of a specially implemented OOK transmitter is designed and simulated. The wireless transmitter is designed and surface optimized by using fewer internal and external parts, so it can fit inside the capsule. The design and simulation of the proposed wireless transmitter was done with the Cadence tool-set in 16nm FinFet technology. Furthermore, there is another field of research in WCE. Until now, the research in WCE was focused on the internal part of the WCE, the pill. In our research project, the main effort was focused not only on the internal part of the WCE but also in the external part. The external part consists of a receiver, and a recorder, both of them placed in a jacket that the patient wears. In this research work a multi-receiver system is proposed. Existed systems use only one receiver and this introduces the possibility to receive false or corrupted data. For this reason, existing systems use bi-directional RF transmission link. While using bidirectional RF transmission link in the capsule, both transmitter and receiver should be placed. This needs more space and consumes more energy for operation. The proposed multi-receiver system eliminates the need of bi-directional link. Less space and power inside the capsule needed because there is no receiver inside the capsule to operate. Multi-receiver system is used also to increase the BER performance of the wireless link. While using multi-receivers, weak or corrupted reception of RF signal is eliminated because always there will be receivers with good reception that can capture the correct data. Furthermore, error correction algorithm is used that compares the received data of all the receivers which improves the decision making method for the transmitted data. Finally, the entire WCE with a set of special lenses with a 360º degree angle of view is proposed. High resolution images can be captured and transmitted wirelessly. The set of special lens, low complexity and low energy compression algorithm, along with the use of the high data rate, low power transmitter, pushes the WCE technology to the next level.