Medical images are often characterized by high complexity and consist of high resolution image files, introducing thus several issues regarding their handling. Current compression schemes produce high compression rates, sacrificing however the image quality and leading this way to unenviable examination. Region of Interest (ROI) coding has been introduced as an efficient technique for addressing such issues, by performing advanced image compression and preserving quality in diagnostically critical regions. This chapter discusses the basic ROI approaches and provides an overview of state of the art ROI coding techniques for medical images along with corresponding results.
Introducing Basic Concepts Of Roi Coding
The functionality of Region of Interest (ROI) is important in medical applications where certain parts of the image are of higher diagnostic importance than others. In such a case, these regions need to be encoded at higher quality than the background. During image transmission for telemedicine purposes, these regions are required to be transmitted first or at a higher priority. In transformation-based ROI coding methods, the coefficients associated with the ROI are transferred ahead of those associated with the background. Therefore, when an image is coded with an emphasis of ROI, it is necessary to identify the coefficients required for the reconstruction of the ROI. Thus, a ROI mask is introduced to indicate which coefficients have to be transmitted exactly in order for the receiver to reconstruct the ROI. Usually, the wavelet transform (Burrus et. al., 1998; I. Daubechies, 1998) is applied to the image at the encoder side and the resulting coefficients not associated with the ROI are scaled down (shifted down) so that the ROI associated bits are placed in higher bit planes The mask in wavelet domain is a map pointing out all the related coefficients for the reconstruction of the ROI. The corresponding locations of the coefficients in next scale are calculated from the current scale. An example calculation of the ROI mask is as follows (Liu et. al. 2004):
Let Rn the wavelet domain of an image and ΩRn the Region of Interest. The characteristic function is defined as: (1)
Then the ROI mask will be generated according to: (2)
Where stands for the wavelet operator for the ith subband, Λ is the index set of all subbands and is identity operator equipped with down-sampling operation respectively.
Key Terms in this Chapter
Medical Image Coding: Refers to image compression as the application of data compression on digital images. In effect, the objective is to reduce redundancy of the image data in order to be able to store or transmit data in an efficient form. Image compression can be lossy or lossless. Lossy compression methods, especially when used at low bit rates, introduce compression artifacts. Lossless compression methods may also be preferred for high value content, such as medical imagery or image scans made for archival purposes.
Wavelet Coding: Wavelet coding or compression is a form of data compression well suited for image compression (sometimes also video compression and audio compression). Wavelet compression can be either perfect (lossless) or lossy, where a certain loss of quality is accepted. Using a wavelet transform, the wavelet compression methods are adequate for representing transients, such as percussion sounds in audio, or high-frequency components in two-dimensional images, for example an image of stars on a night sky. This means that the transient elements of a data signal can be represented by a smaller amount of information than would be the case if some other transform, such as the more widespread discrete cosine transform, had been used.
Volumetric Medical Image: A typical 3D data set is a group of 2D slice images acquired by a CT or MRI scanner. Usually these are acquired in a regular pattern (e.g., one slice every millimeter) and usually have a regular number of image pixels in a regular pattern. This is an example of a regular volumetric grid, with each volume element, or voxel represented by a single value that is obtained by sampling the immediate area surrounding the voxel.
ROI Coding: The Region Of Interest (ROI) coding it is a function that enables a non-uniform distribution of the image quality between a selected region (the ROI) and the rest of the image (background). ROI coding of medical images allows the compression of diagnostically important regions at better quality without affecting the visual assessment procedure, whereas areas like the background can be coded at lower quality in order to decrease image size and improve storage and/or transmission procedures.
Wavelet: A one-dimensional pulse, usually the basic response from a single reflector. Its key attributes are its amplitude, frequency and phase. The wavelet originates as a packet of energy from the source point, having a specific origin in time, and is returned to the receivers as a series of events distributed in time and energy. The distribution is a function of velocity and density changes in the subsurface and the relative position of the source and receiver.
Distributed Telemedicine: Store-and-forward telemedicine involves acquiring medical data (like medical images, biosignals etc) and then transmitting this data to a doctor or medical specialist at a convenient time for assessment offline. Dermatology, radiology, and pathology are common specialties that are conducive to asynchronous telemedicine.
Complete Chapter List
Themis P. Exarchos, Athanasios Papadopoulos, Dimitrios I. Fotiadis
Ioannis Dimou, Michalis Zervakis, David Lowe, Manolis Tsiknakis
William Hsu, Alex A.T. Bui, Ricky K. Taira, Hooshang Kangarloo
Spyretta Golemati, John Stoitsis, Konstantina S. Nikita
Christos V. Bourantas, Katerina Naka, Dimitrios Fotiadis, Lampros Michalis
Stavroula Mougiakakou, Ioannis Valavanis, Alexandra Nikita, Konstantina S. Nikita
Marotesa Voultsidou, J. Michael Herrmann
Dimitrios C. Karampinos, Robert Dawe, Konstantinos Arfanakis, John G. Georgiadis
Dimitrios G. Tsalikakis, Petros S. Karvelis, Dimitrios I. Fotiadis
Katia Marina Passera, Luca Tommaso Mainardi
Lena Costaridou, Spyros Skiadopoulos, Anna Karahaliou, Nikolaos Arikidis, George Panayiotakis
E. Kyriacou, C.I. Christodoulou, C. Loizou, M.S. Pattichis, C.S. Pattichis, S. Kakkos
Marios Neofytou, Constantinos Pattichis, Vasilios Tanos, Marios Pattichis, Eftyvoulos Kyriacou
Thomas V. Kilindris, Kiki Theodorou
Ioannis Tsougos, George Loudos, Panagiotis Georgoulias, Konstantina S. Nikita, Kiki Theodorou
Evanthia E. Tripoliti, Dimitrios I. Fotiadis, Konstantia Veliou
Anastasios Koutlas, Dimitrios I. Fotiadis
Aristotelis Chatziioannou, Panagiotis Moulos
Nikolaos Giannakeas, Dimitrios I. Fotiadis
Petros S. Karvelis, Dimitrios I. Fotiadis
O. Lezoray, G. Lebrun, C. Meurie, C. Charrier, A. Elmotataz, M. Lecluse
Michael Haefner, Alfred Gangl, Michael Liedlgruber, A. Uhl, Andreas Vecsei, Friedrich Wrba
C. Papaodysseus, P. Rousopoulos, D. Arabadjis, M. Panagopoulos, P. Loumou
Alexia Giannoula, Richard S.C. Cobbold
Valentina Russo, Roberto Setola
George K. Matsopoulos
Alberto Taboada-Crispi, Hichem Sahli, Denis Hernandez-Pacheco, Alexander Falcon-Ruiz
C. Delgorge-Rosenberger, C. Rosenberger
Charalampos Doukas, Ilias Maglogiannis