A new method of analysing the intracerebral haemorrhage signal intensity on brain MRI images using frequency domain techniques

Abstract

Diffusion Weighted Magnetic Resonance Imaging (DW-MRI) has become an established part of neuroimaging and is used to diagnose and characterize several neurologic disorders. Intracerebral Haemorrhage (ICH) is a severe medical condition, which may develop quickly into a life-threatening situation, and thereby requires prompt medical attention. Early and reliable identification of the age of haemorrhage is essential when choosing the correct treatment, and estimating patient's diagnosis and outcome. Diffusion Weighted (DW) images presents a variation in the image signal intensity characteristics relative to the different stages of ICH. In the present paper, an effort is made to document the variation in the image signal intensity characteristics of ICH at evolving stages, for 30 subjects, using High Frequency Power (HFP) parameter. Results showed that the difference in the HFP values on DW images for the subjects with ICH compared to their contralateral normal hemisphere, were highly significant (p < 0.

Computer Based Method to Automatically Identify the Regions of Intracerebral Haemorrhage on Diffusion Weighted Magnetic Resonance Images

“…However, in recent times numerous methods for automated diagnostic systems have been proposed to overcome the problems faced due to large number of patients, and the need for constant observation of certain medical conditions. Studies have been carried out in the past that investigate into the signal intensity characteristics of ICH on DW brain images, and further quantify these signal intensity variations to derive useful information about the state of brain tissue in the ICH subject [17], [18]. However, identification of the regions of signal intensity variations that characterize the occurrence of ICH on DW brain images has been done manually by visual inspection.…”

“…Towards this end the present work makes an effort to develop a computer based method to automatically identify the regions of ICH (regions of signal intensity variations) on DW brain images of the ICH subjects, without any human intervention. The proposed method along with the work carried out earlier [17], [18] would provide a complete automated system that would identify the ICH regions on DW brain images and also classify the ICH subjects into different stages of haemorrhage. Thereby the adoption of the present work in the medical diagnosis of ICH subjects would ease the estimation process and help in providing important details for the further management and medicinal decisions, even in the absence of a medical expert.…”