Enhanced vascular and osseous information fusion: disagreement of quantitative and qualitative analysis

  • PDF / 1,173,496 Bytes
  • 11 Pages / 595.276 x 790.866 pts Page_size
  • 40 Downloads / 261 Views

DOWNLOAD

REPORT


(0123456789().,-volV)(0123456789().,-volV)

RECENT ADVANCES IN DEEP LEARNING FOR MEDICAL IMAGE PROCESSING

Enhanced vascular and osseous information fusion: disagreement of quantitative and qualitative analysis Ayush Dogra1 • Bhawna Goyal1 • Sunil Agrawal1 • Urcun John Tanik2 • Sanjeev Kumar3 • Ramesh Sunder Nayak4 Received: 25 October 2018 / Accepted: 14 May 2019  Springer-Verlag London Ltd., part of Springer Nature 2019

Abstract The computer-aided algorithms are becoming indispensable in the field of digital subtraction angiography for better subsequent clinician diagnosis and feature extraction. These tools find widespread application for localization and highlighting potential malfunctioned sites and abnormal vascular pathology. For this purpose, an enhancement filter function based on Hessian matrix is often employed which produces a response close to ideal enhancement function in the literature. However, while predicting the abnormal pathology, both vessel and osseous details are elementary to patient diagnosis and treatment. In this manuscript, with an aim to parallelize the visualization of enhanced vessel details and osseous information for quick diagnosis, an image fusion algorithm is proposed based on the pre-enhancement of source images which generates highly pleasant visual results free from noise and ghostly artefacts outperforming the ten other image fusion techniques. Besides, this study presents a potential research challenge of disagreement between objective and subjective evaluation of fused image. Keywords Osseous  Vascular  Aneurysms  Fusion  DSA  Hessian matrix

1 Introduction The segmentation and visualization of vascular pathologies are essential for effective diagnosis of vascular diseases such as embolisms, stenoses and other malformations affecting vital anatomical regions primarily cardiac, pulmonary, retinal and cerebral. Vascular diseases have emerged as one of the single most responsible cause of death and strokes in the urban areas around the globe. Therefore, the imaging of vessel structures is a noninvasive clinical tool for the detailed investigation of the cardiac and improper blood circulation conditions [1–3]. The primary imaging technology which finds key application in this & Ayush Dogra [email protected]; [email protected] 1

UIET, Panjab University, Chandigarh, India

2

Computer Science and Information Systems, Texas A&M University-Commerce, Commerce, TX, USA

3

CSIR-Central Scientific Instruments Organization (CSIRCSIO), Chandigarh, India

4

Information Science and Engineering, Canara Engineering College, Mangaluru, India

field is digital subtraction angiography. From the beginning of the twentieth century, the DSA imaging technology has emerged as the discrete referral standard for visualization of malfunction and blocking of vascular pathologies and helps in the follow-up for documentation of the progress of the subjugated treatment as these images possess high diagnostic quality and high resolution. The computer-aided analysis and machine vision enable