A video processing pipeline for intraoperative analysis of cerebral blood flow

ABSTRACT

In this paper, we propose a pipeline for non-invasive, high-resolution analysis of cerebral blood flow. We use a 3-CMOS HD-Videocamera to record video samples that are processed by a stabilization and colour amplification pipeline to visualize changes caused by the blood pressure cycle. To our knowledge, this is the first description of a non-invasive, only software-based algorithm which comprises information about brain perfusion.

KEYWORDS:

• Image processing and analysis
• medical imaging and visualization
• applications of imaging and visualization

Acknowledgement

This work was partially supported by Federal Ministry of Education and Research [grant number 16SV7236].

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. We use the term detector to refer to a pair of detection and description methods.

2. If filter kernels are actually convolved or correlated in convolutional neural networks is an implementation detail differing across libraries.

3. Disturbance by noise in the key frame is neglected for ease of reading.

4. If this condition would not hold, amplification would still be possible if $E(w)$ is equal for all frames, because of amplifying the differences.

5. www.blender.org.

6. In reality, this might never be achieved, for example, due to noise caused by interpolation.

Additional information

Notes on contributors

Sören Klemm

Sören Klemm received a Diploma in Electrical Engineering from Federal Armed Forces University Munich, Germany and an MSc. in Digital Signal and Image Processing from University of Central Lancashire, United Kingdom. Since 2015 he is a PhD candidate at University of Münster. His research focus lies in the field of Computer Vision and applications of Machine Learning.

Robin Rexeisen

Robin Rexeisen studied Computer Science at University of Münster, he received a BSc. in 2015 and MSc. in 2018. Currently he is pursuing a Bachelor degree in Mathematics in Münster.

Walter Stummer

Walter Stummer was trained as neurosurgeon at the University of Munich, Germany. He became certified in the year 2000. In 2003 he was appointed as Vice-Chairman of Neurosurgery at the University of Düsseldorf. 2009 he was appointed to the position of Chairman and Director of the Department of Neurosurgery at the University of Münster. Professor Stummer has served as President of the German Society of Neurosurgery from 2016 to 2018. He specializes in brain tumor surgery and has pioneered fluorescence-guided surgery. For this work he has received numerous awards.

Xiaoyi Jiang

Xiaoyi Jiang studied Computer Science at Peking University and received his Ph.D. and Venia Docendi (Habilitation) degree in Computer Science from University of Bern, Switzerland. He was an Associate Professor at Technical University of Berlin, Germany. Since 2002 he is a Full Professor at University of Münster, Germany. Currently, he is Editor-in-Chief of International Journal of Pattern Recognition and Artificial Intelligence. In addition, he also serves on the Advisory Board and Editorial Board of several journals, including IEEE Transactions on Medical Imaging, International Journal of Neural Systems, and Pattern Recognition. He is a Senior Member of IEEE and Fellow of IAPR.

Markus Holling

Markus Holling studied Medicine at the Universities of Marburg and Muenster, Germany and is working as an Associate Professor for Neurosurgery at the Department of Neurosurgery, University Hospital Muenster, Germany with a clinical and research focus on vascular Neurosurgery. In 2011 he received the Wilhelm-Tönnis-Grant from the German Society for Neurosurgery (DGNC). He completed research fellowships at the Departments of Neurosurgery of Professor Hernesniemi (Helsinki, Finland) and Professor Charbel (Chicago, USA). For his student teaching at the Faculty of Medicine, University of Muenster he received the “teacher of the year” – award twice (2012 & 2017).