Authors:
Venkata Praveen Karri
1
;
JungHwan Oh
1
;
Wallapak Tavanapong
2
;
Johnny Wong
2
and
Piet C. de Groen
3
Affiliations:
1
University of North Texas, United States
;
2
Iowa State University, United States
;
3
College of Medicine and Mayo Clinic, United States
Keyword(s):
Colonoscopy, CUDA, GPU, CPU multi-threading, Informative frame filtering.
Related
Ontology
Subjects/Areas/Topics:
Applications and Services
;
Biomedical Engineering
;
Biomedical Signal Processing
;
Computer Vision, Visualization and Computer Graphics
;
Medical Image Detection, Acquisition, Analysis and Processing
Abstract:
Colonoscopy is an endoscopic technique that allows a physician to inspect the mucosa of the human colon. It has contributed to a marked decline in the number of colorectal cancer related deaths. However, recent data suggest that there is a significant (4-12%) miss-rate for the detection of even large polyps and cancers. To address this, we have investigated automated post-procedure and real-time quality measurements by analyzing colonoscopy videos. One of the fundamental steps is separating informative frames from non-informative frames, a process called Informative Frame Filtering (IFF). Non-informative frames comprise out-of-focus frames and frames lacking typical features of the colon. We introduce a new IFF algorithm in this paper, which is much more accurate than our previous one. Also, we exploit the many-core GPU (Graphics Processing Unit) to create an IFF software module for High Performance Computing (HPC). Code optimizations embedded in the many-core GPU resulted in a 40-fo
ld acceleration compared to CPU-only implementation for our IFF software module.
(More)