Recognition of colonic dysplasia at colonoscopy using advanced technology
Author(s)
Ignjatovic, Ana
Type
Thesis
Abstract
Colonoscopy remains the gold standard for detection of dysplasia in the colon.
Resection of dysplasia at colonoscopy has been shown to reduce the subsequent
risk of colorectal cancer. With the recent advent of National Bowel Cancer Screening
Programme the number of colonoscopies performed in the United Kingdom has
increased significantly. In addition, the recent drive for quality improvement has
resulted in increased number of dysplastic lesions detected, resected and sent for
histopathology. This has led to substantial increases in the cost associated with the
procedure. This thesis investigates how advanced technology can improve detection
and characterisation of dysplasia at colonoscopy, potentially contributing to time and
cost savings.
The introduction reviews the literature on dysplasia detection at colonoscopy, both in
terms of basic withdrawal technique and how advanced technology can help reduce
miss rates. The significance of small colorectal polyps is reviewed and the accuracy
of recent colonoscopic technologies in polyp characterization considered.
Three studies examined detection of dysplasia: a small randomized control trial
showed that magnetic endoscope imager did not improve caecal intubation or
dysplasia detection rates in selected patients when used by experienced
colonoscopists. Two studies, one of which was in patients with long standing colitis,
did not demonstrate the benefit of narrow band imaging for detection of dysplasia.
A cohort study revealed that in vivo characterization of dysplasia can be done
accurately and is feasible in routine clinical practice although the inter-observer
agreement is moderate at best as assessed by a retrospective study. A computer
algorithm was devised to attempt to reduce the inter-observer variability and has
shown promising results in a pilot study. A training module on the use of narrow
band imaging in characterization was developed and validated leading to improved
accuracy and reduced variability amongst the participants. Finally, a quality
assurance study confirmed that the still photographs of polyps are adequate record
of optical diagnosis.
Resection of dysplasia at colonoscopy has been shown to reduce the subsequent
risk of colorectal cancer. With the recent advent of National Bowel Cancer Screening
Programme the number of colonoscopies performed in the United Kingdom has
increased significantly. In addition, the recent drive for quality improvement has
resulted in increased number of dysplastic lesions detected, resected and sent for
histopathology. This has led to substantial increases in the cost associated with the
procedure. This thesis investigates how advanced technology can improve detection
and characterisation of dysplasia at colonoscopy, potentially contributing to time and
cost savings.
The introduction reviews the literature on dysplasia detection at colonoscopy, both in
terms of basic withdrawal technique and how advanced technology can help reduce
miss rates. The significance of small colorectal polyps is reviewed and the accuracy
of recent colonoscopic technologies in polyp characterization considered.
Three studies examined detection of dysplasia: a small randomized control trial
showed that magnetic endoscope imager did not improve caecal intubation or
dysplasia detection rates in selected patients when used by experienced
colonoscopists. Two studies, one of which was in patients with long standing colitis,
did not demonstrate the benefit of narrow band imaging for detection of dysplasia.
A cohort study revealed that in vivo characterization of dysplasia can be done
accurately and is feasible in routine clinical practice although the inter-observer
agreement is moderate at best as assessed by a retrospective study. A computer
algorithm was devised to attempt to reduce the inter-observer variability and has
shown promising results in a pilot study. A training module on the use of narrow
band imaging in characterization was developed and validated leading to improved
accuracy and reduced variability amongst the participants. Finally, a quality
assurance study confirmed that the still photographs of polyps are adequate record
of optical diagnosis.
Date Issued
2011-08
Date Awarded
2013-04
Copyright Statement
Attribution NoDerivatives 4.0 International Licence (CC BY-ND)
Advisor
Saunders, Brian
Hart, Alisa
Publisher Department
Surgery and Cancer
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Medicine (Research) MD (Res)