Johnston, Dan (2005) The VPS ReplaySuite: development and evaluation of a novel, Internet based telepathology tool. PhD thesis, Dublin City University.
Abstract
The ReplaySuite is a web-based telepathology tool that replicates the doubleheaded microscope environment online, enabling a reviewing pathologist to ‘replay’ an archived virtual slide examination. Examination-tracking data obtained by the Virtual Pathology Slide (VPS) virtual slide viewer is exploited, allowing a remote pathologist to review an examination conducted at a different time and location. This removes temporal and spatial issues associated with double-headed microscopy.
In order to conduct a preliminary evaluation of the technology, 9 pathologists used the ReplaySuite to review examination replays and diagnostic data from archived examinations of 10 needlecore breast biopsies. Diagnostically difficult cases were most frequently evaluated, either via diagnostic concordance graphs or examination replays, and all 3 participants who replayed more than 10 examinations stated the ReplaySuite to be of some or great benefit in pathology training and quality assurance. Of those who replayed an examination by another pathologist, 83% (5/6) agreed that replays provided an insight into the examining pathologists diagnosis, and 33% (2/6) reconsidered their own diagnosis for at least one case. Of those who reconsidered their original diagnosis, all reclassified either concordant with group consensus or original glass slide diagnosis.
This study demonstrated that the ReplaySuite was of potential benefit in pathology education, however the technology required evaluation in a setting that would facilitate its impact on diagnostic performance. Accordingly, a redeveloped VPS and ReplaySuite were incorporated into the EQUALIS External Quality Assurance (EQA) study in chronic hepatitis staging and grading.
During the study, 9 Swedish pathology departments examined and scored digital representations of liver needlecore biopsies during two sessions, with 10 cases per session and two digital slides per case. Between scoring sessions, participants were provided with access to two supplementary electronic resources: the ReplaySuite, and a library of pre-selected reference images. Comparison of concordance with gold standard (KVAST group) scoring before and after electronic resource use facilitated the elucidation of impact on diagnostic performance.
Between scoring sessions, participant concordance with KVAST staging increased by 18% (49%-67%), while concordance with KVAST grading increased by 20% (34%-54%). Mean staging un-weighted kappa improved from 0.347 to 0.554 (+0.207), or from ‘fair’ to ‘moderate’ exact agreement with KVAST staging. Linear weighted staging kappa improved from 0.603 to 0.688 (+0.085), indicating close agreement in both sessions. Mean grading unweighted kappa increased from 0.132 to 0.412 (+0.280), or from a ‘poor’ to ‘moderate’ level o f exact agreement with KVAST, while linear weighted kappa improved from 0.328 to 0.624 (+0.295), or from ‘fair’ to ‘good’ level of approximate agreement with KVAST.
Subsequent to the EQA scheme, an expert liver pathologist used the ReplaySuite to evaluate study examinations, assessing examination technique and identifying sources of error. Examinations scoring concordant with KVAST were observed to exhibit acceptable examination technique more frequently than discordantly scoring examinations. When grading, 28% (46% - 18%) more concordant than discordant examinations were considered to have viewed sufficient tissue, and at the appropriate magnification. A similar disparity of 24% (59% - 35%) was observed in staging, suggesting that examination technique was important both when determining the degree of necroinflammation within a biopsy, and when ascertaining the extent of fibrosis.
In assessing sources of error, the expert pathologist identified a potential source in 50% of grading examinations, with misinterpretation of observed pathology cited in 19%, and missed pathology (oversight) cited in 31% of grading examinations. Of the 41% of staging examinations in which a source was identified, misinterpretation of observed pathology was cited in 20% of examinations, and missed pathology (oversight) in 21% of examinations.
This study demonstrated that the use of supplementary electronic resources could result in improvements in diagnostic performance. It also illustrated the significant ‘add on’ value that could be provided by the ReplaySuite in EQA, by providing means to assess not only diagnostic concordance, but also diagnostic technique and identify sources of error.
In order to assess Irish trainee pathologist’s perceptions of computer-assisted learning (CAL), a number of commercial systems were utilised to incorporate digital slides into a postgraduate seminar series, and provide subsequent access to seminar digital slides, diagnoses and expert annotations online. All surveyed trainees considered the use of digital slides and expert annotations of benefit in pathology training, and considered the potential implementation of expert examination replays, online self-assessment and the capability to search online for material by organ, diagnosis or pathological feature of benefit.
The work described herein illustrates that both expert and trainee pathologists alike consider the use of supplementary electronic resources of benefit in pathology education, and demonstrates that their use can improve diagnostic performance. The ability to evaluate participation in EQA studies via the ReplaySuite provides significant additional value to education schemes, providing a depth of assessment not possible with conventional microscopy.
Metadata
Item Type: | Thesis (PhD) |
---|---|
Date of Award: | 2005 |
Refereed: | No |
Supervisor(s): | O'Shea, Donal |
Uncontrolled Keywords: | Pathology; ReplaySuite; Pathology Education |
Subjects: | Biological Sciences > Biotechnology Humanities > Biological Sciences > Biotechnology |
DCU Faculties and Centres: | DCU Faculties and Schools > Faculty of Science and Health > School of Biotechnology |
Use License: | This item is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 3.0 License. View License |
ID Code: | 17936 |
Deposited On: | 24 Apr 2013 12:59 by Celine Campbell . Last Modified 24 Apr 2013 12:59 |
Documents
Full text available as:
Preview |
PDF
- Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
10MB |
Downloads
Downloads
Downloads per month over past year
Archive Staff Only: edit this record