Wetenschappelijke stage GNK/BMW: Improving treatment decisions
Wetenschappelijke stage GNK/BMW: Improving treatment decisions
-
| laatst bewerkt:Tijn Gerrits · 14 · Wetenschappelijke stage GNK/BMW: Improving treatment decisions
Research internship bachelor / master Biomedical Sciences or Medicine
Computational pathology - Improving treatment decisions in colorectal cancer
Fields of interest
- Artificial intelligence
- Computational Pathology (gastrointestinal)
- Health technology assessment
- Implementation
Background
Over the past decade, the application of digital pathology has increased tremendously. Pathology laboratories are exchanging microscopes for whole slide image scanners, enabling digital image analysis. Moreover, a new generation of extremely powerful pattern recognition algorithms, e.g. deep learning techniques, have enabled development of computer aided diagnosis systems that approach the accuracy of pathologists for certain specific applications (computational pathology).
Computational pathology algorithms can potentially increase pathologists’ efficiency by automating repetitive task of low complexity, and may be helpful in aiding more complex diagnostics, e.g. grading in oncology, which are often accompanied by significant inter-observer variability. The techniques promise to reduce workload for the pathologist, increase objectivity of diagnoses and therefore improve patient outcomes.
Problem statement
Despite the promising results of computational pathology algorithms, a large scale implementation strategy within routine oncological diagnostics remains to be defined. In a previous qualitative interview study (Swillens et al. submitted) important factors influencing the use of computational pathology in clinical practice were explored. The next step is to determine which of these factors are considered as essential in the path to the clinic among a broad panel of pathologists. Therefore in this study, we will conduct a Delphi study consisting of 3 consecutive rounds, resulting in a set of preconditions.
Objective
Delphi study consisting of three rounds to determine essential preconditions of computational pathology algorithms in colorectal cancers among a broad panel of pathologists in Europe.
Tasks
Performing literature research, define and contact the panel of experts, question development, set-up the diverse rounds for the Delphi procedure to identify the preconditions and controversies that can be explored in next rounds, produce a summary statement and sent back to the panel statistical analysis, and writing draft manuscript, with the option to co-author the resulting publication.
Time period
20 -24 weeks, full time. Starting date: spring 2023 (March/April)
Supervision
Dr. Marcia Tummers (Health Evidence)
Dr. Julie Swillens (IQ healthcare)
In close collaboration with the department of Pathology
Contact: Marcia.Tummers@radboudumc.nl
Reacties toevoegen is niet mogelijk