20180116TconAgendaMinutes
- Version 6
- by (unknown)
- Version 7
- by (unknown)
Deletions or items before changed
Additions or items after changed
| 1 | Brandon announced that he will be transitioning from email as the means of communications to announcements made through our eeDAP studies group on the NCIPhub. https://nciphub.org/groups/eedapstudies | |||
|---|---|---|---|---|
| 2 | ||||
| 3 | To stay involved and receive the communications, please join this group. | |||
| 4 | ||||
| 5 | Please also join the WSI working group https://nciphub.org/groups/wsi_working_group | |||
| 6 | ||||
| 7 | == Recent studies at MSKCC == | |||
| 8 | ||||
| 9 | * Have an eeDAP system on loan from FDA ([https://nciphub.org/groups/eedapstudies/wiki/20171121update/File:MSK6501_FE.pdf Link to MTA], [https://www.ott.nih.gov/sites/default/files/documents/policy/MTA.pdf Link to Template]). | |||
| 10 | ||||
| 11 | * Have conducted eeDAP studies, including a study on a 14-head microscope ([StudyFDAMSKCC14headMicroscope Link]). | |||
| 12 | - | + | * One question asked what was the purpose of the study. This study was a proof of concept study and an opportunity to see the pros and cons of conducting a study on a multi-head microscope. We also want to use this data to study different endpoints for reader agreement and algorithm validation. This dataset could be a resource for algorithm developers to train or test their algorithms. The Tupac16 challenge included the mitosis detection task, so there is an opportunity to recruit algorithms for mitotic detection and counting.
|
|
| 13 | - | + | ||
| 14 | * One question was raised to clarify that Brandon and Qi used eeDAP (with !ImageScope integration) to transcribe the data collected on paper. This included transcribing the circles drawn on paper to marks on !ImageScope. | |||
| 15 | ||||
| 16 | * Update from Yukako. [[File(FDAyagi.pdf)]] | |||
| 17 | ||||
| 18 | == eeDAP installation at CSHL == | |||
| 19 | ||||
| 20 | * Update from Partha. Installation successful. CSHL is planning psychophysics studies to study pathologist behavior. Partha introduced colleagues. | |||
| 21 | - | |||
| 22 | * Adam Kepecs: Scientist at CSHL. Phychophysics background. | |||
| 23 | * Jim Crawford: Pathologist at Northwell Health | |||
| 24 | * Craig Abbey: Scientist at UC Santa Barbara. Psychophysics background. | |||
| 25 | ||||
| 26 | * Discussions to begin on performance endpoints. [PerformanceEndpointsDiscussion Link to discussion.] | |||
| 27 | ||||
| 28 | == Update on registration accuracy studies and MDDT == | |||
| 29 | ||||
| 30 | * Presenting registration accuracy data at SPIE Medical Imaging Conference, final paper will be uploaded to [RegistrationAccuracy the current wiki page on registration accuracy.] | |||
| 31 | ||||
| 32 | * Plan to revise MDDT submission: eliminate discussion of reader studies, make context of use about registration of WSI and microscope | |||
| 33 | ||||
| 34 | == Project proposal bringing together eeDAP reader studies and AI grand challenges == | |||
| 35 | ||||
| 36 | * Welcome/introduce grand challenge organizers Jeroen van der Laak and Mitko Veta | |||
| 37 | * Hiring to support project [JobsDigitalPathologyImageandAlgorithmEvaluation Job Announcement] | |||
| 38 | * Basic info at wiki [HighThroughputTruthing Link] | |||
| 39 | * Quick review of proposal | |||
| 40 | * First steps: digest materials, understand performance metrics, discuss needs | |||