Case study
Accurate Cell Labelling, by Medical Students
University of Zambia
School of Medicine
Microscopy
instance segmentation
On time
every batch delivered
The Challenge
SuperAnnotate's model produced first-pass cell segmentations on microscopy images, but with errors and missed cells. Fixing them takes real knowledge of cell structure and cell types.
What We Did
We partnered with the University of Zambia's School of Medicine to build a team of qualified doctors, recent graduates and final-year medical students, interviewed every candidate, and trained the cohort on the project, with SuperAnnotate training them on its tool. The team corrected the model's annotations and added the cells it had missed.
The Result
Every batch delivered on time by a team of qualified doctors and medical graduates, with hundreds of cells accurately segmented to train SuperAnnotate's medical AI.