VARACADEMY
Reading mammograms in the age of AI
Varacademy is Vara's training programme for mammography readers: real, anonymised cases to read, to understand how to best work with AI
Train your eye across the full screening mix
CASE PRACTICE
Train on anonymised cases from Vara's case collection, which includes data from the PRAIM study: from clearly visible findings to the ones where even the most experienced readers struggle.
CASES COVER
Different lesion types: masses, microcalcifications, architectural distortions
All breast densities, including dense breasts where findings hide
Normal cases, because recognising a normal is part of the skill
See your time and result on every case
RESULTS
After each case compare your assessment to other human readers, to Vara AI, and finally to the confirmed outcome. A session closes with your decisions and reading times in one summary.
Session ended ·
The Reading Room
Example session
18 / 24
correct decisions
across 24 cases
Correct decisions
75%
7 of 9 cancers localized
Total reading time
35:48
Avg per case
1:30
Fastest
Case 17 · 41s
Slowest
Case 6 · 3:12
Case
Time
Result
Case 1
1:12
Correct
Case 2
2:05
Correct
Case 3
48s
Incorrect
Case 4
1:31
Skipped
Case 5
1:54
Correct
Sample figures, for illustration. Session results are for personal review only, and are not a clinical assessment or certification.
AI LITERACY
Get familiar with AI scores in a training setting using real cases
Varacademy shows you how the AI behaves on real cases: what its scores mean, where it is most useful, and how to read alongside it.
YOU'LL LEARN TO
Interpret AI scores and classifications, and what they do and don't mean
Understand the behaviour of AI on various types of findings
Compare AI scores and classifications with that of other readers and confirmed histology (where available)
See how your day-to-day reading changes when you read with AI.
Train your eye across the full screening mix
CASE PRACTICE
Train on anonymised cases from Vara's case collection, which includes data from the PRAIM study: from clearly visible findings to the ones where even the most experienced readers struggle.
CASES COVER
Different lesion types: masses, microcalcifications, architectural distortions
All breast densities, including dense breasts where findings hide
Normal cases, because recognising a normal is part of the skill
See your time and result on every case
RESULTS
After each case compare your assessment to other human readers, to Vara AI, and finally to the confirmed outcome. A session closes with your decisions and reading times in one summary.
Session ended ·
The Reading Room
Example session
18 / 24
correct decisions
across 24 cases
Correct decisions
75%
7 of 9 cancers localized
Total reading time
35:48
Avg per case
1:30
Fastest
Case 17 · 41s
Slowest
Case 6 · 3:12
Case
Time
Result
Case 1
1:12
Correct
Case 2
2:05
Correct
Case 3
48s
Incorrect
Case 4
1:31
Skipped
Case 5
1:54
Correct
Sample figures, for illustration. Session results are for personal review only, and are not a clinical assessment or certification.
AI LITERACY
Get familiar with AI scores in a training setting using real cases
Varacademy shows you how the AI behaves on real cases: what its scores mean, where it is most useful, and how to read alongside it.
YOU'LL LEARN TO
Interpret AI scores and classifications, and what they do and don't mean
Understand the behaviour of AI on various types of findings
Compare AI scores and classifications with that of other readers and confirmed histology (where available)
See how your day-to-day reading changes when you read with AI.
WHAT THE CASES INCLUDE
Different lesion types: masses, microcalcifications, architectural distortions
All breast densities, including dense breasts where findings hide
A mix of normal, benign and malignant cases from real screening rounds
CASE PRACTICE
Train your eye across the full screening mix
Train on anonymised cases from Vara's case collection, which includes data from the PRAIM study: from clearly visible findings to the ones where even the most experienced readers struggle.
RESULTS
See your time and result on every case
After each case compare your assessment to other human readers, to Vara AI, and finally to the confirmed outcome. A session closes with your decisions and reading times in one summary.
Session ended · The Reading Room
Example session
18 / 24
correct decisions across 24 cases
Correct decisions
75%
7 of 9 cancers localized
Total reading time
35:48
Avg per case
1:30
Fastest
Case 17 · 41s
Slowest
Case 6 · 3:12
Case
Time
Result
Case 1
1:12
Correct
Case 2
2:05
Correct
Case 3
48s
Incorrect
Case 4
1:31
Skipped
Case 5
1:54
Correct
Sample figures, for illustration. Session results are for personal review only, and are not a clinical assessment or certification.
Session ended · The Reading Room
Example session
18 / 24
correct decisions across 24 cases
Correct decisions
75%
7 of 9 cancers localized
Total reading time
35:48
Avg per case
1:30
Fastest
Case 17 · 41s
Slowest
Case 6 · 3:12
Case
Time
Result
Case 1
1:12
Correct
Case 2
2:05
Correct
Case 3
48s
Incorrect
Case 4
1:31
Skipped
Case 5
1:54
Correct
Sample figures, for illustration. Session results are for personal review only, and are not a clinical assessment or certification.
THE TRAINING PLATFORM
Cases grouped by what they teach
Each room has its own focus, from a 100-case mix to a single finding type.
Track 01 · Mixed cases
The Reading Room
100 cases · about 2.5 hours
38 cancers, 22 benign findings and 40 normal exams, with no AI shown. Every finding is histologically confirmed.
Track 02 · AI literacy
AI Blind Spots
14 cases · about 25 minutes
Read cases Vara AI scored below the recall threshold, then see where its attention fell and what histology showed. These are deliberately selected examples and do not reflect typical screening prevalence or overall AI performance.
Track 03 · Skill training
Microcalcifications
3 cases
Focused practice on microcalcifications, reading their morphology and distribution.
More rooms are in development: Architectural Distortions, Asymmetries, The Density Challenge and Prior Comparison Skills.
All cases are anonymised before use.
TRY IT
Make the call on one case
Mark anything you would recall, then compare your decision with the AI score and the biopsy result.
Example case, anonymised. No patient-identifying data is shown.
No decision recorded yet.
Staying Sharp Challenge
POWERED BY VARACADEMY
The Staying Sharp Challenge is a hands-on reading session at radiology conferences and workshops. Read real cases, see what the AI sees, and find out where you and the AI agree and disagree.
Staying Sharp Challenge
POWERED BY VARACADEMY
The Staying Sharp Challenge is a hands-on reading session at radiology conferences and workshops. Read real cases, see what the AI sees, and find out where you and the AI agree and disagree.
Staying Sharp Challenge
POWERED BY VARACADEMY
The Staying Sharp Challenge is a hands-on reading session at radiology conferences and workshops. Read real cases, see what the AI sees, and find out where you and the AI agree and disagree.
Upcoming
Event
Location
Date
GISMa workshop
Florence, Italy
17 October 2026
Starting out in breast imaging?
FOR RESIDENTS
Reading experience takes years and thousands of cases. Varacademy gives you a head start: real screening cases, organised by what you need to practise, with ground truth and histology after every case.
FOR RESIDENTS
Starting out in breast imaging?
Reading experience takes years and thousands of cases. Varacademy gives you a head start: real screening cases, organised by what you need to practise, with ground truth and histology after every case.
Get familiar with AI scores in a training setting using real cases
AI LITERACY
Varacademy shows you how the AI behaves on real cases: what its scores mean, where it is most useful, and how to read alongside it.
YOU'LL LEARN TO
Interpret AI scores and classifications, and what they do and don't mean
Understand the behaviour of AI on various types of findings
Compare AI scores and classifications with that of other readers and confirmed histology (where available)
See how your day-to-day reading changes when you read with AI.
FOR RESIDENTS
Starting out in breast imaging?
Reading experience takes years and thousands of cases. Varacademy gives you a head start: real screening cases, organised by what you need to practise, with ground truth and histology after every case.
Learn with Varacademy
Learn with Varacademy
Varacademy is growing. Leave your details and we'll let you know about new case collections, webinars on working with AI, and upcoming Sharp Eyes Challenges.