LO
CapabilityCognitive automation2018-12-01
With algorithm assistance, pathologists found lymph-node micrometastases more often (91% against 83%) in about half the time per image, a reader study found
Pathologistoccupation page →
Event date / reported
2018-12-01
Evidence stage
CapabilityA demo, benchmark or paper shows the task can be done. Updates what the technology can do — not what employers will do.
Tasks this bears on
Searching slides for small findings
Looking through many slides for small things, such as metastases in lymph nodes.
Being augmented✓ Evidence-backed
Where this applies
A multireader, multicase study in which six pathologists reviewed 70 digitised lymph-node slides with and without an algorithm outlining likely tumour. Assisted sensitivity for micrometastases was 91% against 83%, and average review time per image was 61 against 116 seconds for micrometastases and 111 against 137 seconds for negative images. The authors who built and tested the algorithm are employees of Alphabet.
What this means
Searching for small metastases is where AI help makes pathologists both more accurate and faster.
What it does not yet show
A 70-slide reader study by the developers; not a measure of clinical workloads.
What you can check
Open the American Journal of Surgical Pathology article on deep learning assistance for lymph nodes and find "91% vs. 83%".
Does it change the assessment?
No — and this stage does not move it either. A "Capability" record is real evidence, but it does not upgrade a task judgement on its own. The 1 linked judgement above stand where they were.
Source
Steiner et al. — Impact of Deep Learning Assistance on the Histopathologic Review of Lymph Nodes for Metastatic Breast Cancer, American Journal of Surgical Pathology 42(12) (December 2018) · verified 2026-09-30 · Claude (VOLO agent) · interpreted 2026-09-30 · Claude (VOLO agent)
Primary source — published by the party that did this, or the authority of record. No co-signature needed.