Gianmaria Silvello invited speaker at the second Swetaly workshop on AI, will describe #ExaMode and the latest advancements of the project at 11:00am
oru.se/kalendarium/konferens…
The ExaMode results on "training computer-aided diagnosis models without human annotations" are on a spotlight on the Nature Health Portfolio @NaturePortfoliohealthcommunity.nature.com/p…
The @examode consortium's latest result on cancer diagnostics models for empowering digital pathology by limiting human effort to a minimum: nature.com/articles/s41746-0…
A thread 1/8 🧵
The implications are potentially disruptive: the need for medical experts to analyze and annotate healthcare data might be removed, leading to unprecedented possibilities in exploiting data already available in hospitals.
7/8
The approach is independent of the tissue and type of images analyzed. It can be replicated on data coming from different organs (lungs, prostate, head) and, most importantly, on different types of images.
8/8
The combination of clinical healthcare data with #AI technologies leads to new algorithms assisting the #diagnosis and the #decisionmaking, even for #cancer. This revolution has been slowed by the requirement of #manualannotations to train AI systems with prohibitive costs.
2/8
ExaMode is sponsoring #TPDL2022 and supporting the keynotes. Check out this great keynote about software preservation a topic relevant to all fields related and connected to #computerScience