English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

During the diagnostic process, doctors incorporate multimodal informationincluding imaging and the medical history - and similarly medical AIdevelopment has increasingly become multimodal. In this paper we tackle a moresubtle challenge: doctors take a targeted medical history to obtain only themost pertinent pieces of information; how do we enable AI to do the same? Wedevelop a wrapper method named MINT (Make your model INTeractive) thatautomatically determines what pieces of information are most valuable at eachstep, and ask for only the most useful information. We demonstrate the efficacyof MINT wrapping a skin disease prediction model, where multiple images and aset of optional answers to $25$ standard metadata questions (i.e., structuredmedical history) are used by a multi-modal deep network to provide adifferential diagnosis. We show that MINT can identify whether metadata inputsare needed and if so, which question to ask next. We also demonstrate that whencollecting multiple images, MINT can identify if an additional image would bebeneficial, and if so, which type of image to capture. We showed that MINTreduces the number of metadata and image inputs needed by 82% and 36.2%respectively, while maintaining predictive performance. Using real-world AIdermatology system data, we show that needing fewer inputs can retain usersthat may otherwise fail to complete the system submission and drop off withouta diagnosis. Qualitative examples show MINT can closely mimic the step-by-stepdecision making process of a clinical workflow and how this is different forstraight forward cases versus more difficult, ambiguous cases. Finally wedemonstrate how MINT is robust to different underlying multi-model classifiersand can be easily adapted to user requirements without significant modelre-training.

MINT: A wrapper to make multi-modal and multi-image AI models interactive