Health AI Central
Share:
Peer-Reviewed Publication
J Imaging Inform Med2026August 31, 2026Journal Article

From Redaction to Restoration: Deep Learning for Medical Image Deidentification and Reconstruction.

Adrienne Kline1,2,3,4, Abhijit Gaonkar5,6, Daniel Pittman7,8,9,10, Chris Kuehn10, Nils Forkert11,12
1Center for Artificial Intelligence, BCVI, Northwestern Medicine, Chicago, IL, USA. adrienne.kline@northwestern.edu.
2Department of Electrical and Computer Engineering, Northwestern University, Chicago, IL, USA. adrienne.kline@northwestern.edu.
3Department of Surgery, Northwestern University, Chicago, IL, USA. adrienne.kline@northwestern.edu.
4Xtasis Inc., Chicago, IL, USA. adrienne.kline@northwestern.edu.
5Medtronic, Minneapolis, MN, USA.
6ScaleCapacity, Burbank, CA, 91502, USA.
7Center for Artificial Intelligence, BCVI, Northwestern Medicine, Chicago, IL, USA.
8Department of Electrical and Computer Engineering, Northwestern University, Chicago, IL, USA.
9Department of Surgery, Northwestern University, Chicago, IL, USA.
10Xtasis Inc., Chicago, IL, USA.
11Department of Radiology, University of Calgary, Calgary, Canada.
12Hotchkiss Brain institute, University of Calgary, Calgary, Canada.
PubMed DOI

Abstract

Removing patient-identifying information from medical images is a prerequisite for sharing image data directly, as in public dataset release and open benchmarks, where the images themselves, rather than only model updates must leave the originating institution. However, many methods currently used for de-identification, e.g., cropping or blacking out image regions to eliminate burned-in text, can…

Create a free account to keep reading

Free members get 10 full research views every month across publications, clinical trials, FDA clearances, adverse events, and NIH grants. No credit card required.

Create Free AccountSign In

Want unlimited research access? See Pro plans

Data Accuracy Notice: Research intelligence on Health AI Central is aggregated from public sources (PubMed, ClinicalTrials.gov, FDA, NIH, CMS, and others) and refreshed nightly. Classifications and derived metrics are produced by automated methods described in our Methodology. We recommend verifying critical data points against the primary sources before making decisions.