Comparison of AI-generated radiology impressions: a multi-stakeholder evaluation.

Sharang Phadke¹, Nivedita Suresh¹, Zachary Allen¹, Anjali Balagopal¹, Stephen Chan¹, Anish Shah², Megan Winter², Cesar Lam², Trevor Rose², Cyrillo Araujo², Abraham Ahmed², Iman Imanirad², Lincoln Berland¹, Andrew Del Gaizo^3,4

¹Rad AI, San Francisco, CA, USA.

²Moffitt Cancer Center, Tampa, FL, USA.

³Rad AI, San Francisco, CA, USA. andrew.delgaizo@moffitt.org.

⁴Moffitt Cancer Center, Tampa, FL, USA. andrew.delgaizo@moffitt.org.

PubMed PMC DOI PDF

Abstract

A retrospective, blinded evaluation of 200 oncologic computed tomography reports compared original radiologist-authored impressions, impressions generated by a custom domain-specific AI model fine-tuned on institutional data, and impressions generated by a general-purpose large language model. Ten clinicians, including original radiologists (n = 4), independent radiologists (n = 3), and oncologist…

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 Account Sign 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.