Back

Anatomical Accuracy of Generative AI for Congenital Heart Disease Illustrations: Gemini NanoBanana Versus ChatGPT Models in a Blinded Comparative Study

2026-02-23 cardiovascular medicine Title + abstract only
View on medRxiv
Show abstract

BackgroundGenerative artificial intelligence (AI) systems are increasingly used to produce medical illustrations for education; however, their anatomical accuracy in complex domains such as congenital heart disease (CHD) remains insufficiently validated. MethodsIn an assessor-blinded comparative study, we evaluated AI-generated CHD illustrations from two contemporary text-to-image platforms (ChatGPT-5/ChatGPT-Images and Gemini NanoBanana) against human-modified educational images. Twenty differ...

Predicted journal destinations

1
Scientific Reports
701 training papers
Top 16% 10.8%
2
npj Digital Medicine
85 training papers
Top 2% 8.9%
3
Circulation
37 training papers
Top 2% 7.7%
4
Frontiers in Cardiovascular Medicine
33 training papers
Top 3% 6.3%
5
European Heart Journal - Digital Health
15 training papers
Top 0.4% 6.3%
6
Journal of the American Heart Association
92 training papers
Top 6% 6.3%
7
PLOS ONE
1737 training papers
Top 77% 5.9%
8
Open Heart
18 training papers
Top 1% 4.9%
9
Heart Rhythm
16 training papers
Top 2% 3.8%
10
BMJ Open
553 training papers
Top 40% 3.8%
11
The American Journal of Cardiology
15 training papers
Top 3% 3.0%
12
Computers in Biology and Medicine
39 training papers
Top 2% 2.7%
13
Circulation: Genomic and Precision Medicine
30 training papers
Top 4% 2.1%
14
PLOS Digital Health
88 training papers
Top 10% 1.9%
15
Nature Medicine
88 training papers
Top 5% 1.5%
16
Journal of Clinical Medicine
77 training papers
Top 12% 1.5%
17
Journal of the American Medical Informatics Association
53 training papers
Top 5% 1.5%
18
eLife
262 training papers
Top 29% 1.5%
19
Nature
58 training papers
Top 4% 1.3%
20
Nature Communications
483 training papers
Top 47% 1.3%
21
Frontiers in Physiology
18 training papers
Top 1% 1.3%
22
Atherosclerosis
16 training papers
Top 5% 0.9%
23
Cureus
64 training papers
Top 17% 0.9%
24
Hypertension
20 training papers
Top 6% 0.9%
25
Frontiers in Neurology
74 training papers
Top 14% 0.7%