What is acromegaly and how is it detected by AI?

Acromegaly is a rare condition caused by excessive growth hormone production, leading to abnormal bone and organ growth. An AI system now detects it using images of the back of the hand and clenched fist.

Why does the AI use hand images instead of facial photos for acromegaly detection?

The AI uses hand images to protect patient privacy by avoiding facial recognition, which can raise concerns, and because acromegaly often causes visible changes in the hands.

How accurate is the AI in diagnosing acromegaly compared to doctors?

The AI model has demonstrated very high diagnostic accuracy, even outperforming experienced endocrinologists who evaluated the same hand photographs.

Home / Technology / AI Spots Rare Disease Using Only Hands

AI Spots Rare Disease Using Only Hands

4 Mar

Summary

AI analyzes hand images to detect acromegaly, a slow-progressing growth hormone disorder.
Privacy-focused AI uses back-of-hand images, avoiding facial recognition concerns.
AI model demonstrated higher diagnostic accuracy than experienced endocrinologists.

A novel artificial intelligence tool is poised to revolutionize the early detection of acromegaly, a rare condition characterized by excessive growth hormone production. This disorder, often developing gradually over many years, can lead to significant health complications if left untreated. Researchers at Kobe University have pioneered an AI system that analyzes images of the back of the hand and clenched fist to identify acromegaly.

This privacy-conscious approach was chosen to address patient concerns regarding facial recognition, a common method in previous AI diagnostic attempts. By focusing on hands, which exhibit distinct acromegaly-related changes, the AI can screen for the disease without compromising patient identity. Over 11,000 images from 725 patients across 15 Japanese medical institutions were used to train and test the model.

Remarkably, the AI model achieved high sensitivity and specificity, even outperforming experienced endocrinologists in diagnostic accuracy when evaluating the same hand photographs. Researchers believe this technology can serve as a valuable assistant to physicians, complementing clinical expertise and reducing diagnostic delays.

The success of this hand-image AI opens doors for its application in detecting other conditions with visible hand manifestations, such as rheumatoid arthritis and anemia. The development team envisions this technology as a key component in comprehensive health check-ups, helping to connect suspected cases to specialists and reduce healthcare disparities.

Disclaimer: This story has been auto-aggregated and auto-summarised by a computer program. This story has not been edited or created by the Feedzop team.

Home / Technology / AI Spots Rare Disease Using Only Hands

AI Spots Rare Disease Using Only Hands

4 Mar

•

Summary

AI analyzes hand images to detect acromegaly, a slow-progressing growth hormone disorder.
Privacy-focused AI uses back-of-hand images, avoiding facial recognition concerns.
AI model demonstrated higher diagnostic accuracy than experienced endocrinologists.

Disclaimer: This story has been auto-aggregated and auto-summarised by a computer program. This story has not been edited or created by the Feedzop team.