Science

Breast cancer screening: AI manages to imitate the interpretation processes of expert radiologists

A new AI platform mimics radiologists’ interpretation processes when detecting cancerous lesions on mammograms.

The platform, which was developed by radiologists and computer scientists at Duke University, also shows All the steps that led to this or that diagnosis.

Unlike other types of AI that perform the same cancer detection tasks, this new assistant is completely transparent. This is of great importance for physicians and radiologists, who can thus validate or invalidate diagnostics which will emit

“We need algorithms that not only work, but also explain and provide the cases on which they base their conclusions,” said Joseph Lo, a professor of radiology at Duke.

“In this way, AI helps make better decisionsWhether or not the doctor agrees with your cancer diagnosis,” he concluded.

An AI assistant that mimics expert radiologists

In order for this new AI assistant to be able to perform analyzes like a radiologist, he first had to be trained as such. Thus he was taught to recognize suspicious lesions and to ignore all surrounding tissue; but also to focus on certain specific parts of a lesion so that it detects any abnormality more effectively.

In summary, the essential skills for a skilled radiologist have served as learning database.

In detail, artificial intelligence was able to train with 1,136 images of 484 patients with cancerous lesions and known as benign. This allowed him to distinguish minor anomalies from those that warrant further investigation.

“Other RNs don’t try to imitate radiologists; develop their own methods to answer their own questions, which are often irrelevant,” said Alina Barnett, a doctoral student in computer science at Duke.

Cancer detection: AI does not do it better than radiologists

However, it should be emphasized that when AI performance was compared to that of radiologists, they were not better. However, they were just as good as comparable AI platforms.

However, the platform developed by researchers at Duke University has the advantage of allowing radiologists to see exactly which part of the AI ​​interpretation is wrong, because it works transparently.

“This is what transparency AI for medical imaging, and that’s what doctors should be doing demand for any radiological tests” aimed at detecting cancer, said Cynthia Rudin, a professor of electrical engineering at Duke.

PHP Script, Elementor Pro Weadown, WordPress Theme, Fs Poster Plugin Nulled, Newspaper – News & WooCommerce WordPress Theme, Wordfence Premium Nulled, Dokan Pro Nulled, Plugins, Elementor Pro Weadown, Astra Pro Nulled, Premium Addons for Elementor, Yoast Nulled, Flatsome Nulled, Woocommerce Custom Product Ad, Wpml Nulled,Woodmart Theme Nulled, PW WooCommerce Gift Cards Pro Nulled, Avada 7.4 Nulled, Newspaper 11.2, Jannah Nulled, Jnews 8.1.0 Nulled, WP Reset Pro, Woodmart Theme Nulled, Business Consulting Nulled, Rank Math Seo Pro Weadown, Slider Revolution Nulled, Consulting 6.1.4 Nulled, WeaPlay, Nulledfire

Back to top button