How Is AI Transforming Radiology?

Why It Matters

Accurate, AI‑generated metrics turn radiology into a cost‑effective, data‑driven service, guiding high‑price therapies and easing staffing pressures across hospitals.

Key Takeaways

• AI augments, not replaces, radiologists within next five years.
• Imaging volume growing exponentially while radiologist training remains flat.
• Narrative reports discard quantitative data, creating a critical information gap.
• AI segmentation can measure kidney volume in minutes, not hours.
• Precise volume metrics guide tolvaptan use, saving costly ineffective treatments.

Summary

In a recent talk, a Cornell researcher outlined how artificial intelligence is reshaping radiology. He disclosed advisory roles with two Cornell‑spun companies and emphasized an ethic of open, transparent science. The presentation set out to contrast early, overly‑optimistic forecasts with the current, more nuanced reality.

Radiology today is already a computational pipeline: image acquisition, reconstruction, and interpretation rely on algorithms. However, the field faces a perfect storm—exponential growth in scan volume, stagnant radiologist training, and a data bottleneck that compresses millions of pixels into narrative reports that lack quantitative rigor. This mismatch creates diagnostic uncertainty and legal risk.

The speaker illustrated the challenge with autosomal dominant polycystic kidney disease (ADPKD). Traditional kidney‑volume measurement uses three linear dimensions, yielding ~10 % error and taking ten minutes per sequence. His team built an AI segmentation model that delineates kidney contours across multiple MRI sequences, delivering precise volumes in minutes and enabling reliable monitoring of tolvaptan therapy, a $200 k‑per‑year drug effective in only half of patients.

These advances signal a shift from AI‑driven automation to AI‑enabled augmentation. By extracting quantitative biomarkers at scale, radiology can improve treatment selection, reduce unnecessary drug spend, and alleviate radiologist workload. Within five years, AI is expected to become a standard decision‑support layer rather than a replacement for human expertise.

Original Description

FOLLOW US

Bluesky: https://bsky.app/profile/cornelltech.bsky.social

Facebook: https://www.facebook.com/CornellTech

Instagram: https://www.instagram.com/cornell_tech

LinkedIn: https://www.linkedin.com/school/cornell-tech

TikTok: https://www.tiktok.com/@cornelltech

X: https://twitter.com/cornell_tech

Website: https://tech.cornell.edu

ABOUT CORNELL TECH

At Cornell Tech, we develop the leaders and technologies of tomorrow through foundational and applied research, postgraduate education, and new ventures. Through rigorous, practical research and interdisciplinary innovation, we’re advancing lasting economic and social prosperity for New York City and the world.

In 2010, the City of New York issued a challenge to top institutions from around the world to propose a new or expanded applied sciences and engineering campus in New York City. Cornell University and the Technion­–Israel Institute of Technology— two of the world’s leading institutions in the fields of science, engineering, technology, and research—were selected from among 18 proposals to make this vision a reality. Shortly thereafter, the Jacobs Technion-Cornell Institute was founded based on a $133 million gift from Dr. Irwin Mark Jacobs, Founding Chairman and CEO Emeritus of Qualcomm, and his wife, Joan Klein Jacobs.

Cornell Tech first launched in 2012 in Google’s New York City building. In 2017, it migrated to its permanent location on Roosevelt Island—one of the most technically and environmentally advanced campuses in the world. Today, it is home to close to 50 tenure track and nontenured faculty members, approximately 500 master’s students across eight programs, and more than 100 doctoral students in five fields. We are committed to developing principled and innovative leaders who are practiced in interdisciplinary collaboration and entrepreneurship.

As part of a land-grant institution, Cornell Tech is focused on giving back to New York City. We apply our technologies in industries or sectors where New York City is leading the country and the world, including health, urban systems, media and democracy, cryptocurrencies and contracts, and public interest technology. Many of our graduates also stay in New York, supporting the continued growth of a vibrant tech ecosystem.