Automated CT Scans May Speed Up Clinical Assessments

NIH-funded research team develops versatile machine learning model to analyze medical scans and predict disease onset.

Published on Mar. 5, 2026

Got story updates? Submit your updates here. ›

A research team funded by the National Institutes of Health (NIH) has developed a versatile machine learning model called Merlin that can analyze 3D abdominal computed tomography (CT) scans to accomplish a wide range of tasks, from identifying anatomical features to predicting the onset of chronic diseases years in advance. Merlin outperformed specialized models in various diagnostic, prognostic, and quality assessment activities, showcasing its potential to streamline medical workflows and assist in clinical decision-making.

Why it matters

The use of CT scans is common in the early stages of medical evaluations, but the process of interpreting the results and reaching a diagnosis can be lengthy and cumbersome, especially with the growing shortage of physicians in the United States. Merlin's ability to automate and expedite this process could significantly improve the efficiency of clinical assessments and lead to earlier disease detection and intervention.

The details

The researchers trained Merlin on a large dataset of over 15,000 3D abdominal CT scans paired with their radiology reports and nearly one million diagnostic codes. This allowed the model to learn the relationships between visual and written data. Merlin was then tested on more than 50,000 previously unseen abdominal CT scans from four different hospitals. On average, Merlin successfully predicted which of two scans was more likely to be associated with a particular diagnostic code over 81% of the time, outperforming several other models. The team also found that Merlin could identify patients at higher risk of developing chronic diseases like diabetes, osteoporosis, and heart disease in the next five years 75% of the time, compared to 68% for other models.

  • The research was supported by grants from the National Institutes of Health (NIH) and its various institutes, including the National Institute of Biomedical Imaging and Bioengineering (NIBIB), the National Heart, Lung, and Blood Institute (NHLBI), and the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS).
  • The study was conducted by a research team at Stanford University, with the work being carried out by graduate students Louis Blankemeier and Ashwin Kumar.

The players

Merlin

A versatile machine learning model developed by the research team that can analyze 3D abdominal computed tomography (CT) scans and perform a wide range of tasks, from identifying anatomical features to predicting the onset of chronic diseases.

Bruce Tromberg, Ph.D.

The director of the National Institute of Biomedical Imaging and Bioengineering (NIBIB) at the National Institutes of Health (NIH).

Louis Blankemeier, Ph.D.

A co-first author of the study, who conducted the research while a graduate student at Stanford University.

Ashwin Kumar

A co-first author of the study, who is a graduate student at Stanford University.

Akshay Chaudhari, Ph.D.

The senior author of the study, who is a professor of radiology and biomedical data science at Stanford University.

Got photos? Submit your photos here. ›

What they’re saying

“Rich datasets like this are necessary to push the limits of what artificial intelligence models can accomplish in medicine. This work exemplifies how meticulously crafted training data can enable remarkable insights that significantly streamline workflows and assist in clinical decision-making.”

— Bruce Tromberg, Ph.D., Director of the National Institute of Biomedical Imaging and Bioengineering (NIBIB) at the National Institutes of Health (NIH) (Mirage News)

“With Merlin, you could potentially go beyond traditional radiology and jump straight from imaging to a possible diagnosis. And that's just one potential use.”

— Louis Blankemeier, Ph.D., Co-first author of the study (Mirage News)

“Merlin tackled some tasks, such as predicting diagnosis codes, head-on, while other more complicated tasks, such as drafting radiology reports from scratch or identifying and outlining organs in a 3D space, called for additional training.”

— Ashwin Kumar, Co-first author of the study (Mirage News)

“Our model and the data will provide the community a robust backbone to build upon. From here, the sky's the limit.”

— Akshay Chaudhari, Ph.D., Senior author of the study (Mirage News)

What’s next

The researchers plan to refine Merlin to better handle more complicated challenges, such as report writing, and encourage users to fine-tune the model with their own data to address their specific needs.

The takeaway

Merlin's ability to automate and expedite the process of interpreting medical scans and predicting disease onset could significantly improve the efficiency of clinical assessments, leading to earlier disease detection and intervention. This research highlights the potential of advanced AI models to transform medical workflows and assist clinicians in making more informed decisions.