60/100

Moderate evidence

Cite
More
EmailLinkedInX

Automated CT analysis in 127 breast cancer patients suggests muscle fat tracks heart risk better than BMI

Study title: Computed tomography derived cardiometabolic biomarkers in high-risk breast cancer using fully automated deep learning algorithms

Evidence type
Observational study
Sample
n = 127
Sources
1 sourcePeer-reviewed
Published
2 min read

Study title: Computed tomography derived cardiometabolic biomarkers in high-risk breast cancer using fully automated deep learning algorithms

Researchers at Johns Hopkins applied fully automated deep learning algorithms to baseline abdominal CT scans from 127 women diagnosed with breast cancer between 2016 and 2018. The algorithms derived measures of fat, muscle, liver fat and aortic calcification without manual work.

Read the study abstract

Purpose Cardiometabolic disease and obesity are important factors when considering breast cancer risk and survivorship. We aimed to utilize a fully automated deep learning algorithm to evaluate CT-derived cardiometabolic biomarkers in patients with breast cancer. Methods We applied fully automated deep learning algorithms to derive cardiometabolic markers from baseline computed tomography (CT) abdomen scans of females with breast cancer diagnosed between 2016 - 2018 at Johns Hopkins. Results Among 127 females (6…

Computed tomography derived cardiometabolic biomarkers in high-risk breast cancer using fully automated deep learning algorithms
"Prélèvement sérologie au Laboratoire Hopital Douala Cameroun 10" by Happi Raphael, CC BY-SA 4.0.

The gain

In a study of 127 women with breast cancer at Johns Hopkins, automated deep learning on CT scans measured body composition, liver fat and aortic plaque, and found myosteatosis and menopausal status tracked adiposity and aortic calcification where BMI did not.

In brief

BMI was positively correlated with CT-measured adipose tissue area and muscle area and negatively correlated with liver intensity, but it was not associated with aortic calcification. Myosteatosis and menopausal status were linked to both adiposity, especially visceral fat, and aortic calcification.

Myosteatosis was also associated with liver steatosis. The authors conclude the algorithms can be a tool for assessing metabolic disease risk in patients with breast cancer, which matters for survivorship.

Main points

  1. The study covered 127 women with breast cancer diagnosed between 2016 and 2018 at Johns Hopkins.
  2. Of the women, 64.6% were postmenopausal; 57.5% were White and 26.8% were Black.
  3. BMI was positively correlated with CT-derived adipose tissue area and muscle area.
  4. BMI was negatively correlated with liver intensity and not associated with aortic calcification.
  5. Myosteatosis and menopausal status were associated with visceral adiposity and aortic calcification, and myosteatosis with liver steatosis.

The full story

The team used fully automated deep learning algorithms on baseline CT abdomen scans to derive cardiometabolic markers in women with breast cancer, where obesity and cardiometabolic disease bear on risk and survivorship.

BMI tracked fat and muscle area and liver intensity but missed aortic calcification. Myosteatosis and menopausal status captured both fat distribution and calcification, so they outperformed BMI as markers in this group.

The authors conclude the algorithms can feasibly quantify body composition, liver fat and abdominal aortic plaque directly, offering a way to evaluate metabolic disease risk in this patient population.

Sources

1 on file
  1. Breast Cancer Research and Treatmentdoi:10.1007/s10549-026-08122-0
    Peer-reviewed2026-11-14

Talk it over

Three questions to start on, two more if you have time
BMI tracked fat and muscle area but was not associated with aortic calcification, while myosteatosis and menopausal status were. How would you weigh the benefit of automated CT measures against relying on a simple BMI figure?

Start with a question

Automated CT analysis in 127 breast cancer patients suggests muscle fat tracks heart risk better than BMIhttps://truvace.com/item/computed-tomography-derived-cardiometabolic-biomarkers-in-high-risk-breast-cance

  1. BMI tracked fat and muscle area but was not associated with aortic calcification, while myosteatosis and menopausal status were. How would you weigh the benefit of automated CT measures against relying on a simple BMI figure?

    Answer in the thread
  2. The study covered 127 women at Johns Hopkins, of whom 57.5% were White and 26.8% were Black. Who might be missing from that group, and how could that change the findings?

    Answer in the thread
  3. The women were scanned at baseline and the study reports associations, not heart outcomes. What follow-up results would show that these automated measures predict heart disease in breast cancer survivors?

    Answer in the thread
If you have more time
  • If you were a woman in this study, what would you want to be told about your own myosteatosis or aortic calcification measures, and by whom?

    Answer in the thread
  • Which hospitals or patient groups would you want to repeat this analysis, and what would you compare the algorithms against?

    Answer in the thread

What readers say