Gender‐specific data‐driven adiposity subtypes using deep‐learning‐based abdominal computed tomography (CT) segmentation
Authors' conclusion
This study identified gender-specific abdominal adiposity subgroups, which may help clinicians to distinguish diabetes risk quickly and automatically.
Study context
- Population
- Participants of the Pinggu Metabolic Disease Study
- Participants
- 3,817 participants
- Exposure
- Data-driven adiposity subtypes, by gender
- Outcome
- Diabetes risk across abdominal adiposity subtypes
Methodological detailsdeep-learning-based abdominal CT segmentation, K-means clustering
- Deep-learning-based abdominal CT segmentation
- K-means clustering
Main finding
As reported in the abstract
Diabetes risk by adiposity subtype
Each group vs SFDAdjusted odds ratio (95% CI)
MFD
1.92(1.32–2.78)
VFD
6.14(4.18–9.03)
- Men: VFD vs SFD higher diabetes risk, adjusted for age and BMI
- 60%
Observational data: associations, not effects.
Further findings
Dice indices, A-CT model vs manual evaluation
- Liver fat
- 0.96
- Muscle fat
- 0.95
- Subcutaneous fat
- 0.92
- Validated in 100 randomly selected cases
Original publication
Gender‐specific data‐driven adiposity subtypes using deep‐learning‐based abdominal CT segmentation. Obesity (Silver Spring). Published 8 May 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →