Related Experiment Video
Updated: Jul 14, 2025
Use of 3D Robotic Ultrasound for In Vivo Analysis of Mouse Kidneys
Published on: August 12, 2021
Test Retest Reproducibility of Organ Volume Measurements in ADPKD Using 3D Multimodality Deep Learning
Xinzi He1, Zhongxiu Hu2, Hreedi Dev2
1School of Electrical and Computer Engineering, Cornell University and Cornell Tech, New York, New York (X.H., R.S.); Department of Radiology, Weill Cornell Medicine, New York, New York (X.H., Z.H., H.D., D.J.R., A.S., S.I.R., S.J.W., K.T., G.S., A.G., R.S., M.R.P.).
Rationale And Objectives:
Following autosomal dominant polycystic kidney disease (ADPKD) progression by measuring organ volumes requires low measurement variability. The objective of this study is to reduce organ volume measurement variability on MRI of ADPKD patients by utilizing all pulse sequences to obtain multiple measurements which allows outlier analysis to find errors and averaging to reduce variability.
Materials And Methods:
In order to make measurements on multiple pulse sequences practical, a 3D multi-modality multi-class segmentation model based on nnU-net was trained/validated using T1, T2, SSFP, DWI and CT from 413 subjects. Reproducibility was assessed with test-re-test methodology on ADPKD subjects (n = 19) scanned twice within a 3-week interval correcting outliers and averaging the measurements across all sequences. Absolute percent differences in organ volumes were compared to paired students t-test.
Results:
Dice similarlity coefficient > 97%, Jaccard Index > 0.94, mean surface distance < 1 mm and mean Hausdorff Distance < 2 cm for all three organs and all five sequences were found on internal (n = 25), external (n = 37) and test-re-test reproducibility assessment (38 scans in 19 subjects). When averaging volumes measured from five MRI sequences, the model automatically segmented kidneys with test-re-test reproducibility (percent absolute difference between exam 1 and exam 2) of 1.3% which was better than all five expert observers. It reliably stratified ADPKD into Mayo Imaging Classification (area under the curve=100%) compared to radiologist.
Conclusion:
3D deep learning measures organ volumes on five MRI sequences leveraging the power of outlier analysis and averaging to achieve 1.3% total kidney test-re-test reproducibility.
Insights
This study developed a 3D deep learning model to measure organ volumes in autosomal dominant polycystic kidney disease (ADPKD) patients using multiple MRI sequences. The model achieved highly reproducible kidney volume measurements, improving accuracy for disease progression tracking.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Medicine
- Nephrology
Background:
- Autosomal dominant polycystic kidney disease (ADPKD) progression monitoring relies on organ volume measurement, which requires high precision.
- Existing methods for measuring organ volumes in ADPKD patients using MRI exhibit significant variability.
- Reducing measurement variability is crucial for accurately tracking disease progression and evaluating treatment efficacy.
Purpose of the Study:
- To develop and validate a 3D deep learning model for accurate organ volume measurement in ADPKD patients.
- To reduce measurement variability in MRI-based organ volume quantification by utilizing multiple pulse sequences.
- To improve the reliability of ADPKD progression assessment through enhanced imaging analysis.
Main Methods:
- A 3D multi-modality, multi-class segmentation model (nnU-net) was trained and validated on T1, T2, SSFP, and DWI MRI sequences from 413 subjects.
- Test-re-test reproducibility was assessed on 19 ADPKD subjects scanned twice within a 3-week interval.
- Organ volumes were measured across five MRI sequences, with outlier analysis and averaging applied to minimize variability.
Main Results:
- The segmentation model achieved high accuracy with Dice similarity coefficients >97% and Jaccard Indices >0.94 across multiple organs and sequences.
- The model demonstrated excellent test-re-test reproducibility for kidney volume measurements, with an average absolute difference of 1.3%.
- This automated approach surpassed the reproducibility of five expert observers and reliably stratified ADPKD patients into Mayo Imaging Classification categories.
Conclusions:
- A 3D deep learning model effectively measures organ volumes from multiple MRI sequences in ADPKD patients.
- The integration of outlier analysis and averaging significantly reduces measurement variability, achieving 1.3% test-re-test reproducibility for kidney volumes.
- This advanced imaging analysis technique offers a more reliable and accurate method for monitoring ADPKD progression.
More Related Videos
06:03Integrating Augmented Reality Tools in Breast Cancer Related Lymphedema Prognostication and Diagnosis
Published on: February 6, 2020
07:35Use of Ultra-high Field MRI in Small Rodent Models of Polycystic Kidney Disease for In Vivo Phenotyping and Drug Monitoring
Published on: June 23, 2015