ASN's Mission

To create a world without kidney diseases, the ASN Alliance for Kidney Health elevates care by educating and informing, driving breakthroughs and innovation, and advocating for policies that create transformative changes in kidney medicine throughout the world.

learn more

Contact ASN

1401 H St, NW, Ste 900, Washington, DC 20005

email@asn-online.org

202-640-4660

The Latest on Twitter

Kidney Week

Abstract: SA-PO006

Development of a Multimodal "Kidney Age" Prediction Based on Automatic Segmentation CT Image in Patients With Normal Renal Function: A Preliminary Report

Session Information

  • Bioengineering
    November 05, 2022 | Location: Exhibit Hall, Orange County Convention Center‚ West Building
    Abstract Time: 10:00 AM - 12:00 PM

Category: Bioengineering

  • 300 Bioengineering

Authors

  • Hou, Zuoxian, Peking Union Medical College Hospital, Beijing, Beijing, China
  • Xia, Peng, Peking Union Medical College Hospital, Beijing, Beijing, China
  • Zhang, Gu-Mu-Yang, Peking Union Medical College Hospital, Beijing, Beijing, China
  • Sun, Hao, Peking Union Medical College Hospital, Beijing, Beijing, China
  • Chen, Limeng, Peking Union Medical College Hospital, Beijing, Beijing, China
Background

The physiology volumes of the kidney cortex and medulla are presumed to change with age. We established a machine learning model to predict the "Kidney Age" in patients with normal serum creatinine (Scr) levels based on clinical data and an auto-segmentation algorithm separating the kidney cortex and medulla using contrast CT images.

Methods

We recruited 238 patients with normal Scr levels and contrast CT images between Oct 2021 and Feb 2022 in Peking Union Medical College Hospital with their demographic and clinical data. An auto-segmentation method was used for both cortex and medullary separation and their volume calculation, respectively. We combined the kidney volume, as well as clinical data for multimodal features of the machine learning model. All data were separated into a training dataset (85%) with ten-fold cross-validation and a test dataset (15%) for accuracy validation. Multiple machine learning models (n=100) with different initial weights are ensembled to reduce the prediction error. The performance of model was measured by the 95% confidential interval generated from the mean value and standard deviation.

Results

A total of 149 female patients and 89 male patients were included, with a mean age of 48.9±14.7 years old and Scr of 64.49±13.97 μmol/L. Their mean total kidney volume was 284.23±55.34 mm3, using the algorithm separating the kidney volumes of cortex and medulla. The predicted "Kidney Age" is approximately close to the patients' true age, with 92% prediction within the 95% confidential interval. The associated factors of the "Kidney Age" were eGFR (r=-0.516, p<0.001), hypertension (r=0.448, p<0.001), diabetic (r=0.364, p<0.001), kidney cortex volume (r=-0.374, p<0.001), and cortex ratio (r=-0.267, p<0.001).

Conclusion

We established a machine learning model for predicting the "Kidney Age" of patients with normal kidney function based on contrast CT images and clinical data.

Process of multimodal machine learning algorithm

Funding

  • Government Support – Non-U.S.