Early Prediction of Chronic Kidney Diseases Using Machine Learning

Chronic Kidney Disease (CKD) is a long-term condition impacting millions of people worldwide, and early detection plays a crucial role in improving patient outcomes. However, conventional diagnostic methods-such as relying on serum creatinine levels to estimate glomerular filtration rate (eGFR)-may overlook early warning signs. This study introduces an integrated deep learning framework aimed at identifying CKD in its early stages by combining multiple data sources: kidney imaging, clinical measurements, and patientreported symptoms. The proposed system utilizes a customized DenseNet-based convolutional neural network (CNN) to analyze kidney scans (e.g., ultrasound, CT) and detect abnormalities. Alongside this, a feedforward neural network processes structured clinical dataincluding lab results, vitals, and demographic information-to assess CKD risk and severity. To incorporate patient experiences, a natural language processing (NLP) module uses TF-IDF techniques to extract meaningful patterns from free-text symptom descriptions. All three data streams are merged into a unified model that predicts CKD status, estimates eGFR, and classifies disease stage. A user-friendly web interface allows clinicians or patients to upload images, enter clinical details, and describe symptoms, with the model returning real-time predictions, confidence levels, and interpretability features. Preliminary results suggest that this multi-modal approach offers improved predictive accuracy compared to models relying on a single data source. This system shows potential as a clinical decision-support tool, especially for facilitating early intervention and personalized care in CKD management.

• Publication year

  2025

• Venue

  2025 IEEE First International Conference on Innovations in Engineering and Next-Generation Technologies for Sustainability (ICINVENTS)

• Publication date

  2025-11-07

• Fields of study

  Not labeled

• Identifiers
  DOI 10.1109/ICINVENTS64613.2025.11401561
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• KidneyNet: A Novel CNN-Based Technique for the Automated Diagnosis of Chronic Kidney Diseases from CT Scans

  2024cited by this paper
• Deep learning-based ultrasonographic classification of canine chronic kidney disease

  2024cited by this paper
• Promises of Big Data and Artificial Intelligence in Nephrology and Transplantation

  2020cited by this paper
• Automation of the kidney function prediction and classification through ultrasound-based kidney imaging using deep learning

  2019cited by this paper
• Scalable and accurate deep learning with electronic health records

  2018cited by this paper
• A survey on deep learning in medical image analysis

  2017cited by this paper
• Densely Connected Convolutional Networks

  2016cited by this paper

• No citing papers are available for this paper.