BD-KDD: A Clinical Dataset of Kidney Disease Diagnosis

This data article introduces BD-KDD (Clinical Dataset for Kidney Disease Diagnosis and Healthy Classification), a comprehensive clinical and laboratory dataset containing 988 patient records. It was developed to address the limited sample size of commonly used renal research benchmarks and provides more than twice the number of instances compared to the standard UCI Chronic Kidney Disease dataset. The dataset includes 26 variables, combining demographic information (Age), physical examination data (Blood Pressure), and 24 laboratory biomarkers, such as Serum Creatinine (Sc), Blood Urea (Bu), Hemoglobin (Hemo), and Specific Gravity (Sg). Each record is annotated with a binary target variable (Class) that categorizes patients as either Healthy (n = 481) or Kidney Disease (n = 507), enabling reliable development and evaluation of diagnostic machine learning models.