Histopathological correlations of CT-based radiomics imaging biomarkers in native kidney biopsy

• Published in: BMC medical imaging 2024-09, Vol.24 (1), p.256-11, Article 256
• Main Authors: Choi, Yoon Ho, Kim, Ji-Eun, Lee, Ro Woon, Kim, Byoungje, Shin, Hyeong Chan, Choe, Misun, Kim, Yaerim, Park, Woo Yeong, Jin, Kyubok, Han, Seungyeup, Paek, Jin Hyuk, Kim, Kipyo
• Format: Article
• Language: English
• Subjects: Abdomen; Adult; Aged; Atrophy; Biomarkers; Biopsy; Chronic kidney disease; Chronic kidney failure; Computed tomography; Correlation; CT imaging; Deep Learning; Diagnostic imaging; Feature extraction; Female; Fibrosis; Histopathology; Humans; Kidney - diagnostic imaging; Kidney - pathology; Kidney diseases; Kidney fibrosis; Kidneys; Lesions; Machine learning; Magnetic resonance imaging; Male; Medical imaging; Medical imaging equipment; Medical research; Medicine, Experimental; Microscopy; Middle Aged; Prediction models; Radiomics; Renal Insufficiency, Chronic - diagnostic imaging; Renal Insufficiency, Chronic - pathology; Resampling; Retrospective Studies; Statistical analysis; Statistical models; Texture; Tomography, X-Ray Computed - methods; Ultrasonic imaging
• ISSN: 1471-2342; 1471-2342
• DOI: 10.1186/s12880-024-01434-x

Kidney biopsy is the standard of care for the diagnosis of various kidney diseases. In particular, chronic histopathologic lesions, such as interstitial fibrosis and tubular atrophy, can provide prognostic information regarding chronic kidney disease progression. In this study, we aimed to evaluate historadiological correlations between CT-based radiomic features and chronic histologic changes in native kidney biopsies and to construct and validate a radiomics-based prediction model for chronicity grade. We included patients aged ≥ 18 years who underwent kidney biopsy and abdominal CT scan within a week before kidney biopsy. Left kidneys were three-dimensionally segmented using a deep learning model based on the 3D Swin UNEt Transformers architecture. We additionally defined isovolumic cortical regions of interest near the lower pole of the left kidneys. Shape, first-order, and high-order texture features were extracted after resampling and kernel normalization. Correlations and diagnostic metrics between extracted features and chronic histologic lesions were examined. A machine learning-based radiomic prediction model for moderate chronicity was developed and compared according to the segmented regions of interest (ROI). Overall, moderate correlations with statistical significance (P