Texture analysis and machine learning algorithms accurately predict histologic grade in small (< 4 cm) clear cell renal cell carcinomas: a pilot study

• Published in: Abdominal imaging Vol. 45; no. 3; pp. 789 - 798
• Main Authors: Haji-Momenian, Shawn, Lin, Zixian, Patel, Bhumi, Law, Nicole, Michalak, Adam, Nayak, Anishsanjay, Earls, James, Loew, Murray
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.03.2020; Springer Nature B.V
• Subjects: Algorithms; Bladder; Carcinoma, Renal Cell - diagnostic imaging; Carcinoma, Renal Cell - pathology; Clear cell-type renal cell carcinoma; Decision trees; Female; Gastroenterology; Hepatology; Histograms; Humans; Imaging; Kidney cancer; Kidney Neoplasms - diagnostic imaging; Kidney Neoplasms - pathology; Kidneys; Learning algorithms; Machine Learning; Male; Mathematical analysis; Medicine; Medicine & Public Health; Middle Aged; Neoplasm Grading; Pilot Projects; Radiographic Image Interpretation, Computer-Assisted - methods; Radiology; Retroperitoneum; Retrospective Studies; Statistical analysis; Support vector machines; Texture; Tomography, X-Ray Computed; Tumors; Ureters; Clear cell renal cell carcinoma; Histology; Texture; Machine learning
• ISSN: 2366-004X; 2366-0058; 2366-0058
• DOI: 10.1007/s00261-019-02336-1

Purpose To predict the histologic grade of small clear cell renal cell carcinomas (ccRCCs) using texture analysis and machine learning algorithms. Methods Fifty-two noncontrast (NC), 26 corticomedullary (CM) phase, and 35 nephrographic (NG) phase CTs of small (< 4 cm) surgically resected ccRCCs were retrospectively identified. Surgical pathology classified the tumors as low- or high-Fuhrman histologic grade. The axial image with the largest cross-sectional tumor area was exported and segmented. Six histogram and 31 texture (gray-level co-occurrences (GLC) and gray-level run-lengths (GLRL)) features were calculated for each tumor in each phase. T testing compared feature values in low- and high-grade ccRCCs, with a (Benjamini–Hochberg) false discovery rate of 10%. Area under the receiver operating curve (AUC) was calculated for each feature to assess prediction of low- and high-grade ccRCCs in each phase. Histogram, texture, and combined histogram and texture data sets were used to train and test four algorithms (k-nearest neighbor (KNN), support vector machine (SVM), random forests, and decision tree) with tenfold cross-validation; AUCs were calculated for each algorithm in each phase to assess prediction of low- and high-grade ccRCCs. Results Zero, 23, and 0 features in the NC, CM, and NG phases had statistically significant differences between low and high-grade ccRCCs. CM histogram skewness and GLRL short run emphasis had the highest AUCs (0.82) in predicting histologic grade. All four algorithms had the highest AUCs (0.97) predicting histologic grade using CM histogram features. The algorithms’ AUCs decreased using histogram or texture features from NC or NG phases. Conclusion The histologic grade of small ccRCCs can be accurately predicted with machine learning algorithms using CM histogram features, which outperform NC and NG phase image data.