Differentiating kidney stones from phleboliths in unenhanced low-dose computed tomography using radiomics and machine learning

Autor: Jacques Klein, Steve P. Martin, Thomas Benoît De Perrot, Simon Burgermeister, Jeremy Hofmeister, Xavier Montet, Grégoire Feutry
Rok vydání: 2018
Předmět:
Male
Lithiasis/diagnostic imaging
Computed tomography
Lithiasis
computer.software_genre
Kidney Calculi/diagnostic imaging
030218 nuclear medicine & medical imaging
Machine Learning
0302 clinical medicine
Radiomics
Positive predicative value
Diagnosis
80 and over
Medicine
Tomography
Neuroradiology
ddc:616
Aged
80 and over
medicine.diagnostic_test
Low dose
Interventional radiology
General Medicine
Middle Aged
Acute Pain
X-Ray Computed/methods
030220 oncology & carcinogenesis
Female
Radiology
Acute Pain/etiology
medicine.symptom
Flank Pain/etiology
Algorithms
Adult
medicine.medical_specialty
Flank Pain
Machine learning
ddc:616.0757
Diagnosis
Differential
03 medical and health sciences
Kidney Calculi
Young Adult
Humans
Radiology
Nuclear Medicine and imaging
Renal colic
Aged
business.industry
medicine.disease
Differential
Kidney stones
Artificial intelligence
business
Tomography
X-Ray Computed
computer
Zdroj: European Radiology, Vol. 29, No 9 (2019) pp. 4776-4782
ISSN: 1432-1084
0938-7994
Popis: Distinguishing between kidney stones and phleboliths can constitute a diagnostic challenge in patients undergoing unenhanced low-dose CT (LDCT) for acute flank pain. We sought to investigate the accuracy of radiomics and a machine-learning classifier in differentiating between kidney stones and phleboliths on LDCT. Radiomics features were extracted following a semi-automatic segmentation of kidney stones and phleboliths for two independent consecutive cohorts of patients undergoing LDCT for acute flank pain. Radiomics features from the first cohort of patients (n = 369) were ultimately used to train a machine-learning model designed to distinguish kidney stones (n = 211) from phleboliths (n = 201). Classification performance was assessed on the second independent cohort (i.e., testing set) (kidney stones n = 24; phleboliths n = 23) using positive and negative predictive values (PPV and NPV), area under the receiver operating curves (AUC), and permutation testing. Our machine-learning classification model trained on radiomics features achieved an overall accuracy of 85.1% on the independent testing set, with an AUC of 0.902, PPV of 81.5%, and NPV of 90.0%. Classification accuracy was significantly better than chance on permutation testing (p
Databáze: OpenAIRE
Externí odkaz: