Advertisement

Automated Machine Learning Segmentation and Measurement of Urinary Stones on CT Scan

  • Author Footnotes
    ⁎ These authors contributed equally to the work and share first co-authorship.
    Rilwan Babajide
    Footnotes
    ⁎ These authors contributed equally to the work and share first co-authorship.
    Affiliations
    University of Chicago Pritzker School of Medicine, Chicago, IL
    Search for articles by this author
  • Author Footnotes
    ⁎ These authors contributed equally to the work and share first co-authorship.
    Katerina Lembrikova
    Footnotes
    ⁎ These authors contributed equally to the work and share first co-authorship.
    Affiliations
    SUNY Downstate College of Medicine, Brooklyn, NY
    Search for articles by this author
  • Justin Ziemba
    Affiliations
    University of Pennsylvania Perelman School of Medicine, Philadelphia, PA

    Department of Surgery, Division of Urology, Hospital of the University of Pennsylvania, Philadelphia, PA
    Search for articles by this author
  • James Ding
    Affiliations
    University of Pennsylvania Perelman School of Medicine, Philadelphia, PA
    Search for articles by this author
  • Yuemeng Li
    Affiliations
    Department of Radiology, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA

    The Center for Biomedical Image Computing and Analytics, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA
    Search for articles by this author
  • Antoine Selman Fermin
    Affiliations
    Department of Surgery, Division of Pediatric Urology, The Children's Hospital of Philadelphia, Philadelphia, PA
    Search for articles by this author
  • Yong Fan
    Affiliations
    Department of Radiology, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA

    The Center for Biomedical Image Computing and Analytics, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA
    Search for articles by this author
  • Gregory E. Tasian
    Correspondence
    Address correspondence to: Gregory E. Tasian M.D., M.S.c., M.S.C.E., Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA 19104.
    Affiliations
    Department of Surgery, Division of Pediatric Urology, The Children's Hospital of Philadelphia, Philadelphia, PA

    Department of Biostatistics, Epidemiology, and Informatics; Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA
    Search for articles by this author
  • Author Footnotes
    ⁎ These authors contributed equally to the work and share first co-authorship.

      Abstract

      Objectives

      To evaluate the performance of an engineered machine learning algorithm to identify kidney stones and measure stone characteristics without the need for human input.

      Methods

      We performed a cross-sectional study of 94 children and adults who had kidney stones identified on non-contrast CT. A previously developed deep learning algorithm was trained to segment renal anatomy and kidney stones and to measure stone features. The performance and speed of the algorithm to measure renal anatomy and kidney stone features were compared to the current gold standard of human measurement performed by 3 independent reviewers.

      Results

      The algorithm was 100% sensitive and 100% specific in detecting individual kidney stones. The mean stone volume segmented by the algorithm was smaller than that of human reviewers and had moderate overlap (Dice score: 0.66). There was substantial variation between human reviewers in total segmented stone volume (Jaccard score: 0.17) and volume of the single largest stone (Jaccard score: 0.33). Stone segmentations performed by the machine learning algorithm more precisely approximated stone borders than those performed by human reviewers on qualitative assessment.

      Conclusion

      An engineered machine learning algorithm can identify and characterize stones more accurately and reliably than humans, which has the potential to improve the precision and efficiency of assessing kidney stone burden.

      Abbreviations:

      ICC (intraclass correlation), IRR (inter-rater reliability), ML (machine learning), NCCT (non-contrast computed tomography)
      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Urology
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

        • Türk C
        • Petrik A
        • Sarika K
        • et al.
        EAU guidelines on diagnosis and conservative management of urolithiasis.
        Eur Urol. 2016; 69: 468-474https://doi.org/10.1016/j.eururo.2015.07.040
        • Assimos D
        • Krambeck A
        • Miller NL
        • et al.
        Surgical management of stones: american urological association/endourological society guideline, PART I.
        J Urol. 2016; 196: 1153-1160https://doi.org/10.1016/j.juro.2016.05.090
        • Assimos D
        • Krambeck A
        • Miller NL
        • et al.
        Surgical management of stones: american urological association/endourological society guideline, PART II.
        J Urol. 2016; 196: 1161-1169https://doi.org/10.1016/j.juro.2016.05.091
        • Iremashvili V
        • Li S
        • Penniston KL
        • et al.
        Role of residual fragments on the risk of repeat surgery after flexible ureteroscopy and laser lithotripsy: single center study.
        J Urol. 2019; 201: 358-363https://doi.org/10.1016/j.juro.2016.05.091
        • Kim JK
        • Yook IH
        • Choi MJ
        • et al.
        A performance comparison on the machine learning classifiers in predictive pathology staging of prostate cancer.
        Stud Health Technol Inform. 2017; 245 (1273-1273)https://doi.org/10.3233/978-1-61499-830-3-1273
        • Algohary A
        • Viswanath S
        • Shiradkar R
        • et al.
        Radiomic features on MRI enable risk categorization of prostate cancer patients on active surveillance: preliminary findings.
        J Magn Reson Imaging. 2018; 48: 818-828https://doi.org/10.1002/jmri.25983
        • Ginsburg SB
        • Algohary A
        • Pahwa S
        • et al.
        Radiomic features for prostate cancer detection on MRI differ between the transition and peripheral zones: Preliminary findings from a multi- institutional study.
        J Magn Reson Imaging. 2017; 46: 184-193https://doi.org/10.1002/jmri.25562
        • Zheng H
        • Ji J
        • Zhao L
        • et al.
        Prediction and diagnosis of renal cell carcinoma using nuclear magnetic resonance-based serum metabolomics and self-organizing maps.
        Oncotarget. 2016; 7: 59189-59198https://doi.org/10.18632/oncotarget.10830
        • Ziemba JB
        • Li P
        • Gurnani R
        • et al.
        A user-friendly application to automate CT renal stone measurement.
        J Endourol. 2018; 32 (2018): 685-691https://doi.org/10.1089/end.2018.0326
        • Demehri S
        • Steigner ML
        • Sodickson AD
        • et al.
        CT-based determination of maximum ureteral stone area: a predictor of spontaneous passage.
        Am J Roentgenol. 2012; 198: 603-608https://doi.org/10.2214/AJR.11.7276
        • Bell JR
        • Posielski NM
        • Penniston KL
        • et al.
        Automated computer software compared with manual measurements for CT-based urinary stone metrics: an evaluation study.
        J Endourol. 2018; 32: 455-461https://doi.org/10.1089/end.2017.0787
        • Jendeberg J
        • Geijer H
        • Alshamari M
        • et al.
        Prediction of spontaneous ureteral stone passage: automated 3D-measurements perform equal to radiologists, and linear measurements equal to volumetric.
        Eur Radiol. 2018; 28: 2474-2483https://doi.org/10.1007/s00330-017-5242-9
        • Viswanath K
        • Gunasundari R.
        Analysis and implementation of kidney stone detection by reaction diffusion level set segmentation using xilinx system generator on FPGA.
        VLSI Des. 2015; 2015e581961https://doi.org/10.1155/2015/581961
        • Duan X
        • Wang J
        • Qu M
        • et al.
        Kidney stone volume estimation from computerized tomography images using a model based method of correcting for the point spread function.
        J Urol. 2012; 188: 989-995https://doi.org/10.1016/j.juro.2012.04.098
        • Lee SR
        • Jeon HG
        • Park DS
        • et al.
        Longitudinal stone diameter on coronal reconstruction of computed tomography as a predictor of ureteral stone expulsion in medical expulsive therapy.
        Urology. 2012; 80: 784-789https://doi.org/10.1016/j.urology.2012.06.032
        • Lidén M
        • Andersson T
        • Broxvall M
        • et al.
        Urinary stone size estimation: a new segmentation algorithm-based CT method.
        Eur Radiol. 2012; 22: 731-737https://doi.org/10.1007/s00330-011-2309-x
        • Patel SR
        • Stanton P
        • Zelinski N
        • et al.
        Automated renal stone volume measurement by noncontrast computerized tomography is more reproducible than manual linear size measurement.
        J Urol. 2011; 186: 2275-2279https://doi.org/10.1016/j.juro.2011.07.091
        • Selby MG
        • Vrtiska TJ
        • Krambeck AE
        • et al.
        Quantification of asymptomatic kidney stone burden by computed tomography for predicting future symptomatic stone events.
        Urology. 2015; 85: 45-50https://doi.org/10.1016/j.urology.2014.08.031
        • Zorba OÜ
        • Ogullar S
        • Yazar S
        • et al.
        CT-based determination of ureteral stone volume: a predictor of spontaneous passage.
        J Endourol. 2016; 30: 32-36https://doi.org/10.1089/end.2015.0481
        • Kurtzman JT
        • Song L
        • Ross ME
        • et al.
        Urology consultation and emergency department revisits for children with urinary stone disease.
        J Urol. 2018; 200: 180-186https://doi.org/10.1016/j.juro. 2018.02.069
        • Rorden C
        • Brett M.
        Stereotaxic display of brain lesions.
        Behav Neurol. 2000; 12: 191-200https://doi.org/10.1155/2000/421719
        • Harris PA
        • Taylor R
        • Minor BL
        • et al.
        The REDCap consortium: building an international community of software platform partners.
        J Biomed Inform. 2019; 95103208https://doi.org/10.1016/j.jbi.2019. 103208
        • Li Y
        • Li H
        • Fan Y.
        ACEnet: anatomical context-encoding network for neuroanatomy segmentation.
        Med Image Anal. 2021; 70101991https://doi.org/10.1016/j.media.2021.101991
        • Thein N
        • Nugroho HA
        • Adji TB
        • et al.
        An image preprocessing method for kidney stone segmentation in CT scan images.
        in: 2018 International Conference on Computer Engineering, Network and Intelligent Multimedia (CENIM). 2018: 147-150https://doi.org/10.1109/CENIM.2018.8710933
        • Cicchetti DV.
        Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology.
        Psychol Assess. 1994; 6: 284-290https://doi.org/10.1037/1040-3590.6.4.284
        • Visser M
        • Muller DMJ
        • van Dujin RJM
        • et al.
        Inter-rater agreement in glioma segmentations on longitudinal MRI.
        NeuroImage Clin. 2019; 22101727https://doi.org/10.1016/j.nicl.2019.101727
        • Deeley MA
        • Chen A
        • Datteri R
        • et al.
        Comparison of manual and automatic segmentation methods for brain structures in the presence of space-occupying lesions: a multi-expert study.
        Phys Med Biol. 2011; 56: 4557-4577https://doi.org/10.1088/0031-9155/56/14/021