Deep learning is a promising technology and seems to be the future of the CT stone evaluation

Danilovic, Alexandre

doi:10.1590/S1677-5538.IBJU.2022.0132.1

COMMENT

Computed tomography (CT) is the current gold standard diagnostic imaging exam for urolithiasis (¹1 Caglayan A, Horsanali MO, Kocadurdu K, Ismailoglu E, Guneyli S. Deep learning model-assisted detection of kidney stones on computed tomography. Int Braz J Urol. 2022;48:830-9.). However, making a CT report is a time-consuming process and requires a specialist. Therefore, an automated model of kidney stones detection would help saving health resources.

The authors of “Deep learning model-assisted detection of kidney stones on computed tomography” showed that a convolution-based algorithm, xResNet50, detected kidney stones with accuracy up to 85.0% for 0-1 cm stones, 89.0% for 1-2 cm stones and 93.0% for > 2 cm stones in CT sagittal section compared to experienced radiologists. Not surprisingly, larger stones are easier to detect (¹1 Caglayan A, Horsanali MO, Kocadurdu K, Ismailoglu E, Guneyli S. Deep learning model-assisted detection of kidney stones on computed tomography. Int Braz J Urol. 2022;48:830-9.). However, the accuracy of this automated model to detect kidney stones seems to be not sufficient to dismiss the specialist analysis. Although detection of stones is a good primary objective for an automated model, the mere detection of a kidney stone is not enough for clinical application. A complete report of the stone features is necessary for the best clinical decision. Also, the automated model algorithm should take in consideration CT settings as tube current and window as it impacts measurements of clinically relevant stone features such as size and density (²2 Danilovic A, Cavalanti A, Rocha BA, Traxer O, Torricelli FCM, Marchini GS, et al. Assessment of Residual Stone Fragments After Retrograde Intrarenal Surgery. J Endourol. 2018;32:1108-13., ³3 Danilovic A, Rocha BA, Marchini GS, Traxer O, Batagello C, Vicentini FC, et al. Computed tomography window affects kidney stones measurements. Int Braz J Urol. 2019;45:948-55.).

However, artificial intelligence is advancing fast. Other authors were able to show good agreement of other automated model algorithm with radiologist results for stone size, volume, location, number and density (⁴4 Elton DC, Turkbey EB, Pickhardt PJ, Summers RM. A deep learning system for automated kidney stone detection and volumetric segmentation on noncontrast CT scans. Med Phys. 2022;49:2545-54., ⁵5 Cui Y, Sun Z, Ma S, Liu W, Wang X, Zhang X, et al. Automatic Detection and Scoring of Kidney Stones on Noncontrast CT Images Using S.T.O.N.E. Nephrolithometry: Combined Deep Learning and Thresholding Methods. Mol Imaging Biol. 2021;23:436-45.). Deep learning is a promising technology and seems to be the future of the CT stone evaluation.

REFERENCES

¹
Caglayan A, Horsanali MO, Kocadurdu K, Ismailoglu E, Guneyli S. Deep learning model-assisted detection of kidney stones on computed tomography. Int Braz J Urol. 2022;48:830-9.
²
Danilovic A, Cavalanti A, Rocha BA, Traxer O, Torricelli FCM, Marchini GS, et al. Assessment of Residual Stone Fragments After Retrograde Intrarenal Surgery. J Endourol. 2018;32:1108-13.
³
Danilovic A, Rocha BA, Marchini GS, Traxer O, Batagello C, Vicentini FC, et al. Computed tomography window affects kidney stones measurements. Int Braz J Urol. 2019;45:948-55.
⁴
Elton DC, Turkbey EB, Pickhardt PJ, Summers RM. A deep learning system for automated kidney stone detection and volumetric segmentation on noncontrast CT scans. Med Phys. 2022;49:2545-54.
⁵
Cui Y, Sun Z, Ma S, Liu W, Wang X, Zhang X, et al. Automatic Detection and Scoring of Kidney Stones on Noncontrast CT Images Using S.T.O.N.E. Nephrolithometry: Combined Deep Learning and Thresholding Methods. Mol Imaging Biol. 2021;23:436-45.

Publication Dates

Publication in this collection
26 Aug 2022
Date of issue
Sep-Oct 2022

History

Received
02 June 2022
Accepted
06 June 2022

This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

[1] ¹
Caglayan A, Horsanali MO, Kocadurdu K, Ismailoglu E, Guneyli S. Deep learning model-assisted detection of kidney stones on computed tomography. Int Braz J Urol. 2022;48:830-9.

[2] ²
Danilovic A, Cavalanti A, Rocha BA, Traxer O, Torricelli FCM, Marchini GS, et al. Assessment of Residual Stone Fragments After Retrograde Intrarenal Surgery. J Endourol. 2018;32:1108-13.

[3] ³
Danilovic A, Rocha BA, Marchini GS, Traxer O, Batagello C, Vicentini FC, et al. Computed tomography window affects kidney stones measurements. Int Braz J Urol. 2019;45:948-55.

[4] ⁴
Elton DC, Turkbey EB, Pickhardt PJ, Summers RM. A deep learning system for automated kidney stone detection and volumetric segmentation on noncontrast CT scans. Med Phys. 2022;49:2545-54.

[5] ⁵
Cui Y, Sun Z, Ma S, Liu W, Wang X, Zhang X, et al. Automatic Detection and Scoring of Kidney Stones on Noncontrast CT Images Using S.T.O.N.E. Nephrolithometry: Combined Deep Learning and Thresholding Methods. Mol Imaging Biol. 2021;23:436-45.

Brasil