Skin cancer detection by deep learning and sound analysis algorithms: A prospective clinical study of an elementary dermoscope

A Dascalu, E O David, A Dascalu, E O David

Abstract

Background: Skin cancer (SC), especially melanoma, is a growing public health burden. Experimental studies have indicated a potential diagnostic role for deep learning (DL) algorithms in identifying SC at varying sensitivities. Previously, it was demonstrated that diagnostics by dermoscopy are improved by applying an additional sonification (data to sound waves conversion) layer on DL algorithms. The aim of the study was to determine the impact of image quality on accuracy of diagnosis by sonification employing a rudimentary skin magnifier with polarized light (SMP).

Methods: Dermoscopy images acquired by SMP were processed by a first deep learning algorithm and sonified. Audio output was further analyzed by a different secondary DL. Study criteria outcomes of SMP were specificity and sensitivity, which were further processed by a F2-score, i.e. applying a twice extra weight to sensitivity over positive predictive values.

Findings: Patients (n = 73) fulfilling inclusion criteria were referred to biopsy. SMP analysis metrics resulted in a receiver operator characteristic curve AUC's of 0.814 (95% CI, 0.798-0.831). SMP achieved a F2-score sensitivity of 91.7%, specificity of 41.8% and positive predictive value of 57.3%. Diagnosing the same set of patients' lesions by an advanced dermoscope resulted in a F2-score sensitivity of 89.5%, specificity of 57.8% and a positive predictive value of 59.9% (P=NS).

Interpretation: DL processing of dermoscopic images followed by sonification results in an accurate diagnostic output for SMP, implying that the quality of the dermoscope is not the major factor influencing DL diagnosis of skin cancer. Present system might assist all healthcare providers as a feasible computer-assisted detection system. FUND: Bostel Technologies. Trial Registration clinicaltrials.gov Identifier: NCT03362138.

Keywords: Artificial intelligence; Deep learning; Dermoscopy; Melanoma; Skin cancer; Sonification; Telemedicine.

Copyright © 2019 The Authors. Published by Elsevier B.V. All rights reserved.

Figures

Fig. 1
Fig. 1
Flowchart of image processing. A dermoscopy image is acquired by a smartphone and conveyed to cloud computing. A deep learning classifier predicts primary findings which are further processed by sonification. Final diagnosis is conferred to user as a malign or benign lesion diagnosis, i.e. excise or not indication.
Fig. 2
Fig. 2
An image from an elementary dermoscope (2a) was compared to the advanced dermoscope version (2b). A Fast Fourier Transformation was applied on images (2c, b), white areas were further thresholded (Fig. 2e, f) and area fraction was measured. The advanced dermoscope possesses higher spatial frequencies (p > .001) and therefore inputs a more detailed image.
Fig. 3
Fig. 3
Comparison of benign nevus (left), dysplastic nevus (middle) and melanoma (right). A. Skin magnifier with polarized light acquires hazy out of focus images, without a clear-cut pattern recognition of fine details required for diagnosis. B. Same lesions images acquired by an advanced dermoscope exhibit clear fine features and edges enabling diagnosis. C. Raw audio waves of a benign nevus, dysplastic nevus and melanoma, imperceptible to the human eye.
Fig. 4
Fig. 4
ROC curve of raw sonified audio files derived from an elementary dermoscope.

References

    1. Ferlay J., Colombet M., Soerjomataram I. Estimating the global cancer incidence and mortality in 2018: GLOBOCAN sources and methods. Int J Cancer. 2019 Apr 15;144(8):1941–1953.
    1. Kandel M., Allayous C., Dalle S. Update of survival and cost of metastatic melanoma with new drugs: Estimations from the MelBase cohort. Eur J Cancer. 2018 Dec;105:33–40. Internet.
    1. Dinnes J., Deeks J.J., Chuchu N. Dermoscopy, with and without visual inspection, for diagnosing melanoma in adults. Cochrane Database Syst Rev. 2018 Dec 4;12:CD011902.
    1. Tschandl P., Hofmann L., Fink C., Kittler H., Haenssle H.A. Melanomas vs. nevi in high-risk patients under long-term monitoring with digital dermatoscopy: do melanomas and nevi already differ at baseline? J Eur Acad Dermatology Venereol. 2017 Jun 1;31(6):972–977.
    1. Annessi G., Bono R., Sampogna F., Faraggiana T., Abeni D. Sensitivity, specificity, and diagnostic accuracy of three dermoscopic algorithmic methods in the diagnosis of doubtful melanocytic lesions: the importance of light brown structureless areas in differentiating atypical melanocytic nevi from thin melanomas. J Am Acad Dermatol. 2007 May 1;56(5):759–767.
    1. Koelink C.J.L., Vermeulen K.M., Kollen B.J. Diagnostic accuracy and cost-effectiveness of dermoscopy in primary care: a cluster randomized clinical trial. J Eur Acad Dermatology Venereol. 2014 Nov 1;28(11):1442–1449.
    1. Matsumoto M., Secrest A., Anderson A. Estimating the cost of skin cancer detection by dermatology providers in a large health care system. J Am Acad Dermatol. 2018 Apr 1;78(4):701–709.e1.
    1. Argenziano G., Cerroni L., Zalaudek I. Accuracy in melanoma detection: a 10-year multicenter survey. J Am Acad Dermatol. 2012 Jul;67(1):54–59.
    1. Celebi M.E., Codella N., Halpern A. Dermoscopy image analysis: overview and future directions. IEEE J Biomed Heal Informatics. 2019 Mar;23(2):474–478.
    1. Codella N., Cai J., Abedini M., Garnavi R., Halpern A., Smith J.R. Springer; Cham: 2015. Deep learning, sparse coding, and SVM for melanoma recognition in dermoscopy images; pp. 118–126.
    1. Navarrete-Dechent C., Dusza S.W., Liopyris K., Marghoob A.A., Halpern A.C., Marchetti M.A. Automated dermatological diagnosis: hype or reality? J Invest Dermatol. 2018 Oct 1;138(10):2277–2279.
    1. Haenssle H.A., Fink C., Schneiderbauer R. Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists. Ann Oncol. 2018;29(8):1836–1842.
    1. Walker B.N., Rehg J.M., Kalra A. Dermoscopy diagnosis of cancerous lesions utilizing dual deep learning algorithms via visual and audio (sonification) outputs: laboratory and prospective observational studies. EBioMedicine. Feb 2019;40:176–183.
    1. Codella N.C.F., Gutman D., Celebi M.E. Skin lesion analysis toward melanoma detection: a challenge at the 2017 international symposium on biomedical Imaging (ISBI), hosted by the international skin imaging collaboration (ISIC) 2017 Oct 13.
    1. Walker B.N., Nees M.A. Theory of sonification. In: Hermann T., Hunt A., Neuhoff J., editors. The sonification handbook. Logos Publishing House; Berlin, Germany: 2011. pp. 9–39. [ISBN 978-3-8325-2819-5]
    1. Celebi M.E., Kingravi H.A., Vela P.A. A comparative study of efficient initialization methods for the k-means clustering algorithm. Expert Syst Appl. 2013 Jan;40(1):200–210.
    1. Berman D., Golomb J.D., Walther D.B. Scene content is predominantly conveyed by high spatial frequencies in scene-selective visual cortex. PLoS One. 2017 Dec 22;12(12)
    1. Malvehy J., Hauschild A., Curiel-Lewandrowski C. Clinical performance of the Nevisense system in cutaneous melanoma detection: an international, multicentre, prospective and blinded clinical trial on efficacy and safety. Br J Dermatol. 2014 Nov;171(5):1099–1107.
    1. Elmore J.G., Barnhill R.L., Elder D.E. Pathologists' diagnosis of invasive melanoma and melanocytic proliferations: observer accuracy and reproducibility study. BMJ. 2017;357:j2813. Jun 28.
    1. Fleming N.H., Egbert B.M., Kim J., Swetter S.M. Reexamining the threshold for reexcision of histologically transected dysplastic nevi. JAMA Dermatol. 2016;152(12):1327–1334.
    1. Benvenuto-Andrade C., Dusza S.W., Agero A.L.C. Differences between polarized light dermoscopy and immersion contact dermoscopy for the evaluation of skin lesions. Arch Dermatol. 2007 Mar 1;143(3):329–338.
    1. Geirhos R., Janssen D.H.J., Schütt H.H., Rauber J., Bethge M., Wichmann F.A. Comparing deep neural networks against humans: object recognition when the signal gets weaker*. arXiv. 11 Dec 2018;1706 06969v2 []

Source: PubMed

Sponsor e collaboratori

Condizioni mediche

Interventi farmacologici

3
Sottoscrivi