PANORAMİK RADYOGRAFLARDA APİKAL PATOLOJİ TEŞHİSİ İÇİN YAPAY ZEKA YETERLİ GÜVENİLİRLİK DÜZEYİNE ULAŞTI MI: FARKLI EŞİK DEĞERLERDE KARŞILAŞTIRMALI ÇALIŞMA

Abstract

Amaç: Panoramik radyografiler üzerinden tespit edilen apikal patolojilerin yapay zekâ temelli ticari bir yazılım ile doğruluğunun tespit edilmesidir. Gereç ve yöntem: En az bir tane apikal patolojinin olduğu 100 panoramik radyograf seçilmiş ve Denti.AI ticari yazılımına yüklenmiştir. Sistemin %30, %60, %90 eşik değerlerinde apikal patolojiyi tespit etmesi sağlanmıştır. Panoramik radyografiler Ağız, Diş ve Çene Radyolojisi alanında uzmanlığını almış iki hekim ve Ağız, Diş ve Çene Radyolojisi uzmanlık eğitimini sürdüren iki araştırma görevlisi tarafından seçilmiştir. Verilerin analizi SPSS 22.0 Paket Veri Programı ile yapılmıştır. Bulgular: Çalışmanın sonunda %30, %60, %90 eşik değerleri için sensivite değerleri sırasıyla %62.6, %38.1 ve %6.8, spesifite değerleri %0, %100 ve %100, tanısal doğruluk değerleri 61.3, 39.3 ve 8.6 olarak bulunmuştur. PPV değerleri %96.8, 100 ve 100, NPV değerleri 0, 3.2 ve 2.1, AUC değerleri ise 0.313, 0.69 ve 0.534 olarak hesaplanmıştır. İstatistiki testlerde p<0.05 değeri anlamlı olarak kabul edilmiştir. Sonuç: Yapay zekâ programı, özellikle eşik değer düşürüldüğü zaman uzman hekimlere yakın sonuçlar vermektedir. Bu şekilde hazırlanan yapay zekâ tabanlı ticari yazılımlarda örnek sayılarının artışı ve bunların geriye dönük tespit edilmesinin doğruluğu artıracağını, bu tip yazılımların klinik tanılarda daha çok yer alacağını ve yoğun kliniklerde başvurulabilecek bir destek sistemi olabileceğini düşünmekteyiz.

Keywords

• apikal patoloji
• derin öğrenme
• panoramik radyografi
• yapay zeka

RELIABILITY ASSESSMENT OF ARTIFICIAL INTELLIGENCE FOR THE DIAGNOSIS OF APICAL PATHOLOGY IN PANORAMIC RADIOGRAPHS: A COMPARATIVE STUDY AT DIFFERENT THRESHOLD VALUES

Abstract

Background: To determine the accuracy of apical pathologies detected on panoramic radiographs with an artificial intelligence-based commercial software. Methods: 100 panoramic radiographs with at least one apical pathology were selected and uploaded to Denti.AI commercial software. The system was enabled to detect apical pathology at 30%, 60%, and 90% threshold values. Panoramic radiographs were selected by two Oral and Maxillofacial Radiologists and two research assistants who are continuing their residency training in Oral and Maxillofacial Radiology. Data analysis was done with SPSS 22.0 Packet Data Program. Results: At the end of the study, sensitivity values were 62.6%, 38.1% and 6.8%, specificity values were 0%, 100% and 100%, and diagnostic accuracy values were 61.3, 39.3 and 8.6 for the threshold values of 30%, 60%, 90%, respectively. PPV values were calculated as 96.8, 100 and 100%, NPV values as 0, 3.2 and 2.1, AUC values as 0.313, 0.69 and 0.534. In statistical tests, p <0.05 value was accepted as significant. Conclusion: The artificial intelligence program gives results close to specialist dentists, especially when the threshold value is lowered. We think that the increase in the number of samples and retrospective detection of artificial intelligence-based commercial software prepared in this way will increase the accuracy, this type of software will be more involved in clinical diagnoses and can be a support system that can be used in intensive clinics.

Keywords

• apical pathology
• artificial intelligence
• deep learning
• panoramic radiography

References

1. White SC, Pharoah MJ. Oral Radyoloji: İlkeler Ve Yorumlama. Çevirenler: Nursel Akkaya, Zuhal Çokaktaş Yandımata. 7. Baskı Ankara: Palme Yayınevi; 2018. s. 315.
2. Koivisto T, Bowles WR, Rohrer M. Frequency and distribution of radiolucent jaw lesions: a retrospective analysis of 9,723 cases. Journal of Endodontics 2012;38(6), 729-32.
3. Kakehashi S, Stanley HR, Fitzgerald RJ. The effects of surgical exposures of dental pulps in germ-free and conventional laboratory rats. Oral Surg Oral Med Oral Pathol 1965;20:340–9.
4. Möller ÅJ, Fabricius L, Dahlen G, Öhman AE, Heyden GUY. Influence on periapical tissues of indigenous oral bacteria and necrotic pulp tissue in monkeys. European Journal of Oral Sciences 1981;89(6), 475-84.
5. Azuma MM, Samuel RO, Gomes‐Filho JE, Dezan‐Junior E, Cintra LTA. The role of IL‐6 on apical periodontitis: a systematic review. International endodontic journal 2014;47(7), 615-21.
6. Braz-Silva PH, Bergamini ML, Mardegan AP, De Rosa CS, Hasseus B, Jonasson P. Inflammatory profile of chronic apical periodontitis: a literature review. Acta Odontologica Scandinavica 2019;77(3), 173-80.
7. Arslan ZB, Demir H, Berker Yıldız D, Yaşar F. (2020). Diagnostic accuracy of panoramic radiography and ultrasonography in detecting periapical lesions using periapical radiography as a gold standard. Dentomaxillofacial Radiology 2020;49, 20190290.
8. Orhan K, Bayrakdar IS, Ezhov M, Kravtsov A, Özyürek T. Evaluation of artificial intelligence for detecting periapical pathosis on cone‐beam computed tomography scans. International Endodontic Journal 2020;53(5), 680-89.

9. Ekert T, Krois J, Meinhold L, Elhennawy K, Emara R, Golla T, et al. Deep learning for the radiographic detection of apical lesions. Journal of endodontics 2019;45(7), 917-22.
10. Wang S, Summers RM. Machine learning and radiology. Medical image analysis 2012;16(5), 933-51.
11. Sklan JE, Plassard AJ, Fabbri D, Landman BA. Toward content-based image retrieval with deep convolutional neural networks. In Medical Imaging 2015: Biomedical Applications in Molecular, Structural, and Functional Imaging (Vol. 9417, p. 94172C). International Society for Optics and Photonics 2015.
12. Hung K, Montalvao C, Tanaka R, Kawai T, Bornstein MM. The use and performance of artificial intelligence applications in dental and maxillofacial radiology: A systematic review. Dentomaxillofacial Radiology 2020;49(1), 20190107.
13. Christ PF, Elshaer MEA, Ettlinger F, Tatavarty S, Bickel M, Bilic P, et al. Automatic liver and lesion segmentation in CT using cascaded fully convolutional neural networks and 3D conditional random fields. In International Conference on Medical Image Computing and Computer-Assisted Intervention . Springer: Cham; 2016. p. 415-23.
14. Lakhani P, Sundaram B. Deep learning at chest radiography: automated classification of pulmonary tuberculosis by using convolutional neural networks. Radiology 2017;284:574–82.
15. Yasaka K, Akai H, Abe O, Kiryu S. Deep learning with convolutional neural network for differentiation of liver masses at dynamic contrast-enhanced CT: a preliminary study. Radiology 2018;286:887–96.
16. Kim KH, Choi SH, Park SH. Improving arterial spin labeling by using deep learning. Radiology 2018;287:658–66.
17. Liu F, Jang H, Kijowski R, Bradshaw T, McMillan AB. Deep learning MR imaging-based attenuation correction for PET/MR imaging. Radiology 2018;286:676–84.
18. Lee JH, Kim DH, Jeon SN, Choi SH. Detection and diagnosis of dental caries using a deep learning-based convolutional neural network algorithm. Journal of dentistry 2018;77, 106-11.
19. Murata M, Ariji Y, Ohashi Y, Kawai T, Fukuda M, Funakoshi T, et al. Deep-learning classification using convolutional neural network for evaluation of maxillary sinusitis on panoramic radiography. Oral radiology 2019;35(3), 301-7.
20. Gulshan V, Peng L, Coram M, Stumpe, MC, Wu D, Narayanaswamy A, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. Jama 2016;316(22), 2402-10.
21. Esteva A, Kuprel B, Novoa RA, Ko J, Swetter SM, Blau HM, et al. Dermatologist-level classification of skin cancer with deep neural networks. Natüre 2017;542(7639), 115-8.
22. Lopez-Garnier S, Sheen P, Zimic M. Automatic diagnostics of tuberculosis using convolutional neural networks analysis of MODS digital images. PloS one 2019;14(2), e0212094.
23. Endres MG, Hillen F, Salloumis M, Sedaghat AR, Niehues SM, Quatela O, et al. Development of a deep learning algorithm for periapical disease detection in dental radiographs. Diagnostics 2020;10(6), 430.
24. Johari M, Esmaeili F, Andalib A, Garjani S, Saberkari H. Detection of vertical root fractures in intact and endodontically treated premolar teeth by designing a probabilistic neural network: an ex vivo study. Dentomaxillofacial Radiology 2017;46(2), 20160107.
25. Krois J, Ekert T, Meinhold L, Golla T, Kharbot B, Wittemeier A, et al. Deep learning for the radiographic detection of periodontal bone loss. Scientific reports 2019;9(1), 1-6.
26. Tuzoff DV, Tuzova LN, Bornstein MM, Krasnov AS, Kharchenko MA, Nikolenko SI, et al. Tooth detection and numbering in panoramic radiographs using convolutional neural networks. Dentomaxillofacial Radiology 2019;48(4), 20180051.
27. Amasya H, Yildirim D, Aydogan T, Kemaloglu N, Orhan K. Cervical vertebral maturation assessment on lateral cephalometric radiographs using artificial intelligence: comparison of machine learning classifier models. Dentomaxillofacial Radiology 2020;49, 20190441.
28. Estrela C, Bueno MR, Leles CR, Azevedo B, Azevedo JR. Accuracy of cone beam computed tomography and panoramic and periapical radiography for detection of apical periodontitis. Journal of endodontics 2008;34(3), 273-79.