EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Artificial Intelligence–Assisted ACR TI-RADS and Bethesda-like Risk Prediction for Indeterminate Thyroid Nodules: A Retrospective Proof-of-Concept Evaluation

Künstlich-Intelligenz-gestützte ACR TI-RADS- und Bethesda-ähnliche Risikoprädiktion bei indeterminierten Schilddrüsenknoten: retrospektive Proof-of-Concept-Bewertung

Duzgun Yildirim iD, Deniz Esin Tekcan Sanli, Meltem Kursun iD, Burcu Bektas Gunes, Aytug Altundag, Leonardo Obinna Iheme iD, Mustafa Hurmuzli iD, Ahmet Necati Sanli iD, Eren Yılmaz iD, Filiz Celebi, Ayten Paşazade iD, Bülent Çolakoğlu und 1 weitere

Diagnostics · 02.11.2026 · 16 (21) · 3616

DOI: 10.3390/diagnostics16213616 ↗

Auf einen Blick

In 52 operativ entfernten Bethesda-III/IV-Knoten zeigte das KI-System TRAICK für die ACR-TI-RADS-Ausgabe eine mit einem Radiologen vergleichbare Trennschärfe und höhere Spezifität. Die Bethesda-ähnliche Ausgabe wirkte ebenfalls vielversprechend.

Fragestellung

Wie gut sagt das KI-System TRAICK bei zytologisch indeterminierten Schilddrüsenknoten ACR TI-RADS- und Bethesda-ähnliche Risikokategorien voraus?

Methode

Retrospektive Proof-of-Concept-Studie mit 52 operativ resezierten Knoten; Referenzstandard war die Histopathologie des Indexknotens. Vergleich der ACR-TI-RADS-Ausgabe mit 2 Radiologen mittels McNemar- und DeLong-Test.

Ergebnisse

52 Knoten, davon 23 maligne (44,2 %). ACR TI-RADS: Accuracy 71,2 % (95 %-KI 57,7–81,7), AUC 0,762 (0,640–0,884). Vergleich mit Radiologe 1: AUC 0,762 vs. 0,777, Spezifität 65,5 % vs. 37,9 %, Sensitivität 78,3 % vs. 91,3 %. Bethesda-like: AUC 0,822 (0,707–0,938), Sensitivität 60,9 %, Spezifität 86,2 %.

Einschränkungen

Kleine, selektierte chirurgische Kohorte; unabhängige prospektive Validierung erforderlich.

Schlussfolgerung der Autor:innen

In dieser kleinen, selektierten chirurgischen Kohorte zeigte die ACR-TI-RADS-Ausgabe eine Trennschärfe ohne signifikanten Unterschied zum besseren von zwei erfahrenen Radiologen bei höherer Spezifität; die Bethesda-ähnliche Ausgabe war vielversprechend.

Kernaussagen

  • 52 operativ resezierte Bethesda-III/IV-Knoten
  • ACR TI-RADS: Accuracy 71,2 %; AUC 0,762
  • Bethesda-like: AUC 0,822; Spezifität 86,2 %

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Fachartikel in der Zeitschrift Diagnostics, online erschienen am 02.11.2026. Verfasst von Duzgun Yildirim und 12 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiology, Acıbadem Taksim Hospital, Istanbul 34373, Turkey. Die Arbeit nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background/Objectives: Cytologically indeterminate thyroid nodules (Bethesda III–IV) remain a diagnostic challenge. We evaluated the TRAICK (version 1.9.4; TRAICK, Istanbul, Türkiye) artificial intelligence (AI) system, which generates an American College of Radiology Thyroid Imaging Reporting and Data System (ACR TI-RADS) category and an imaging-derived Bethesda-like risk category from gray-scale ultrasound images.

Methods

This retrospective proof-of-concept study included 52 surgically resected nodules with Bethesda III (n = 29) or IV (n = 23) cytology, with histopathology of the index nodule as the reference standard. The two TRAICK outputs were analyzed separately, and the ACR TI-RADS output was compared with two radiologists using the McNemar and DeLong tests.

Results

Twenty-three nodules (44.2%) were malignant on index-nodule histopathology. The TRAICK ACR TI-RADS output had an accuracy of 71.2% (95% CI 57.7–81.7) and an area under the ROC curve (AUC) of 0.762 (0.640–0.884). Its AUC did not differ significantly from that of Radiologist 1 (0.762 vs. 0.777, p = 0.783); specificity was higher (65.5% vs. 37.9%, p = 0.021), whereas sensitivity was lower without a statistically significant difference (78.3% vs. 91.3%, p = 0.250). Compared with Radiologist 2, the TRAICK ACR TI-RADS output had higher accuracy (71.2% vs. 51.9%, p = 0.013), specificity (65.5% vs. 24.1%, p < 0.001), and AUC (0.762 vs. 0.629, p = 0.042). The Bethesda-like output had an AUC of 0.822 (0.707–0.938), with a sensitivity of 60.9% and specificity of 86.2% at the V–VI threshold. In sensitivity analyses, the AUC of the ACR TI-RADS output did not differ significantly from that of Radiologist 1, whereas its AUC advantage over Radiologist 2 was not statistically significant when low-risk neoplasms were classified as malignant.

Conclusions

In this small, selected surgical cohort, the TRAICK ACR TI-RADS output showed discrimination that did not differ significantly from that of the more accurate of two experienced radiologists, with higher specificity, and its Bethesda-like output showed promising discrimination. These findings support further study of TRAICK as a complementary aid for preoperative decision-making; independent prospective validation is needed.

Schlagwörter

Stichwörter der Autor:innen

ACR TI-RADSBethesda System for Reporting Thyroid Cytopathologyartificial intelligencediagnostic accuracyindeterminate cytologythyroid nodulesultrasonography

Autor:innen und Institutionen

  • Department of Radiology, Acıbadem Taksim Hospital, Istanbul 34373, Turkey.
  • Department of Radiology, Faculty of Medicine, Gaziantep University, Gaziantep 27310, Turkey.
  • Department of Radiology, Bucak State Hospital, Burdur 15300, Turkey.
  • Department of Computer Engineering, Gedik University, Istanbul 34876, Turkey.
  • Department of Otorhinolaryngology, School of Medicine, Biruni University, Istanbul 34295, Turkey.
  • EPAM Systems Inc., Istanbul 34475, Turkey.

ORCID-Profile: Duzgun Yildirim, Meltem Kursun, Leonardo Obinna Iheme, Mustafa Hurmuzli, Ahmet Necati Sanli, Eren Yılmaz, Ayten Paşazade, Enes Yerdes

Zitieren

APA

Yildirim, D., Tekcan Sanli, D. E., Kursun, M., Gunes, B. B., Altundag, A., Iheme, L. O., Hurmuzli, M., Sanli, A. N., Yılmaz, E., Celebi, F., Paşazade, A., Çolakoğlu, B., & Yerdes, E. (2026). Artificial Intelligence–Assisted ACR TI-RADS and Bethesda-like Risk Prediction for Indeterminate Thyroid Nodules: A Retrospective Proof-of-Concept Evaluation. Diagnostics, 16(21), 3616. https://doi.org/10.3390/diagnostics16213616

Vancouver

Yildirim D, Tekcan Sanli DE, Kursun M, Gunes BB, Altundag A, Iheme LO, et al. Artificial Intelligence–Assisted ACR TI-RADS and Bethesda-like Risk Prediction for Indeterminate Thyroid Nodules: A Retrospective Proof-of-Concept Evaluation. Diagnostics. 2026;16(21):3616. doi:10.3390/diagnostics16213616

BibTeX
@article{Yildirim2026,
  author = {Yildirim, Duzgun and Tekcan Sanli, Deniz Esin and Kursun, Meltem and Gunes, Burcu Bektas and Altundag, Aytug and Iheme, Leonardo Obinna and Hurmuzli, Mustafa and Sanli, Ahmet Necati and Yılmaz, Eren and Celebi, Filiz and Paşazade, Ayten and Çolakoğlu, Bülent and Yerdes, Enes},
  title = {Artificial Intelligence–Assisted ACR TI-RADS and Bethesda-like Risk Prediction for Indeterminate Thyroid Nodules: A Retrospective Proof-of-Concept Evaluation},
  journal = {Diagnostics},
  year = {2026},
  volume = {16},
  number = {21},
  pages = {3616},
  doi = {10.3390/diagnostics16213616}
}
Kostenlos starten →