Probe-based confocal laser endomicroscopy in detecting malignant lesions of vocal folds

Abstract

Probe-based confocal laser endomicroscopy (CLE) is an innovative technique for real-time, non-invasive analysis of the surface epithelium. While being successfully used for diagnosis by experts, this method has not yet been established in clinical routine, partly due to the lack of standards and criteria for classifying various lesions. Our aim was to determine the diagnostic value and inter-rater reliability of CLE in detecting malignant lesions of the vocal cords. 58 video sequences were extracted from the probe-based CLE (GastroFlex probe with a Cellvizio ® laser system) examinations of 3 patients with squamous cell carcinomas and 4 patients with benign alterations of the vocal folds. Two ENT surgeons, who were blinded to the histological result, were asked to identify the sequences representing a carcinoma. We showed an accuracy, sensitivity, specificity, PPV and NPV of 91.38-96.55%, 100%, 87.8-95.2%, 77.27-89.47% and 100%, respectively, with an inter-rater reliability of k = 0.89 ("almost perfect agreement"). Probe-based CLE is a promising method for diagnosis and assessment of vocal fold lesions in vivo. Our results suggest that, with adequate training, the diagnostic value of this technique can be improved and potentially provide important information during oncological surgery. KEY WORDS: Confocal laser endomicroscopy • Laryngeal cancer • Non-invasive histological imaging • Vocal folds RIASSUNTO L'endomicroscopia confocale laser (CLE) tramite sonda è una tecnica innovativa per l'analisi in tempo reale non invasiva dell'epitelio di superficie. Mentre il suo uso diagnostico ha avuto successo tra gli specialisti, questo metodo non si è ancora affermato nella routine clinica, in parte anche per la mancanza di regole e criteri per la classificazione delle varie lesioni. Il nostro obiettivo era quello di determinare il valore diagnostico e l'affidabilità tra valutatori CLE nel rilevamento delle lesioni maligne delle corde vocali. Sono state estratte 58 sequenze video dagli esami CLE tramite sonda (sonda GastroFlex con un sistema laser Cellvizio ®) di 3 pazienti con carcinoma squamocellulare e di 4 pazienti con alterazioni benigne delle corde vocali. A due chirurghi ORL, in cieco per l'esame istologico, è stato chiesto di identificare le sequenze riguardanti un carcinoma.

Konfokale Laser-Endomikroskopie des Kopf-Hals-Plattenepithelkarzinoms: eine systematische Übersicht

“…Darüber hinaus wurde gezeigt, dass die Genauigkeit der Interpretation von CLE-Bildern in hohem Maße von der Erfahrung des klinischen Untersuchers abhängt und eine signifikante Lernkurve für die Auswertung der CLE-Bilder besteht [5,12,13]. Die Interpretation von CLE-Bildern durch Untersucher mit wenig oder kei-ner Erfahrung zeigte für einen Satz von 31 repräsentativen Bildern der Stimmlippen unterschiedliche Genauigkeitsgrade (58,1-77,4 %) und inkonsistente Interrater-Ergebnisse (Fleiss' Kappa k = 0,25) [24]. Nach Absolvierung und Zertifizierung eines Online-Trainingsprogramms (www.cellvizio.net/academy) konnte mit einer Genauigkeit von 91,38-96,55 % und einer Interrater-Reliabilität von k = 0,89 eine Verbesserung der untersucherbasierten Befunderhebung erzielt werden [33].…”

“…Die Interpretation von CLE-Bildern durch Untersucher mit wenig oder kei-ner Erfahrung zeigte für einen Satz von 31 repräsentativen Bildern der Stimmlippen unterschiedliche Genauigkeitsgrade (58,1-77,4 %) und inkonsistente Interrater-Ergebnisse (Fleiss' Kappa k = 0,25) [24]. Nach Absolvierung und Zertifizierung eines Online-Trainingsprogramms (www.cellvizio.net/academy) konnte mit einer Genauigkeit von 91,38-96,55 % und einer Interrater-Reliabilität von k = 0,89 eine Verbesserung der untersucherbasierten Befunderhebung erzielt werden [33]. Für menschliche Begutachtungen sind Videosequenzen aussagekräftiger als Einzelbilder [23].…”

Transferability of Deep Learning Algorithms for Malignancy Detection in Confocal Laser Endomicroscopy Images from Different Anatomical Locations of the Upper Gastrointestinal Tract

“…On a similar, different data set with 7 patients, they found accuracies by Ear, Nose and Throat (ENT) specialists to be between 58.1% and 87.1%, where the non-CLE experienced doctors had a mean accuracy of 67.7% and those with profound CLE experience one of 82.2%. 8 For this work, 47 image sequences from five patients were recorded. All patients underwent biopsy subsequent to CLE and were diagnosed with SCC of the vocal folds using the gold-standard of histo-pathology.…”

“…CLE is being successfully used in diagnostics of the intestine in clinical practice. 6 It was reported, that probe-based CLE (pCLE) can also successfully be used for assessment of malignancy of SCC of the vocal folds 8,10 as well as in epithelium of the oral cavity. 7 However, it is also known that interpretation of CLE images requires significant training.…”

Patch-based Carcinoma Detection on Confocal Laser Endomicroscopy Images - A Cross-site Robustness Assessment

“…Goncalves et al selected 31 representative images 2 Study approved by the ethics committee of the University of Erlangen-Nürnberg; reference number: 60 14 B out of a different vocal cord data set of 7 patients and found rating accuracies by ENT specialists to be between 58.1% and 87.1%, where the non-CLE experienced doctors had a mean accuracy of 67.7% and those with profound experience of CLE one of 82.2% (Goncalves et al, 2017). Also for this data set, the differences in recognition performance are well explained by the learning effect as described by Neumann (Neumann et al, 2011).…”