Flexible transnasal endoscopy with white light or narrow band imaging for the diagnosis of laryngeal malignancy: diagnostic value, observer variability and influence of previous laryngeal surgery

Laryngeal Cancer Screening During Flexible Video Laryngoscopy Using Large Computer Vision Models

OBJECTIVE

Develop an artificial intelligence assisted computer vision model to screen for laryngeal cancer during flexible laryngoscopy.

METHODS

Using laryngeal images and flexible laryngoscopy video recordings, we developed computer vision models to classify video frames for usability and cancer screening. A separate model segments any identified lesions on the frames. We used these computer vision models to construct a video stream annotation system. This system classifies findings from flexible laryngoscopy as "potentially malignant" or "probably benign" and segments any detected lesions. Additionally, the model provides a confidence level for each classification.

RESULTS

The overall accuracy of the flexible laryngoscopy cancer screening model was 92%. For cancer screening, it achieved a sensitivity of 97.7% and a specificity of 76.9%. The segmentation model attained an average precision at a 0.50 intersection-over-union of 0.595. The confidence level for positive screening results can assist clinicians in counseling patients regarding the findings.

CONCLUSION

Our model is highly sensitive and adequately specific for laryngeal cancer screening. Segmentation helps endoscopists identify and describe potential lesions. Further optimization is required to enable the model's deployment in clinical settings for real-time annotation during flexible laryngoscopy.

The influence of clinical experience on reliable evaluation of pharyngeal and laryngeal lesions: comparison of high-definition laryngoscopy using narrow band imaging with fibre-optic laryngoscopy

BACKGROUND

Fibre-optic laryngoscopy is still widely used in daily clinical practice; however, high-definition laryngoscopy using narrow band imaging could be more reliable in characterising pharyngeal and laryngeal lesions.

METHODS

Endoscopic videos were assessed in a tertiary referral hospital by 12 observers with different levels of clinical experience. Thirty pairs of high-definition laryngoscopy with narrow band imaging and fibre-optic laryngoscopy videos were judged twice, with an interval of two to four weeks, in a random order. Inter- and intra-observer reliability, sensitivity and specificity were calculated in terms of detecting a malignant lesion and a specific histological entity, for beginners, trained observers and experts.

RESULTS

Using high-definition laryngoscopy with narrow band imaging, inter-observer reliability for detecting malignant lesions increased from moderate to substantial in trained observers and experts (high-definition laryngoscopy with narrow band imaging κ = 0.66 and κ = 0.77 vs fibre-optic laryngoscopy κ = 0.51 and κ = 0.56, for trained observers and experts respectively) and sensitivity increased by 16 per cent.

CONCLUSION

Inter-observer reliability increased with the level of clinical experience, especially when using high-definition laryngoscopy with narrow band imaging.

The Multimodal Diagnostic Approach Necessary in Detecting Elusive Submucosal Laryngeal Cancer

Submucosal laryngeal lesions have proven themselves to be a diagnostic challenge in the field of medicine, often presenting inconsistently between endoscopic visualization, various imaging modalities, and biopsy. The conflicting clinical picture can lead to a delay in definitive diagnosis and treatment. A variety of laryngeal imaging modalities exist that give a unique perspective of the tumor being evaluated and can be used in combination to clarify discrepancies in presentation.  This report describes the clinical course of an undiagnosed laryngeal squamous cell carcinoma (SCC) presenting with persistent dysphonia, dysphagia, and unilateral vocal fold immobility. A negative head and neck computerized tomography (CT) scan reduced the concern for cancer, so symptomatic treatment with vocal fold augmentation was performed. Augmentation curiously worsened the dysphonia and also may have delayed the process of definitive diagnosis. Upon presenting to the laryngology clinic, stroboscopy demonstrated no vibration of the affected vocal fold. Submucosal vascular irregularity was noted with narrow band imaging with a very subtle keratotic mucosal change raising suspicion for underlying malignancy. Despite two CT scans that failed to visualize the lesion initially, a biopsy revealed keratinizing SCC, which was subsequently staged as T3N0M0. The patient elected to receive radiation therapy alone given his medical comorbidities. This case showcases the elusive ability submucosal laryngeal cancers have in diagnostic workups. Heavy reliance on any single diagnostic modality may be misleading, resulting in delayed diagnosis and treatment. An early, thorough, and multimodal approach that analyzes the cumulative results of a variety of diagnostic tools is essential in identifying and treating these elusive cancers in a timely manner.

Artificial Intelligence in Laryngeal Endoscopy: Systematic Review and Meta-Analysis

Background: Early diagnosis of laryngeal lesions is necessary to begin treatment of patients as soon as possible to preserve optimal organ functions. Imaging examinations are often aided by artificial intelligence (AI) to improve quality and facilitate appropriate diagnosis. The aim of this study is to investigate diagnostic utility of AI in laryngeal endoscopy. Methods: Five databases were searched for studies implementing artificial intelligence (AI) enhanced models assessing images of laryngeal lesions taken during laryngeal endoscopy. Outcomes were analyzed in terms of accuracy, sensitivity, and specificity. Results: All 11 studies included presented an overall low risk of bias. The overall accuracy of AI models was very high (from 0.806 to 0.997). The accuracy was significantly higher in studies using a larger database. The pooled sensitivity and specificity for identification of healthy laryngeal tissue were 0.91 and 0.97, respectively. The same values for differentiation between benign and malignant lesions were 0.91 and 0.94, respectively. The comparison of the effectiveness of AI models assessing narrow band imaging and white light endoscopy images revealed no statistically significant differences (p = 0.409 and 0.914). Conclusion: In assessing images of laryngeal lesions, AI demonstrates extraordinarily high accuracy, sensitivity, and specificity.

LPCANet: Classification of Laryngeal Cancer Histopathological Images Using a CNN with Position Attention and Channel Attention Mechanisms

Laryngeal cancer is one of the most common malignant tumors in otolaryngology, and histopathological image analysis is the gold standard for the diagnosis of laryngeal cancer. However, pathologists have high subjectivity in their diagnoses, which makes it easy to miss diagnoses and misdiagnose. In addition, according to a literature search, there is currently no computer-aided diagnosis (CAD) algorithm that has been applied to the classification of histopathological images of laryngeal cancer. Convolutional neural networks (CNNs) are widely used in various other cancer classification tasks. However, the potential global and channel relationships of images may be ignored, which will affect the feature representation ability. Simultaneously, due to the lack of interpretability, the results are often difficult to accept by pathologists. we propose a laryngeal cancer classification network (LPCANet) based on a CNN and attention mechanisms. First, the original histopathological images are sequentially cropped into patches. Then, the patches are input into the basic ResNet50 to extract the local features. Then, a position attention module and a channel attention module are added in parallel to capture the spatial dependency and the channel dependency, respectively. The two modules produce the fusion feature map to enhance the feature representation and improve network classification performance. Moreover, the fusion feature map is extracted and visually analyzed by the grad-weighted class activation map (Grad_CAM) to provide a certain interpretability for the final results. The three-class classification performance of LPCANet is better than those of the five state-of-the-art classifiers (VGG16, ResNet50, InceptionV3, Xception and DenseNet121) on the two original resolutions (534 * 400 and 1067 * 800). On the 534 * 400 data, LPCANet achieved 73.18% accuracy, 74.04% precision, 73.15% recall, 72.9% F1-score, and 0.8826 AUC. On the 1067 * 800 data, LPCANet achieved 83.15% accuracy, 83.5% precision, 83.1% recall, 83.1% F1-score, and 0.9487 AUC. The results show that LPCANet enhances the feature representation by capturing the global and channel relationships and achieves better classification performance. In addition, the visual analysis of Grad_CAM makes the results interpretable, which makes it easier for the results to be accepted by pathologists and allows the method to become a second tool for auxiliary diagnosis.

Narrow-Band Imaging in oncologic otorhinolaryngology: State of the art

OBJECTIVES

To describe the diagnostic performance of Narrow Band Imaging (NBI) combined with White Light Imaging (WLI) in the diagnosis of mucosal lesions at each location of the upper aerodigestive tract, for detection of primary tumor in case of carcinoma of unknown primary, for determination of intraoperative resection margins, and to describe its main diagnostic pitfalls.

MATERIAL AND METHODS

A PubMed search was carried out according to the PRISMA method.

RESULTS

Four hundred and seventy-seven articles published between 2007 and 2020 were identified, 133 of which met the study inclusion criteria and were assessed.

CONCLUSION

The current literature seems to support the use of NBI in diagnosis and/or follow-up of (pre-)malignant head & neck tumors, and in the determination of intraoperative resection margins.

The value of narrowband imaging using the Ni classification in the diagnosis of laryngeal cancer

INTRODUCTION

Narrowband imaging (NBI) is a special endoscopic optical enhancement setting allowing better visualization of mucosal microvasculature compared to white light endoscopy. This study evaluates the validity of NBI using the Ni classification in the detection and differentiation of severe dysplasia (SD) and glottic squamous cell carcinoma (SCC).

METHODS

Patients with suspicious vocal cord lesions underwent conventional white light endoscopy followed by clinically indicated biopsy. At the same time, NBI images were obtained and graded independently. Lesions were graded from I to V according to the Ni classification and compared to histopathological findings.

RESULTS

Fifty-two patients were included in this study (40 SCC and 12 SD). The sensitivity and specificity of NBI in diagnosing laryngeal cancer was 95.0% (CI, 83.9%-99.4%) and 83.3% (CI, 51.6%-97.9%), respectively. The negative likelihood ratio was 0.06. Higher Ni grades correlated very strongly with more advanced disease.

CONCLUSIONS

NBI using the Ni classification is a sensitive diagnostic tool for the detection and differentiation of early neoplastic and preneoplastic glottic lesions. As higher Ni classification correlates strongly with advanced disease, it serves as a useful adjunct to white light endoscopy in the diagnosis of laryngeal cancer.Level of Evidence: Level IV.

Narrowband Imaging for p16+ Unknown Primary Squamous Cell Carcinoma Prior to Transoral Robotic Surgery

Our purpose was to assess the potential utility of narrowband imaging (NBI) as a tool in diagnosing and treating unknown primary oropharyngeal squamous cell carcinoma (OPSCC) in patients prior to diagnostic resection with transoral robotic surgery (TORS). Between 2016 and March 2019, 29 patients with carcinoma of unknown primary meeting inclusion criteria were identified and treated with TORS. NBI was used preoperatively in 9 of 29 patients. A suspected tumor site was delineated by NBI in 8 of 9 patients (89%). Of the patients imaged with NBI, 8 of 9 (89%) patients had a pathologically confirmed tumor following TORS, corresponding to the same 8 suspected tumor sites identified with NBI. In contrast, a primary tumor was localized following TORS in 15 of 20 (75%) patients not evaluated with NBI. Thus, we see NBI as a potentially useful tool for the diagnosis and management of p16+ carcinoma of unknown primary.Level of Evidence: IIb.

Use of narrowband imaging for the diagnosis and screening of laryngeal cancer: A systematic review and meta-analysis

We evaluated the diagnostic accuracy of narrowband imaging (NBI) in terms of detecting laryngeal cancer compared to that of white light endoscopy (WLE). Two reviewers individually searched the six databases for studies published between the first record date and December 31, 2019. We recorded the numbers of true positives, true negatives, false positives, and false negatives. Quality Assessment of Diagnostic Accuracy Studies ver. 2 software was used to assess the studies. The extent of the inter-rater agreement was also measured. The diagnostic odds ratio (OR) associated with NBI was 87.463 (95% confidence interval [CI]: 46.968, 160.873). The area under the summary receiver operating characteristic curve was 0.954. NBI was more diagnostically accurate than WLE, which was associated with a diagnostic OR of 13.750. NBI affords high diagnostic accuracy, thus supporting a role for NBI in the diagnostic work-up of laryngeal cancer.

Evaluation of Vascular Patterns Using Contact Endoscopy and Narrow-Band Imaging (CE-NBI) for the Diagnosis of Vocal Fold Malignancy

The endoscopic detection of perpendicular vascular changes (PVC) of the vocal folds has been associated with vocal fold cancer, dysplastic lesions, and papillomatosis, according to a classification proposed by the European Laryngological Society (ELS). The combination of contact endoscopy with narrow-band imaging (NBI-CE) allows intraoperatively a highly contrasted, real-time visualization of vascular changes of the vocal folds. Aim of the present study was to determine the association of PVC to specific histological diagnoses, the level of interobserver agreement in the detection of PVC, and their diagnostic effectiveness in diagnosing laryngeal malignancy. The evaluation of our data confirmed the association of PVC to vocal fold cancer, dysplastic lesions, and papillomatosis. The level of agreement between the observers in the identification of PVC was moderate for the less-experienced observers and almost perfect for the experienced observers. The identification of PVC during NBI-CE proved to be a valuable indicator for diagnosing malignant and premalignant lesions.

Value of pre- and intraoperative diagnostic methods in suspected glottic neoplasia

PURPOSE

To evaluate the individual and combined ability of videostroboscopy (VS), high-speed digital imaging (HSDI), enhanced endoscopy (EE) and saline infusion (SI) to predict neoplasia, defined as glottic precursor lesion (GPL) or T1a glottic cancer, in patients suspected for glottic neoplasia.

METHODS

A nationwide prospective cohort study of patients treated by cordectomy for suspected GPL or T1a glottic cancer from August 1st 2016 to October 31st 2018 was conducted in the five Danish University Departments of Head and Neck surgery. Sensitivity, specificity, negative and positive predictive values, and area under Receiver Operating Curves (AUC-ROC) were calculated with 95% confidence intervals with respect to the histological diagnosis. Logistic regression with an imputation model for missing data was applied.

RESULTS

261 patients aged 34-91 years participated; 79 (30.3%) with non-neoplasia (i.e., inflammation, papilloma, hyperkeratosis) and 182 (69.7%) neoplasia, hereof 95 (36.4%) with GPL and 87 (33.3%) with T1a glottic cancer. Data from 188 VS, 60 HSDI, 100 preoperative EE, 209 intraoperative EE, and 234 SI were analyzed. In the complete case analysis the AUC-ROC of each diagnostic test was low, but increased when the tests were combined and especially if the combination included EE. However, multinomial logistic regression with imputation showed significant association (p < 0.05) only between age, male gender, and perpendicular vasculature in intraoperative EE, and the endpoint neoplasia.

CONCLUSIONS

Intraoperative EE was the most accurate diagnostic method in detecting neoplasia. The prediction ability of methods applied preoperatively was more limited, but improved when test modalities were combined.

Efficacy of Transnasal Flexible Videoendoscopy With Narrow Band Imaging for Follow-Up of Patients After Transoral Laser Cordectomy

BACKGROUND AND OBJECTIVES

Transoral laser surgery is the optimal surgical treatment modality for the early stages of glottic cancer. To allow for further treatment to be as effective and as minimal as possible, persistent or recurrent tumors should be detected very soon. The main aim is to minimize the risk of necessity of performing a total laryngectomy. Flexible videoendoscopy with narrow band imaging (NBI) was recommended by the European Laryngological Society as a diagnostic method for the follow-up of patients treated for laryngeal cancer. Nevertheless, the efficacy of this technique has not been extensively studied in patients after transoral laser cordectomies for vocal fold cancer. The aim of this study was to evaluate the efficacy of in-office transnasal NBI flexible videoendoscopy in the follow-up of patients after transoral laser cordectomy for glottic cancer. We also focused on describing the specific characteristics of recurrent tumor appearance.

MATERIALS AND METHODS

The presented study was conceived as a prospective study. Ninety-four consecutive patients who underwent transoral laser cordectomy for severe dysplasia, Tis, T1, and T2 glottic cancer in the period from June 2010 to August 2015 were enrolled in the study. All patients were postoperatively regularly followed using transnasal videoendoscopic examinations with NBI. Whenever a suspect lesion was identified during in-office examination, its nature was proven histologically.

RESULTS

We discovered 23 suspect findings in 21 patients by means of flexible videoendoscopy with NBI. Fifteen (65.2%) of them were histologically confirmed as recurrent tumors, whereas in 8 (34.8%) of them, the recurrent tumor was not proven. In two patients, the recurrent tumor was evident on computed tomography or magnetic resonance imaging, but NBI endoscopy did not reveal relapsing disease. The rest (71 patients) were considered true negative. Of the 15 recurrent tumors detected by NBI-coupled videoendoscopy, 8 (53.3%) were identified as submucosal masses with nonsignificant or no vascular changes, 4 (26.7%) were noted as the progression of leukoplakia without visible vascular changes, only 2 (13.3%) showed typical vascular changes, and, in 1 (6.7%) case, evident, significantly exophytic tumor with pathological vascularization was discovered. The sensitivity, specificity, and positive and negative predictive values of the in-office transnasal videolaryngoscopy with NBI were calculated to be 88%, 92%, 71%, and 97%, respectively.

CONCLUSIONS

The results of the study demonstrate that transnasal endoscopy with NBI in an outpatient setting is an excellent method for the follow-up of patients after transoral laser cordectomy for glottic cancer. The method achieves high sensitivity, specificity, and negative predictive value and a slightly low positive predictive value. Nevertheless, we must bear in mind that recurrent tumors after previous endoscopic resection may have a completely different appearance than new tumors originating from previously untreated tissues. Lasers Surg. Med. © 2019 Wiley Periodicals, Inc.