DWDavid J. Wellenstein

David J. Wellenstein

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


David J. Wellenstein has a total of 60 co-authors. Here are the top co-authors by number of shared publications:

  • Guido B. van Den Broek

    Department of Otorhinolaryngology and Head and Neck Surgery Radboud University Medical Center Nijmegen the Netherlands

    10 shared publications
  • Jimmie Honings

    Department of Otorhinolaryngology and Head and Neck Surgery, Radboud University Medical Center, Nijmegen, The Netherlands

    7 shared publications
  • Robert P. Takes

    Department of Otorhinolaryngology and Head and Neck Surgery Radboud University, Medical Center Nijmegen Nijmegen The Netherlands

    7 shared publications
  • Henri A. M. Marres

    Department of Otorhinolaryngology and Head and Neck Surgery Radboud University Medical Center Nijmegen the Netherlands

    5 shared publications
  • Henrieke W. Schutte

    Department of Otorhinolaryngology and Head and Neck Surgery, Radboud University Medical Center, Nijmegen, The Netherlands

    4 shared publications
  • Boudewijn E. C. Plaat

    Department of Otorhinolaryngology, Head and Neck Surgery, University of Groningen, University Medical Centre Groningen, Groningen, the Netherlands

    2 shared publications

Latest Publications

  • Open Access

    Real-Time-Capable Detection of Glottic, Supraglottic and Hypopharyngeal Lesions Using Artificial Intelligence During Flexible Endoscopy

    by Nathalie F. van Rhee, Celine M. L. H. Wilmes, Hidde K. Krijnen, Mischa Hofman, Jonathan Woodburn, Rosanne C. Schoonbeek, Inge Wegner, Henri A. M. Marres, Michel R. M. San Giorgi, Gyorgy B. Halmos, et al.

    Background/Objectives: Supraglottic and hypopharyngeal carcinomas are aggressive malignancies that are often diagnosed at advanced stages. Timely recognition of these malignancies is influenced by many factors, such as endoscope quality and experience. This study evaluated the potential of artificial intelligence (AI) to support real-time detection and classification of such lesions during flexible endoscopy in the outpatient clinic. Methods: A previously developed deep learning (DL) algorithm was extended from a glottic lesion model to the unified localization and classification of glottic, supraglottic, and hypopharyngeal lesions during flexible endoscopy. Lesion frames were extracted from endoscopy videos obtained at two head and neck oncology centers and one secondary referral center between 2012 and 2024. These frames were annotated and labeled based on histopathological or clinically confirmed reference diagnoses. The primary outcome was the unified model’s performance in detection of lesions per frame. After training (70% of data), the positive predictive value (precision) and sensitivity (recall) of this model were calculated on an independent test set (30% of data), stratified by subsite and tumor (T-) classification. Secondly, the model’s binary classification performance (benign or malignant) was evaluated. Results: From 490 supraglottic and hypopharyngeal endoscopy videos, 40,059 frames with a benign or malignant lesion were extracted and added to the 56,036 glottic lesion frames in the database, comprising 1336 lesion videos and 123 healthy control videos (total n = 1459). On the test set, the model achieved a detection precision of 92.1% (95% CI: 90.9–93.2) and a recall of 73.3% (95% CI: 69.8–76.6). Detection performance increased with higher T-stage. Among correctly detected lesion instances in a malignancy-dominated test set (65.3% of lesion instances), sensitivity for malignancy detection was 94.3% (95% CI: 91.8–96.5). Combined end-to-end performance for correct malignant lesion detection and classification was estimated at 69.1% (object-level recall 73.3% × classification sensitivity 94.3%). Evaluation of 33 lesion-free videos demonstrated a mean frame-level specificity of 46.6% and a median of 19 false positive detections per video. Conclusions: This is the first study to report a DL model for real-time endoscopic detection and classification of benign and malignant laryngeal and pharyngeal lesions. The developed model showed promising lesion detection and cancer classification performance in the evaluated test set. T1 tumor detection remains an important limitation, particularly because early-stage detection is a primary aim of AI-assisted endoscopy. Further model testing is required in real-world lesion prevalence settings, where external validation and clinical usability should be investigated.


  • Open Access

    A Systematic Review of the Clinical Impact of Implementing Artificial Intelligence in Upper Aerodigestive Tract Endoscopy

    by Celine M. L. H. Wilmes, Arsen Goril Bsc, Henri A. M. Marres, David J. Wellenstein, Guido B. van Den Broek

    Background: Endoscopy is essential in upper aerodigestive tract (UADT) examination, particularly in the early detection of laryngopharyngeal lesions. However, UADT endoscopy remains operator‐dependent and lacks standardized quality metrics. Recent advancements in artificial intelligence (AI) have generated interest in applications within UADT endoscopy. This review evaluates the clinical impact of AI in UADT endoscopy. Methods: A literature review was conducted up to December 31, 2024. Studies were evaluated using the modified Quality Assessment of Diagnostic Accuracy Studies (QUADAS)‐2 tool. Results: Eighty‐three studies were included. Results indicate that AI in UADT endoscopy achieves diagnostic accuracy, sensitivity, and specificity rates comparable to experts, with optimal outcomes combined with human expertise. AI also demonstrated significantly faster inference times. Conclusions: This review highlights AI's potential to enhance clinical impact in UADT endoscopy, especially when combined with human expertise. However, the limited focus on real‐time clinical translation underscores the need for further research to enable effective integration into clinical practice.


  • Open Access

    Is ultraviolet light disinfection fit to be the future standard for the disinfection of flexible endoscopes without a working channel?

    by Yana Halmans, David J. Wellenstein, Michael Romijn, Suzan Cremers, Jannie J. Smit, Joost Hopman, Robert P. Takes, Guido B. van Den Broek

    Objective: To investigate colony-forming unit (CFU) reduction on contaminated flexible endoscopes (FEs) without a working channel after UV-C light disinfection, compared to the current disinfection method with the endoscope washer disinfector (EWD).Design, setting and participants: After pharyngolaryngoscopy, a manual pre-cleaning with tap water was performed. A culture was then collected by rolling the distal 8-10 cm of the FE over an Agar plate. The FE was disinfected using the D60 (60-s disinfection process with UV-C light) or the EWD (gold standard reprocessing process with water and chemicals). Another culture was then taken. After incubation, a CFU count was performed.Results: A total of 200 FEs without a working channel were divided equally between the two disinfection groups. After clinical use and manual pre-cleaning, 84 of the 100 (84.0%) (UV-C light group) and 79 of the 100 (79.0%) (EWD) FEs were contaminated with at least 1 CFU. FEs that showed no contamination after use were excluded from further analysis. After disinfection with UV-C light, 72 (85.7%) FEs showed no contamination (i.e., 0 CFUs) versus 66 (83.5%) FEs after reprocessing with the EWD.Conclusion: There is no difference in CFUs reduction on contaminated FEs without a working channel between UV-C light disinfection and the current gold standard, the EWD.


  • Top Cited Papers
    Open Access

    Detection of laryngeal carcinoma during endoscopy using artificial intelligence

    by David J. Wellenstein, Jonathan Woodburn, Henri A. M. Marres, Guido B. van Den Broek

    Background: The objective of this study was to assess the performance and application of a self‐developed deep learning (DL) algorithm for the real‐time localization and classification of both vocal cord carcinoma and benign vocal cord lesions.Methods: The algorithm was trained and validated upon a dataset of videos and photos collected from our own department, as well as an open‐access dataset named “Laryngoscope8”.Results: The algorithm correctly localizes and classifies vocal cord carcinoma on still images with a sensitivity between 71% and 78% and benign vocal cord lesions with a sensitivity between 70% and 82%. Furthermore, the best algorithm had an average frame per second rate of 63, thus making it suitable to use in an outpatient clinic setting for real‐time detection of laryngeal pathology.Conclusion: We have demonstrated that our developed DL algorithm is able to localize and classify benign and malignant laryngeal pathology during endoscopy.


  • Open Access

    UV-C light: the future of disinfection of flexible endoscopes without a working channel?

    by Yana Halmans, David Wellenstein, Michael Romijn, Suzan Cremers, Jannie J. Smit, Joost Hopman, Robert Takes, Guido van Den Broek

    Objective: To investigate Colony-Forming Unit (CFU) reduction on contaminated flexible endoscopes without a working channel after UV-C light disinfection compared to the current disinfection method with the Endoscope Washer Disinfector. Design, setting: After pharyngolaryngoscopy, a manual pre-cleaning with tap water was performed. A culture was then collected by rolling the distal 8-10 cm of the flexible endoscope over an Agar plate. The flexible endoscope was disinfected using the D60 (60-second disinfection process with UV-C light) or the Endoscope Washer Disinfector (golden standard reprocessing process with water and chemicals). Another culture was then taken. After incubation, a CFU count was performed. Results: Two hundred flexible endoscopes without a working channel were divided equally between the two disinfection groups. After clinical use and manual pre-cleaning, 84 of the 100 (84.0%) (UV-C light group) and 79 of the 100 (79.0%) (EWD) flexible endoscopes were contaminated with at least 1 CFU. Flexible endoscopes that showed no contamination after use were excluded from further analysis. After disinfection with UV-C light, 72 (85.7%) flexible endoscopes showed no contamination (i.e. 0 CFUs) versus 66 (83.5%) flexible endoscopes after reprocessing with the Endoscope Washer Disinfector. Conclusion: There is no difference in CFUs reduction on contaminated flexible endoscopes without a working channel between UV-C light disinfection and the current gold standard, the Endoscope Washer Disinfector.


  • Open Access

    Optimizing Settings for Office‐Based Endoscopic CO2 Laser Surgery Using an Experimental Vocal Cord Model

    by Anouk S. Schimberg, Tim M. Klabbers, David J. Wellenstein, Floris Heutink, Jimmie Honings, Ilse van Engen‐Van Grunsven, Rudolf M. Verdaasdonk, Robert P. Takes, Guido B. van Den Broek

    Objectives/Hypothesis: To provide insight in the thermal effects of individual laser settings in target tissues to optimize flexible endoscopic CO2 laser surgery treatment.Study Design: Experimental laboratory study.Methods: Thermal effects of the CO2 laser using a fiber delivery system were visualized using the color Schlieren technique in combination with a polyacrylamide gel tissue model. Variable settings were used for emission mode, power, laser fiber distance, and laser duration, which were evaluated in every possible combination. Collateral thermal expansion and incision depth were measured. To validate the model, the results were compared to histology after CO2 laser irradiation of ex vivo human vocal cords, and the intraclass correlation coefficient was calculated. Thermal damage and incision depth were measured by a blinded pathologist.Results: Of all parameters studied, duration of laser irradiation had the greatest effect on thermal expansion. Increased distance between laser tip and target tissue resulted in significantly reduced incision depth and increased thermal expansion. Pulsed emission modes led to increased incision depths. The intraclass correlation coefficient for consistency between the model setup and the ex vivo human vocal cords was classified as “fair.”Conclusions: By using high‐intensity pulsed lasers at minimal distance to the target tissue, exposure times and subsequent damage to surrounding tissue can be reduced. If an evaporation technique is used, lower power in continuous wave at a larger distance to the target tissue will lead to superficial but broader thermal effects. The model setup used in this study is a valid model to investigate laser‐induced thermal effects in vocal cord tissue.Level of Evidence: NA Laryngoscope, 130:E680–E685, 2020


  • Open Access

    Office‐based CO2 laser surgery for benign and premalignant laryngeal lesions

    by David J. Wellenstein, Jimmie Honings, Anouk S. Schimberg, Henrieke W. Schutte, Jasmijn M. Herruer, Frank J.A. van Den Hoogen, Robert P. Takes, Guido B. van Den Broek

    Objective: Patients with laryngeal pathology are often treated with CO2 laser surgery, usually in the operating room under general anesthesia. Although office‐based laser surgery using several other laser types has been investigated, prospective studies on office‐based CO2 laser surgery are scarce. Our goal was to investigate the feasibility of office‐based CO2 laser surgery for benign and premalignant laryngeal pathology by analyzing completion rate, safety, effect on voice quality, and success rate (i.e., no residual or recurrent disease).Methods: A prospective cohort study was performed of 30 consecutive procedures. Inclusion started in June 2016 and was completed in August 2018. Adult patients with clinically benign or premalignant laryngeal lesions who could not undergo transoral laser microsurgery in the operating room under general anesthesia were included. Reasons were either contraindications for general anesthesia, previously failed therapeutic laryngoscopy under general anesthesia, and preference of a procedure under topical anesthesia by the patient. The mean follow‐up was 9 months.Results: Thirty procedures were performed in 27 patients (24 males) with an average age of 62 years. Twenty‐nine (97%) procedures were fully completed without complications. The mean preoperative Voice Handicap Index (VHI) score (VHI 44) significantly decreased 2 months (VHI 28, P = 0.032) and 6 months (VHI 14, P < 0.001) after the procedure. Almost two‐thirds of patients showed no residual or recurrent disease at their follow‐up visits.Conclusion: Office‐based CO2 laser surgery is a feasible and safe procedure that results in significant voice‐quality improvement. Almost two‐thirds of patients did not require further treatment.Level of Evidence: 2 Laryngoscope, 130:1503–1507, 2020


  • Topical Anesthesia for Endoscopic Office-based Procedures of the Upper Aerodigestive Tract

    by David J. Wellenstein, Raymond A.B. van der Wal, Henrieke W. Schutte, Jimmie Honings, Frank J.A. van Den Hoogen, Henri A.M. Marres, Robert P. Takes, Guido B. van Den Broek


  • Open Access

    Cost analysis of office-based transnasal esophagoscopy

    by David J. Wellenstein, Jimmie Honings, Henrieke W. Schutte, Jasmijn M. Herruer, Frank J. A. van Den Hoogen, Henri A. M. Marres, Robert P. Takes, Guido B. van Den Broek

    Although office-based transnasal esophagoscopy has been investigated extensively, a cost analysis is still lacking. We performed a cost analysis combined with feasibility study for two diagnostic processes: patients with globus pharyngeus and/or dysphagia, and hypopharyngeal carcinoma.