Errors in the assessment of tympanic membrane perforations

A Proof-of-Concept Computer Vision Approach for Measurement of Tympanic Membrane Perforations

OBJECTIVE

Size, an important characteristic of a tympanic membrane perforation (TMP), is commonly assessed with gross estimation via visual inspection, a practice which is prone to inaccuracy. Herein, we demonstrate feasibility of a proof-of-concept computer vision model for estimating TMP size in a small set of perforations.

METHODS

An open-source deep learning architecture was used to train a model to segment and calculate the area of a perforation and the visualized tympanic membrane (TM) in a set of endoscopic images of mostly anterior and relatively small TMPs. The model then computed relative TMP size by calculating the ratio of perforation area to TM area. Model performance on the test dataset was compared to ground-truth manual annotations. In a validation survey, otolaryngologists were tasked with estimating the size of TMPs from the test dataset. The primary outcome was the average absolute error of model size predictions and clinician estimates compared to sizes determined by ground-truth manual annotations.

RESULTS

The model's average absolute error for size predictions was a 0.8% overestimation for all test perforations. Conversely, among the 38 survey respondents, the average clinician error was a 11.0% overestimation (95% CI, 5.2-16.7%, p = 0.003).

CONCLUSIONS

In a small sample of TMPs, we demonstrated a computer vision approach for estimating TMP size is feasible. Further validation studies must be done with significantly larger and more heterogenous datasets.

LEVEL OF EVIDENCE

N/A Laryngoscope, 134:2906-2911, 2024.

Assessment of tympanic membrane perforation size using ImageJ software by ENT clinicians of different grades

OBJECTIVE

To compare visual estimation versus ImageJ calculation of tympanic membrane perforation size in the paediatric population between clinicians of different experience.

METHODS

Five images of tympanic membrane perforations in children, captured using an otoendoscope, were selected. The gold standard was the ImageJ results by one consultant otologist. Consultants, registrars and Senior House Officers or equivalent were asked to visually estimate and calculate the perforation size using ImageJ software.

RESULTS

The mean difference in variation from gold standard between visual estimation and ImageJ calculation was 12.16 per cent, 95 per cent CI (10.55, 13.78) p < 0.05, with ImageJ providing a more accurate estimation of perforation. Registrars were significantly more accurate at visual estimation than senior house officers. There was no statistically significant difference in ImageJ results between the different grades.

CONCLUSION

Using ImageJ software is more accurate at estimating tympanic membrane perforation size than visual assessment for all ENT clinicians regardless of experience.

Tympanic membrane perforations: the importance of etiology, size and location

PURPOSE

The ability to predict the degree of a conductive hearing loss caused by a tympanic membrane perforation is important for every otologist, as it may require additional diagnostic tests and prevent unexpected intraoperative findings. The aim of this study was to correlate the various characteristics of a perforation (etiology, size, location, involvement of the manubrium or umbo) with the degree and frequency predominance of the consequent hearing loss.

METHODS

A transversal study in a tertiary hospital center was conducted between July 2019 and December 2020. Fifty-eight patients with 65 tympanic perforations underwent a comprehensive medical and audiological evaluation, which included an otoendoscopy. An image processing software (ImageJ®) was used to measure the perforated area. The qualitative variables were etiology, affected quadrants, presence of myringosclerosis and involvement of umbo or manubrium of the malleus. The air-bone gap was measured at 250, 500, 1000, 2000 and 4000 Hz. Primary outcomes (mean air-bone gap and pure-tone average) were evaluated to find clinical factors associated with worse hearing.

RESULTS

Data collected from 50 ears was included. Perforation size showed a positive statistically significant correlation with the air-bone gap (r = .508; p < .001) and pure-tone average (r = .375; p < .001). Higher air-bone gaps were found in perforations involving the posterior quadrants and the manubrium (p < .001 and p = .031, respectively). Inflammatory causes showed higher bone and air conduction pure-tone averages (p = .031 and p = .084, respectively) compared to traumatic or iatrogenic. An "inverted V" pattern of the air-bone gap, with the 2 kHz frequency being the least affected, was a consistent finding. However, it was not due to the Carhart's notch in bone conduction.

CONCLUSION

The conductive hearing loss resulting from a tympanic membrane perforation is etiology, size and location-dependent, with higher losses occurring for inflammatory backgrounds, large perforations and when the posterior quadrants or the manubrium are involved. If the "inverted V" is absent, additional middle ear pathology should be investigated.

Functional correlations of tympanic membrane perforation size

OBJECTIVE

The correlation between tympanic membrane perforations and hearing loss was studied.

STUDY DESIGN

Prospective data from 220 patients, who underwent primary surgery for simple chronic otitis media with a perforated eardrum, were analyzed.

SETTING

Tertiary referral center.

PATIENTS

One hundred fifty-one patients with 155 eardrum perforations, which were checked for correct diagnosis, normal middle-ear status, and integrity of the ossicular chain, were included.

INTERVENTIONS

All patients underwent primary myringoplasty.

MAIN OUTCOME MEASURES

Preoperative conductive hearing loss due to eardrum perforations.

RESULTS

Hearing loss shows a linear relationship with increasing eardrum perforation size. Umbo involvement shows a worsening of the hearing by 5 to 6 dB (p < 0.0001). The least impact of a perforation is seen at the resonance frequency of 2 kHz. Above and below 2 kHz, an 'inverted V shape' of the air-bone gap is a consistent finding. If the air-bone gap exceeds the 'inverted V-shape' pattern, additional pathology behind the eardrum must be assumed and addressed.

CONCLUSION

We propose using standardized photographs or drawings to document preoperative perforation sizes. A linear relationship between the size of a perforation and the conductive hearing loss does exist. Umbo involvement at the perforation margin may worsen the hearing by 5 to 6 dB, whereas the position of the perforation itself does not play a role. The least impact of a perforation is seen at the resonance frequency of 2 kHz. An "inverted V-shape" pattern, above and below 2 kHz, of the air-bone gap is a consistent finding. If the air-bone gap exceeds this pattern, additional pathology behind the eardrum perforation must be assumed and addressed.

[Temporal bone surgery. Surgical sequelae and complications]

The saying "no surgery, no surgical complications" is certainly true for all specialties. Three categories of undesired events may occur following surgery: surgical sequelae, failure to cure and complications. A critical self-analysis of surgical complications often reveals that these arise in vexing fashion according to Murphy's law, i.e. "what can go wrong, will go wrong". Incomplete preoperative evaluation, insufficient exposure of the surgical field, failure to identify surgical landmarks and misjudgment of the patient's preoperative complaints may culminate in an undesired surgical outcome. Modern preoperative radiology, the possibilities of actual or even virtual temporal bone drilling in the laboratory and the surgeon's level of experience all ensure the relative rarity of relevant surgical complications in otology.

Quantitative analysis of tympanic membrane perforation: a simple and reliable method

Background:

Accurate assessment of the features of tympanic membrane perforation, especially size, site, duration and aetiology, is important, as it enables optimum management.

Aim and objectives:

To describe a simple, cheap and effective method of quantitatively analysing tympanic membrane perforations.

Materials and methods:

The system described comprises a video-otoscope (capable of generating still and video images of the tympanic membrane), adapted via a universal serial bus box to a computer screen, with images analysed using the Image J geometrical analysis software package. The reproducibility of results and their correlation with conventional otoscopic methods of estimation were tested statistically with the paired t-test and correlational tests, using the Statistical Package for the Social Sciences version 11 software.

Results:

The following equation was generated: P/T × 100 per cent = percentage perforation, where P is the area (in pixels2) of the tympanic membrane perforation and T is the total area (in pixels2) for the entire tympanic membrane (including the perforation). Illustrations are shown. Comparison of blinded data on tympanic membrane perforation area obtained independently from assessments by two trained otologists, of comparative years of experience, using the video-otoscopy system described, showed similar findings, with strong correlations devoid of inter-observer error (p = 0.000, r = 1). Comparison with conventional otoscopic assessment also indicated significant correlation, comparing results for two trained otologists, but some inter-observer variation was present (p = 0.000, r = 0.896). Correlation between the two methods for each of the otologists was also highly significant (p = 0.000).

Conclusion:

A computer-adapted video-otoscope, with images analysed by Image J software, represents a cheap, reliable, technology-driven, clinical method of quantitative analysis of tympanic membrane perforations and injuries.

Assessments of the Size of Tympanic Membrane Perforations: A Comparison of Clinical Estimations with Video-Otoscopic Calculations

We conducted a study to determine how accurate various ENT specialists were in estimating the size of 100 tympanic membrane (TM) perforations with standard otoscopy. The specialists included, in descending order of rank, 2 Consultant Surgeons, 2 Senior Registrars, and 2 Registrars, all of whom had confirmed good vision. We compared their estimates, which were made independently and expressed as a percentage of the total area of the TM, with exact measurements calculated with computer-based video-otoscopy. We found that the video-otoscopic calculations were far superior to the estimates of the specialists, even the most experienced Consultants (p < 0.01). We recommend that video-otoscopy he used whenever possible.

Tympanic membrane perforations in children

We have assessed children with perforations of the pars-tensa presenting to an otolaryngology clinic over a 7-year period. We found that over 70% of the perforations followed the insertion of a ventilation tube. Using a novel method of dividing the tympanic membrane into quadrants, we were able to assess the sites and grade the size of perforations. The most common site of involvement, in all perforations, was the antero-inferior quadrant. With the exception of perforations following the use of T-tubes, we found little difference between the sites, sizes and hearing levels of ventilation tube and non-ventilation tube related perforations. Larger perforations resulted in a greater degree of hearing loss.