Computer-aided diagnosis of external and middle ear conditions: A machine learning approach.

In medicine, a misdiagnosis or the absence of specialists can affect the patient's health, leading to unnecessary tests and increasing the costs of healthcare. In particular, the lack of specialists in otolaryngology in third world countries forces patients to seek medical attention from general practitioners, whom might not have enough training and experience for making correct diagnosis in this field. To tackle this problem, we propose and test a computer-aided system based on machine learning models and image processing techniques for otoscopic examination, as a support for a more accurate diagnosis of ear conditions at primary care before specialist referral; in particular, for myringosclerosis, earwax plug, and chronic otitis media. To characterize the tympanic membrane and ear canal for each condition, we implemented three different feature extraction methods: color coherence vector, discrete cosine transform, and filter bank. We also considered three machine learning algorithms: support vector machine (SVM), k-nearest neighbor (k-NN) and decision trees to develop the ear condition predictor model. To conduct the research, our database included 160 images as testing set and 720 images as training and validation sets of 180 patients. We repeatedly trained the learning models using the training dataset and evaluated them using the validation dataset to thus obtain the best feature extraction method and learning model that produce the highest validation accuracy. The results showed that the SVM and k-NN presented the best performance followed by decision trees model. Finally, we performed a classification stage -i.e., diagnosis- using testing data, where the SVM model achieved an average classification accuracy of 93.9%, average sensitivity of 87.8%, average specificity of 95.9%, and average positive predictive value of 87.7%. The results show that this system might be used for general practitioners as a reference to make better decisions in the ear pathologies diagnosis.

External Auditory Canal Dimensions, Age, and Cerumen Retention or Impaction in Persons With Down Syndrome.

OBJECTIVE: To investigate associations between age, external auditory canal (EAC) dimensions, and cerumen retention/impaction among persons with Down syndrome (DS). METHODS: This cross-sectional study evaluated EAC dimensions, cerumen retention/impaction, and middle ear status with pneumatoscopy after extraction in 130 persons with DS. Descriptive and inferential statistics correlated age, presence of impacted/retained cerumen, and EAC diameter. RESULTS: Of 260 ears in 67 males and 63 females with average age of 9.48 years, 72.3% (188) had EAC of ≤4 mm. Those ≤1 year were 4.97 times more likely to have cerumen problems than those >1 year (95% CI, 1.45-17.02, P = .011). The odds of having cerumen problems with an EAC diameter of ≤4 mm were 3.31 times higher than with a diameter of 5 mm (95% CI, 1.46-7.50, P = .004), and odds of having cerumen impaction were as much as 6.19 times higher (95% CI, 2.38-16.08, P < .001). Male gender and low-lying external ear were also associated with increased odds of cerumen problems. CONCLUSION: There is a high prevalence of cerumen retention/impaction in persons with DS compared to the general Philippine population and a higher prevalence rate for EAC stenosis than elsewhere. A canal diameter of 4 mm and below and age 1 year or less are associated with a significantly higher likelihood of cerumen retention/impaction.

Cerumen impaction shown by brain magnetic resonance imaging in patients with cognitive impairment.

AIM: Hearing loss is a risk factor for the progression of dementia. Cognitive improvement is occasionally found after removal of cerumen impaction. Because patients with dementia do not usually complain about cerumen impaction, detection methods are important. The present study aimed to investigate whether cerumen impaction is observable using brain magnetic resonance imaging. METHODS: Participants were six patients whose hearing level improved 15 dB or more unilaterally or bilaterally after the removal of cerumen impaction. A radiologist who was blind to the impaction side and whether magnetic resonance imaging scans were taken before or after impaction removal classified cerumen impaction as positive, negative or unclear. RESULTS: Three ears classified as impaction positive and five ears classified as impaction negative corresponded accurately to the presence or absence of cerumen impaction. Among four ears classified as unclear, two did and two did not have cerumen impaction. CONCLUSION: Careful examination of the external ear canal on brain magnetic resonance imaging can be used to detect cerumen impaction.

Investigation of age, sex and menstrual stage variation in human cerumen lipid composition by high performance thin layer chromatography.

OBJECTIVE: The objective of the study was to correlate quantitative changes in the lipid composition of human cerumen with changes in age, sex and menstrual cycle stage. DESIGN: Cerumen samples were collected from the external ear canal and analysed using sequential, one dimensional, high performance thin layer chromatography. SUBJECTS: The following age groups of both sexes were investigated: one to 10 years; 11-18 years; 19-40 years; and 40 years and over. Additionally, cerumen samples from subjects in three stages of the menstrual cycle were compared. RESULTS: In the cerumen samples, the peak values for wax ester and cholesterol occured between the ages of one and 10 years for both sexes. However, squalene and triglyceride content reached maximum levels at puberty. Men aged 19-40 years had a significantly greater percentage of cerumen lipid squalene content than women from the same age group; however, their cholesterol content was found to be lower. Regarding the various menstrual cycle stages, cerumen samples taken at the follicular stage from women aged 19-40 years had a significantly lower free fatty acids content, and higher cholesterol and squalene levels, compared with samples taken in the luteal or menstrual stages. CONCLUSION: The proportions of the lipid constituents of cerumen varied with age, sex and menstrual stage. In cerumen, the main lipid constituent stimulated at puberty appears to be squalene, not wax esters as reported for sebum. The relevance of lipid constituents to cerumen's protective role is discussed.