Usefulness of imputation for the analysis of incomplete otoneurologic data

Machine Learning Techniques for Differential Diagnosis of Vertigo and Dizziness: A Review

Vertigo is a sensation of movement that results from disorders of the inner ear balance organs and their central connections, with aetiologies that are often benign and sometimes serious. An individual who develops vertigo can be effectively treated only after a correct diagnosis of the underlying vestibular disorder is reached. Recent advances in artificial intelligence promise novel strategies for the diagnosis and treatment of patients with this common symptom. Human analysts may experience difficulties manually extracting patterns from large clinical datasets. Machine learning techniques can be used to visualize, understand, and classify clinical data to create a computerized, faster, and more accurate evaluation of vertiginous disorders. Practitioners can also use them as a teaching tool to gain knowledge and valuable insights from medical data. This paper provides a review of the literatures from 1999 to 2021 using various feature extraction and machine learning techniques to diagnose vertigo disorders. This paper aims to provide a better understanding of the work done thus far and to provide future directions for research into the use of machine learning in vertigo diagnosis.

Prevalence of hearing loss among noise-exposed workers within the services sector, 2006-2015

OBJECTIVE

The purpose of this study is to estimate the prevalence of hearing loss among noise-exposed U.S. workers within the Services sector.

METHODS

Audiograms for 1.9 million workers (158,436 within Services) from 2006 to 2015 were examined. Prevalence and adjusted risk for hearing loss as compared with a reference industry were estimated for the Services sector/sub-sectors, and all industries combined.

RESULTS

The prevalence of hearing loss within Services was 17 compared to 16% for all industries combined. However, many sub-sectors greatly exceeded the overall prevalence (10-33% higher) and/or had adjusted risks significantly higher than the reference industry. Workers in Administration of Urban Planning and Community and Rural Development had the highest prevalence (50%), and workers in Solid Waste Combustors and Incinerators had more than double the risk, the highest of any sub-sector. Some sub-sectors traditionally viewed as 'low-risk' also had high prevalences and risks.

CONCLUSIONS

Large numbers of workers within Services have an elevated risk of hearing loss and need immediate hearing conservation efforts. Additional research and surveillance are needed for sub-sectors for which there is low awareness of hearing hazards or a lack of hearing data.

Prevalence of hearing loss among noise-exposed workers within the Mining and Oil and Gas Extraction sectors, 2006-2015

BACKGROUND

The purpose of this study was to estimate the prevalence of hearing loss (HL) among noise-exposed US workers within the Mining, and Oil and Gas Extraction (OGE) sectors.

METHODS

Audiograms of 1.9 million workers across all industries (including 9389 in Mining and 1076 in OGE) from 2006 to 2015 were examined. Prevalence and adjusted risk as compared to a reference industry (Couriers and Messengers) were estimated for all industries combined and the Mining and OGE sectors and subsectors.

RESULTS

The prevalences of HL in Mining and OGE were 24% and 14%, respectively, compared with 16% for all industries combined. Many Mining and one OGE subsector exceeded these prevalences and most had an adjusted risk (prevalence ratio) significantly greater than the reference industry. Some subsectors, particularly in OGE, could not be examined due to low sample size. The prevalences in Construction Sand and Gravel Mining and Natural Gas Liquid Extraction were 36% and 28%, respectively. Workers within Support Activities for Coal Mining had double the risk of HL than workers in the reference industry.

CONCLUSIONS

The many subsectors identified with high prevalences and/or worker risks for HL well above risks in the reference industry need critical attention to conserve worker hearing and maintain worker quality of life. Administrative and engineering controls can reduce worker hazardous noise exposures. Noise and ototoxic chemical exposure information is needed for many subsectors, as is audiometric testing results for OGE workers. Additional research is also needed to further characterize exposures and improve hearing conservation measures.

Prevalence of hearing loss among noise-exposed workers within the agriculture, forestry, fishing, and hunting sector, 2003-2012

BACKGROUND

The purpose of this study was to estimate the prevalence of hearing loss among noise-exposed US workers within the Agriculture, Forestry, Fishing, and Hunting (AFFH) sector.

METHODS

Audiograms for 1.4 million workers (17 299 within AFFH) from 2003 to 2012 were examined. Prevalence, and the adjusted risk for hearing loss as compared with the reference industry (Couriers and Messengers), were estimated.

RESULTS

The overall AFFH sector prevalence was 15% compared to 19% for all industries combined, but many of the AFFH sub-sectors exceeded the overall prevalence. Forestry sub-sector prevalences were highest with Forest Nurseries and Gathering of Forest Products at 36% and Timber Tract Operations at 22%. The Aquaculture sub-sector had the highest adjusted risk of all AFFH sub-sectors (PR = 1.70; CI = 1.42-2.04).

CONCLUSIONS

High risk industries within the AFFH sector need continued hearing conservation efforts. Barriers to hearing loss prevention and early detection of hearing loss need to be recognized and addressed.

Trends in worker hearing loss by industry sector, 1981-2010

BACKGROUND

The purpose of this study was to estimate the incidence and prevalence of hearing loss for noise-exposed U.S. workers by industry sector and 5-year time period, covering 30 years.

METHODS

Audiograms for 1.8 million workers from 1981-2010 were examined. Incidence and prevalence were estimated by industry sector and time period. The adjusted risk of incident hearing loss within each time period and industry sector as compared with a reference time period was also estimated.

RESULTS

The adjusted risk for incident hearing loss decreased over time when all industry sectors were combined. However, the risk remained high for workers in Healthcare and Social Assistance, and the prevalence was consistently high for Mining and Construction workers.

CONCLUSIONS

While progress has been made in reducing the risk of incident hearing loss within most industry sectors, additional efforts are needed within Mining, Construction and Healthcare and Social Assistance.

Do hearing protectors protect hearing?

BACKGROUND

We examined the association between self-reported hearing protection use at work and incidence of hearing shifts over a 5-year period.

METHODS

Audiometric data from 19,911 workers were analyzed. Two hearing shift measures-OSHA standard threshold shift (OSTS) and high-frequency threshold shift (HFTS)-were used to identify incident shifts in hearing between workers' 2005 and 2009 audiograms. Adjusted odds ratios were generated using multivariable logistic regression with multi-level modeling.

RESULTS

The odds ratio for hearing shift for workers who reported never versus always wearing hearing protection was nonsignificant for OSTS (OR 1.23, 95% CI 0.92-1.64) and marginally significant for HFTS (OR 1.26, 95% CI 1.00-1.59). A significant linear trend towards increased risk of HFTS with decreased use of hearing protection was observed (P = 0.02).

CONCLUSION

The study raises concern about the effectiveness of hearing protection as a substitute for noise control to prevent noise-induced hearing loss in the workplace.

Prevalence of workers with shifts in hearing by industry: a comparison of OSHA and NIOSH Hearing Shift Criteria

OBJECTIVE

The purpose of this study was to compare the prevalence of workers with National Institute for Occupational Safety and Health significant threshold shifts (NSTS), Occupational Safety and Health Administration standard threshold shifts (OSTS), and with OSTS with age correction (OSTS-A), by industry using North American Industry Classification System codes.

METHODS

From 2001 to 2010, worker audiograms were examined. Prevalence and adjusted prevalence ratios for NSTS were estimated by industry. NSTS, OSTS, and OSTS-A prevalences were compared by industry.

RESULTS

Twenty percent of workers had an NSTS, 14% had an OSTS, and 6% had an OSTS-A. For most industries, the OSTS and OSTS-A criteria identified 28% to 36% and 66% to 74% fewer workers than the NSTS criteria, respectively.

CONCLUSIONS

Use of NSTS criteria allowing for earlier detection of shifts in hearing is recommended for improved prevention of occupational hearing loss.

Prevalence of hearing loss in the United States by industry

BACKGROUND

Twenty-two million workers are exposed to hazardous noise in the United States. The purpose of this study is to estimate the prevalence of hearing loss among U.S. industries.

METHODS

We examined 2000-2008 audiograms for male and female workers ages 18-65, who had higher occupational noise exposures than the general population. Prevalence and adjusted prevalence ratios (PRs) for hearing loss were estimated and compared across industries.

RESULTS

In our sample, 18% of workers had hearing loss. When compared with the Couriers and Messengers industry sub-sector, workers employed in Mining (PR = 1.65, CI = 1.57-1.73), Wood Product Manufacturing (PR = 1.65, CL = 1.61-1.70), Construction of Buildings (PR = 1.52, CI = 1.45-1.59), and Real Estate and Rental and Leasing (PR = 1.61, CL = 1.51-1.71) [corrected] had higher risks for hearing loss.

CONCLUSIONS

Workers in the Mining, Manufacturing, and Construction industries need better engineering controls for noise and stronger hearing conservation strategies. More hearing loss research is also needed within traditional "low-risk" industries like Real Estate.

Predictive capability of historical data for diagnosis of dizziness

OBJECTIVE

The study examines categorical responses to questions on a comprehensive dizziness questionnaire, to find the overall predictive power of the questionnaire, and to identify which question(s) are most predictive of each diagnosis.

STUDY DESIGN

Retrospective chart review.

SETTING

Specialized dizziness and balance center at a tertiary care hospital.

PATIENTS

A total of 619 patients (aged 19-89 yr, of whom 60% are women and 40% are men) diagnosed with 1 of 23 types of dizziness or postural instability.

INTERVENTION

All patients were administered a standard 163-item dizziness questionnaire (including 77 review of systems items).

OUTCOME MEASURES

Predicted diagnoses from the questionnaire, as determined by binary and multinomial logistic regressions, are compared with an ultimate clinical diagnosis made by an expert neurotologist based on full interview, examination, and clinical tests.

RESULTS

Significant question groupings exist for each of the main diagnoses. A subset of 47 questions under multinomial logistic regression gave high predictive accuracies for migraine (92%), benign paroxysmal positional vertigo (90%) and Ménière's disease (86%), and fair predictive power for vestibular neuritis (63%), contributing to an overall predictive accuracy of 84%. A smaller subset of 32 questions gave an overall predictive accuracy of 71%.

CONCLUSION

The capability of historical data to accurately predict the ultimate diagnosis for dizziness emphasizes the importance of a structured questionnaire in the evaluation of such patients. Future developments include the formulation of a computer-based program to generate a differential diagnosis for the practitioner to consider.

Techniques for biased data distributions and variable classification with neural networks applied to otoneurological data

It is frequently useful and advantageous to investigate not only the classification efficacy of neural networks, but also the reasons for misclassification and relations between input variables and output classes. We have developed novel techniques to disentangle these dilemmas: a network structure and learning strategy for biased output class distributions, a method to measure the classification information incorporated in variables and variable groups, and methods to express properties learned by a network from its structure. We tested these techniques with otoneurological data from the conjunction with vertiginous diseases that we have explored in our previous neural network studies.