Materiały źródłowe

• #1 Prediction of hearing recovery with deep learning algorithm in sudden sensorineural hearing loss | Scientific Reports

  https://www.nature.com/articles/s41598-024-70436-0

  This study aimed to establish a deep learning-based predictive model for the prognosis of idiopathic sudden sensorineural hearing loss (SSNHL). […] One of the key challenges in managing SSNHL is predicting treatment outcomes or prognoses, so that a personalized, more aggressive interventions for patients with anticipated poor outcomes. […] Recent advancements in machine learning, particularly deep learning algorithms, offer promising ways to improve prognostic accuracy in medical decision making. […] Our results showed promising accuracy across both Siegel’s criteria and AAO-HNS classification. […] These findings demonstrate the potential of deep learning algorithms for prognosis prediction in SSNHL. […] The deep learning model identified several significant features instrumental in predicting hearing recovery.

• #1 Predicting Sudden Sensorineural Hearing Loss Recovery with Patient-Personalized Seigel’s Criteria Using Machine Learning

  https://www.mdpi.com/2075-4418/14/12/1296

  Predicting Sudden Sensorineural Hearing Loss Recovery with Patient-Personalized Seigel’s Criteria Using Machine Learning […] Background: Accurate prognostic prediction is crucial for managing Idiopathic Sudden Sensorineural Hearing Loss (ISSHL). Previous studies developing ISSHL prognosis models often overlooked individual variability in hearing damage by relying on fixed frequency domains. This study aims to develop models predicting ISSHL prognosis one month after treatment, focusing on patient-specific hearing impairments. […] The main poor prognostic factors identified are severe hearing loss, high-frequency hearing loss, recovery starting duration greater than 2 weeks, advanced age, history of vertigo, and late initiation of treatment. […] Previous studies have developed machine learning models to accurately predict the prognosis of ISSHL. […] However, relying solely on these fixed-frequency domains to assess hearing recovery may overlook the nuances of patient-specific hearing impairment, potentially resulting in imprecise evaluations. […] The primary objective of this study is to develop machine learning models for predicting the prognosis of patients with ISSHL one month after treatment, focusing on hearing-impaired frequency domains. […] We elucidated the effects of clinical variables on the prediction of the model. […] The recovery status of ISSHL, which is the target of the machine learning model, was determined using Patient-Personalized Siegel’s criteria. […] The application of traditional Siegel’s criteria, focusing on the 0.5, 1, 2, and 3 kHz frequency domains, identified 334 patients (57.5%) as non-recovery and 247 patients (42.5%) as recovery. […] This shift underscores the importance of considering patient-specific impaired frequency ranges in hearing assessments. […] The soft voting classifier emerged as the most effective model, achieving the highest AUROC and precision scores among all the models tested in both validation and test evaluations. […] The distribution shift in the recovery assessment was not statistically significant based on the result of the chi-square test. […] Our study underscores the significance of the ISSHL prognostic model, which accounts for patient-specific hearing impairments. Further research is required to validate our findings in larger and more diverse patient populations.

• #1 Prediction of hearing recovery in unilateral sudden sensorineural hearing loss using artificial intelligence | Scientific Reports

  https://www.nature.com/articles/s41598-022-07881-2

  Our findings strongly suggest that high-frequency hearing loss could affect the prognosis of ISSNHL. […] The next important feature was the initial hearing result. […] Interestingly, initial hearing levels of the unaffected ear were also important for the prognosis. […] To the best of our knowledge, no study has identified serum BUN as a prognostic factor. […] Tg could be a poor prognostic factor as seen when the value was high. […] Regarding age, older patients tend to have negative mean SHAP values, as the cut-off value was 50.7. […] The duration from onset to treatment is known to be a prognostic factor. Early treatment had a positive mean SHAP value. […] In conclusion, our machine and deep learning models showed superior performance in predicting the prognosis of ISSNHL.

• #1 Prediction of hearing recovery in unilateral sudden sensorineural hearing loss using artificial intelligence | Scientific Reports

  https://www.nature.com/articles/s41598-022-07881-2

  Despite the significance of predicting the prognosis of idiopathic sudden sensorineural hearing loss (ISSNHL), no predictive models have been established. […] This study used artificial intelligence to develop prognosis models to predict recovery from ISSNHL. […] Using the SHapley Additive exPlanation method, we found that the initial audiogram shape is the most important prognostic factor. Machine/deep learning methods were successfully established to predict the prognosis of ISSNHL. […] The LightGBM achieved an AUROC CI of 0.915 (95% CI 0.8640.967) and a balanced accuracy (BACC) of 0.84. […] The MLP achieved an AUROC of 0.911 (95% CI 0.8590.964) and a BACC of 0.823. […] We confirmed that LightGBM and MLP were significantly superior to conventional LR as regards AUROC. […] The initial audiogram shape had the top feature importance.

• #1 Predicting Sudden Sensorineural Hearing Loss Recovery with Patient-Personalized Seigel’s Criteria Using Machine Learning

  https://pmc.ncbi.nlm.nih.gov/articles/PMC11202901/

  This shift underscores the importance of considering patient-specific impaired frequency ranges in hearing assessments. […] The Patient-Personalized Siegel criteria led to notable changes in ISSHL recovery classification in 41 patients (7%). […] These findings emphasize the need for personalized assessments to accurately determine the prognosis of patients with ISSHL. […] The soft voting classifier emerged as the most effective model, achieving the highest AUROC and precision scores among all the models tested in both validation and test evaluations. […] Factors that negatively impact prognosis were discovered based on the SHAP value, which included the following: average hearing threshold, dizziness, delay in treatment onset, and descending audiogram type. […] Our study underscores the significance of the ISSHL prognostic model, which accounts for patient-specific hearing impairments.

• #1 Prediction of hearing recovery with deep learning algorithm in sudden sensorineural hearing loss | Scientific Reports

  https://www.nature.com/articles/s41598-024-70436-0

  Our deep learning model also emphasized lesser-known predictors, such as tinnitus or ear fullness, over well-known predictors such as time to treatment onset or the presence of vertigo. […] Using data from 1108 SSNHL patients, we developed a deep learning model to predict hearing recovery with high accuracy, suggesting its usefulness for personalized patient counseling.

• #1 Hearing recovery prediction and prognostic factors of idiopathic sudden sensorineural hearing loss: a retrospective analysis with a deep neural network model | Brazilian Journal of Otorhinolaryngology

  https://www.elsevier.es/en-revista-brazilian-journal-otorhinolaryngology-english-edition–497-articulo-hearing-recovery-prediction-prognostic-factors-S1808869423000356

  Various machine learning methods used to predict hearing recovery in ISSHL patients. […] The deep neural network method showed the highest predictive performance. […] Initial and early post-treatment hearing levels were significant prognostic factors. […] Machine learning may help predict ISSHL prognosis, as shown in the study. […] Idiopathic Sudden Sensorineural Hearing Loss (ISSHL) is an otologic emergency, and an early prediction of prognosis may facilitate proper treatment. […] Recovery was predicted by various machine learning models. […] The deep neural network model showed the highest predictive performance (accuracy, 88.81%; area under the receiver operating characteristic curve, 0.9448). […] Initial hearing level of affected and non-affected ear, post-treatment (2-weeks) hearing level of affected ear were significant prognostic factors.

• #1 Extended Preoperative Audiometry for Outcome Prediction and Risk Analysis in Patients Receiving Cochlear Implants

  https://www.mdpi.com/2077-0383/12/9/3262

  Extended Preoperative Audiometry for Outcome Prediction and Risk Analysis in Patients Receiving Cochlear Implants […] Background: The outcome of cochlear implantation has improved over the last decades, but there are still patients with less benefit. […] The aim of this study was therefore to examine preoperative pure-tone audiogram and speech comprehension as well as aetiology, to investigate their relationship with postoperative speech comprehension in CI recipients. […] The strongest correlation with postoperative speech comprehension was seen for monosyllabic words with hearing aid at 80 dB. […] Meningitis and congenital diseases were strongly negatively associated with postoperative word comprehension. […] The multivariate model was able to explain 40% of postoperative variability. […] Speech comprehension with hearing aid at 80 dB can be used as a supplementary preoperative indicator of CI-aided speech comprehension and should be measured regularly in the clinical routine. […] Combining audiological and aetiological variables provides more insights into the variability of the CI outcome, allowing for better patient counselling.

• #1 Extended Preoperative Audiometry for Outcome Prediction and Risk Analysis in Patients Receiving Cochlear Implants

  https://www.mdpi.com/2077-0383/12/9/3262

  In clinical practice, preoperative diagnostics and anamnesis are used as a predictive basis for the postoperative result. […] For that test, Hoppe et al. have shown that unaided maximum word recognition (WRSmax) can serve as a predictor for the minimum expected benefit with a CI. […] Furthermore the monosyllabic speech comprehension at 65 dB with a hearing aid (WRS65(HA)) can be used as well. […] In their multicentre cohort, a positive correlation was found between the time elapsed since CI implantation and postoperative speech comprehension. […] The outcome was positively correlated with genetic aetiology and the presence of Ménière’s disease. […] An updated model with more influencing factors, including the pure-tone average, raised the explanatory power for postoperative speech comprehension to 22% within the same patient collective. […] These results demonstrate that it appears reasonable to include etiological aspects, on the basis of their correlation with speech comprehension. […] The model showed a positive correlation between the postoperative outcome and (i) preoperative speech comprehension with optimum hearing-aid fitting and (ii) age at onset of deafness. […] A negative correlation was found between speech comprehension and meningitis. […] Overall, the positive association between preoperative speech comprehension and postoperative outcome observed by other research groups was confirmed in this study. […] The predictive value is limited by the high proportion of patients with 0% word comprehension. […] The search for variables that can explain and predict interindividual differences in postoperative outcomes is a task for future research.

• #1 Severe hearing loss in seniors: Surgical risks and audiological outcome after cochlear implantation

  https://www.oatext.com/Severe-hearing-loss-in-seniors-Surgical-risks-and-audiological-outcome-after-cochlear-implantation.php

  Background: Aim of study was the evaluation of hearing improvement, quality of life and surgical risks in geriatric patients receiving a cochlear implant. […] Conclusion: Geriatric patients over 70 yrs. significantly benefit from cochlear implantation, though outcome in speech perception is marginally inferior to younger patients. […] All groups benefitted significantly from cochlear implantation in everyday speech perception according to the APHAB questionnaire and audiological results. […] Seniors younger than 70 years showed no difference in speech perception after CI to the younger control group. This implies that cochlear implantation for this group can be considered as well as for younger patients. With regards to speech perception, seniors older than 70 years seem to benefit marginally less than the younger ones.

• #1 Severe hearing loss in seniors: Surgical risks and audiological outcome after cochlear implantation

  https://www.oatext.com/Severe-hearing-loss-in-seniors-Surgical-risks-and-audiological-outcome-after-cochlear-implantation.php

  Our results imply that rising age itself only has a small impact on risk factors and development of speech comprehension in people aged 70 years and above. The patients older than 80 yrs. had a similar outcome in speech perception and complications and were comparable to those aged 70-79 yrs. […] However speech perception after CI showed a significant increase with an overall perception of 53% for monosyllables and 98% for bisyllables – a good postoperative result.

• #1 Preoperative variables affecting outcome of cochlear implant | The Egyptian Journal of Otolaryngology | Full Text

  https://ejo.springeropen.com/articles/10.1186/s43163-024-00563-y

  Cochlear implants made a great impact in the management of severe-to-profound hearing loss in both children and adults. […] However, the final outcome in pediatric implantation is not predictable as there are large number of factors which will affect the outcome of cochlear implantation. […] Based on our findings, two most important factors affecting outcome of cochlear implantation were the age at implantation and the hearing age. […] The end result in pediatric implantation is still not totally predicted because there are numerous factors that, either individually or collectively, will determine how well the cochlear implant works. […] By classifying these factors, practitioners are able to provide informed preoperative prognosis and may be willing to manipulate factors in an effort to get the best results.

• #1 Preoperative variables affecting outcome of cochlear implant | The Egyptian Journal of Otolaryngology | Full Text

  https://ejo.springeropen.com/articles/10.1186/s43163-024-00563-y

  Poor performance has been linked to a number of known factors, including implantation at a late age, minimal nerve survival, inappropriate fitting, insufficient cognitive ability, social and educational circumstances that stress interpersonal interaction, and little parental support. […] This study was carried out to gain a better understanding of the prognostic factors that promote language development and auditory abilities after cochlear implantation. […] Age of the studied children at time of cochlear implantation was statistically significant predictor for CI outcome as regard receptive language quotient and also for N1 latency. […] Hearing age of the studied children was a statistically significant predictor for CI outcome as regards P1 latency. […] Significant negative correlation was found between age of the studied children and their hearing age with their outcome measures, while statistically significant positive correlation was found between these outcomes and their age of implantation.

• #1 Preoperative variables affecting outcome of cochlear implant | The Egyptian Journal of Otolaryngology | Full Text

  https://ejo.springeropen.com/articles/10.1186/s43163-024-00563-y

  Earlier age of implantation was associated with decrease in average FFA threshold and decrease the latencies of S-CAEPS. […] We performed multivariable regression analysis of the variables hierarchically proven to be predictors of CI outcome, and we found that the statistically significant predictors of CI outcome as regard P1 latency were hearing age, N1 latency was age at implantation, and as regard receptive language, quotients were age in month and age of implantation, meaning that earlier age of implantation and longer duration of CI use were highly significant predictors of language output. […] Age of the studied children at time of cochlear implantation and hearing age were the most important predictors that affect speech and language development.

• #1 Assessing hearing loss in older adults with a single question and person characteristics; Comparison with pure tone audiometry in the Rotterdam Study | PLOS One

  https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0228349

  Hearing loss (HL) is a frequent problem among the elderly and has been studied in many cohort studies. […] The ability of the question to classify hearing loss, reflected in the area under the curve (AUC), was 0.70 (95% CI 0.68, 0.71) for mild hearing loss or worse and 0.86 (95% CI 0.85, 0.87) for moderate hearing loss or worse. […] Self-reported hearing loss using a single question has a good ability to detect hearing loss in older adults, especially when age is accounted for. A single question cannot substitute audiometry, but it can assess hearing loss on a population level with reasonable accuracy. […] The ability of the question to identify hearing loss increases substantially when other factors are taken into account, with age being the most important one. […] Self-reported hearing loss, using the question Do you have any difficulty with your hearing (without hearing aids)?, has reasonable ability to detect both mild and moderate hearing loss in older adults, especially when the age of the individual is factored into the answer. This finding is mainly of importance for large population-based studies in which audiometry is absent but hearing loss still has to be quantified. A single question cannot substitute regular audiometry, but it is able to assess hearing on a population level with reasonable accuracy, adjusted for the age of the individuals.

• #1 Sudden Hearing Loss: Practice Essentials, Pathophysiology, Epidemiology

  https://emedicine.medscape.com/article/856313-overview

  Prognosis for some hearing recovery is good. […] The value of steroids in the treatment of idiopathic sudden sensorineural hearing loss (ISSNHL) remains unclear. […] Prognosis and treatment of sudden sensorineural hearing loss. […] Prognostic Factors for Recovery from Sudden Sensorineural Hearing Loss: A Retrospective Study. […] Prognosis of Bilateral Sudden Sensorineural Hearing Loss: A Systematic Review and Meta-Analysis. […] Clinical Features and Prognosis of Sudden Sensorineural Hearing Loss Secondary to Intralabyrinthine Hemorrhage.

• #2 Hearing recovery prediction and prognostic factors of idiopathic sudden sensorineural hearing loss: a retrospective analysis with a deep neural network model | Brazilian Journal of Otorhinolaryngology

  https://www.elsevier.es/en-revista-brazilian-journal-otorhinolaryngology-english-edition–497-articulo-hearing-recovery-prediction-prognostic-factors-S1808869423000356

  The prognosis of ISSHL is generally predicted using conventional statistical models, such as logistic regression. […] Significant differences in various factors were identified between the recovery and non-recovery groups. […] A poor initial hearing threshold of the affected ear has been demonstrated to be associated with a poor prognosis. […] Initial hearing threshold and post-treatment (2-weeks) hearing thresholds of the affected ear were also identified as significant prognostic factors of ISSHL in the DNN method. […] The DNN method showed the highest predictive power (accuracy: 88.81%, AUC=0.9448) and was considered the most useful for prognostic prediction.

• #2 Predicting Sudden Sensorineural Hearing Loss Recovery with Patient-Personalized Seigel’s Criteria Using Machine Learning

  https://pmc.ncbi.nlm.nih.gov/articles/PMC11202901/

  This shift underscores the importance of considering patient-specific impaired frequency ranges in hearing assessments. […] The Patient-Personalized Siegel criteria led to notable changes in ISSHL recovery classification in 41 patients (7%). […] These findings emphasize the need for personalized assessments to accurately determine the prognosis of patients with ISSHL. […] The soft voting classifier emerged as the most effective model, achieving the highest AUROC and precision scores among all the models tested in both validation and test evaluations. […] Factors that negatively impact prognosis were discovered based on the SHAP value, which included the following: average hearing threshold, dizziness, delay in treatment onset, and descending audiogram type. […] Our study underscores the significance of the ISSHL prognostic model, which accounts for patient-specific hearing impairments.

• #2 Predicting Sudden Sensorineural Hearing Loss Recovery with Patient-Personalized Seigel’s Criteria Using Machine Learning

  https://www.mdpi.com/2075-4418/14/12/1296

  Predicting Sudden Sensorineural Hearing Loss Recovery with Patient-Personalized Seigel’s Criteria Using Machine Learning […] Background: Accurate prognostic prediction is crucial for managing Idiopathic Sudden Sensorineural Hearing Loss (ISSHL). Previous studies developing ISSHL prognosis models often overlooked individual variability in hearing damage by relying on fixed frequency domains. This study aims to develop models predicting ISSHL prognosis one month after treatment, focusing on patient-specific hearing impairments. […] The main poor prognostic factors identified are severe hearing loss, high-frequency hearing loss, recovery starting duration greater than 2 weeks, advanced age, history of vertigo, and late initiation of treatment. […] Previous studies have developed machine learning models to accurately predict the prognosis of ISSHL. […] However, relying solely on these fixed-frequency domains to assess hearing recovery may overlook the nuances of patient-specific hearing impairment, potentially resulting in imprecise evaluations. […] The primary objective of this study is to develop machine learning models for predicting the prognosis of patients with ISSHL one month after treatment, focusing on hearing-impaired frequency domains. […] We elucidated the effects of clinical variables on the prediction of the model. […] The recovery status of ISSHL, which is the target of the machine learning model, was determined using Patient-Personalized Siegel’s criteria. […] The application of traditional Siegel’s criteria, focusing on the 0.5, 1, 2, and 3 kHz frequency domains, identified 334 patients (57.5%) as non-recovery and 247 patients (42.5%) as recovery. […] This shift underscores the importance of considering patient-specific impaired frequency ranges in hearing assessments. […] The soft voting classifier emerged as the most effective model, achieving the highest AUROC and precision scores among all the models tested in both validation and test evaluations. […] The distribution shift in the recovery assessment was not statistically significant based on the result of the chi-square test. […] Our study underscores the significance of the ISSHL prognostic model, which accounts for patient-specific hearing impairments. Further research is required to validate our findings in larger and more diverse patient populations.

• #2 Hearing recovery prediction and prognostic factors of idiopathic sudden sensorineural hearing loss: a retrospective analysis with a deep neural network model | Brazilian Journal of Otorhinolaryngology

  https://www.elsevier.es/en-revista-brazilian-journal-otorhinolaryngology-english-edition–497-articulo-hearing-recovery-prediction-prognostic-factors-S1808869423000356

  Various machine learning methods used to predict hearing recovery in ISSHL patients. […] The deep neural network method showed the highest predictive performance. […] Initial and early post-treatment hearing levels were significant prognostic factors. […] Machine learning may help predict ISSHL prognosis, as shown in the study. […] Idiopathic Sudden Sensorineural Hearing Loss (ISSHL) is an otologic emergency, and an early prediction of prognosis may facilitate proper treatment. […] Recovery was predicted by various machine learning models. […] The deep neural network model showed the highest predictive performance (accuracy, 88.81%; area under the receiver operating characteristic curve, 0.9448). […] Initial hearing level of affected and non-affected ear, post-treatment (2-weeks) hearing level of affected ear were significant prognostic factors.

• #2 Extended Preoperative Audiometry for Outcome Prediction and Risk Analysis in Patients Receiving Cochlear Implants

  https://www.mdpi.com/2077-0383/12/9/3262

  In clinical practice, preoperative diagnostics and anamnesis are used as a predictive basis for the postoperative result. […] For that test, Hoppe et al. have shown that unaided maximum word recognition (WRSmax) can serve as a predictor for the minimum expected benefit with a CI. […] Furthermore the monosyllabic speech comprehension at 65 dB with a hearing aid (WRS65(HA)) can be used as well. […] In their multicentre cohort, a positive correlation was found between the time elapsed since CI implantation and postoperative speech comprehension. […] The outcome was positively correlated with genetic aetiology and the presence of Ménière’s disease. […] An updated model with more influencing factors, including the pure-tone average, raised the explanatory power for postoperative speech comprehension to 22% within the same patient collective. […] These results demonstrate that it appears reasonable to include etiological aspects, on the basis of their correlation with speech comprehension. […] The model showed a positive correlation between the postoperative outcome and (i) preoperative speech comprehension with optimum hearing-aid fitting and (ii) age at onset of deafness. […] A negative correlation was found between speech comprehension and meningitis. […] Overall, the positive association between preoperative speech comprehension and postoperative outcome observed by other research groups was confirmed in this study. […] The predictive value is limited by the high proportion of patients with 0% word comprehension. […] The search for variables that can explain and predict interindividual differences in postoperative outcomes is a task for future research.

Zobacz więcej