JAJoong Ho Ahn

Joong Ho Ahn

*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


Joong Ho Ahn has a total of 95 co-authors. Here are the top co-authors by number of shared publications:

  • Jong Woo Chung

    Department of Otorhinolaryngology-Head and Neck Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea

    10 shared publications
  • Jong Woo Chung

    Department of Otorhinolaryngology - Head & Neck Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea

    9 shared publications
  • Hong Ju Park

    Department of Otorhinolaryngology - Head & Neck Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea

    8 shared publications
  • Hong Ju Park

    Department of Otorhinolaryngology-Head and Neck Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea

    8 shared publications
  • Woo Seok Kang

    Department of Otorhinolaryngology-Head and Neck Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea

    8 shared publications
  • Woo Seok Kang

    Department of Otorhinolaryngology - Head & Neck Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea

    6 shared publications

Latest Publications

  • Open Access

    Deep Learning–Assisted Diagnosis of Middle Ear Diseases in Otoendoscopy Across Clinician Experience Levels: Retrospective Reader Study

    by Se Eun Yi, Jihye Chae, Keunwoo Park, Yeon Joo Choi, Jihoon Kweon, Joong Ho Ahn

    Background: Accurate diagnosis of middle ear diseases remains challenging due to the specialized nature of otoscopic interpretation and the global shortage of otolaryngologists. Objective: We aimed to evaluate the impact of a deep learning–based AI support system on the diagnostic performance of clinicians with varying levels of expertise, specifically focusing on its ability to bridge the gap between trainees and experts. Methods: In this retrospective reader study, a deep learning model (EfficientNet-B4) was developed using a large-scale dataset of 7335 endoscopic images. The model was designed for multi-class classification, simultaneously identifying primary categories (chronic otitis media, otitis media with effusion, or none [indicating the absence of the first two]) and secondary features (attic retraction, myringitis, otomycosis, and ventilation tube). To assess clinical impact, diagnostic performance was compared among 3 otology professors, 5 senior residents, and 5 junior residents using an independent clinical assessment set of 100 images. Participants performed diagnoses both with and without AI-generated predictions, with a time constraint of 1 minute per image to simulate real-world clinical pressure. Results: In 5-fold cross-validation, the deep learning model achieved 94.8% accuracy for the primary classification. The stand-alone AI achieved diagnostic accuracies ranging from 94.0% to 99.0% in the clinical assessment set. AI support significantly enhanced diagnostic accuracy across all groups for primary classification, with improvements of 11.0% for otology professors, 28.8% for senior residents, and 32.2% for junior residents ( P <.001). Accuracy increased from 78.7% to 89.7% for otology professors, from 65.0% to 93.8% for senior residents, and from 54.4% to 86.6% for junior residents. The accuracy gap between the most experienced (otology professors) and least experienced (junior residents) groups was reduced from 24.3% to 3.1%. For secondary categories, significant improvements were observed in diagnosing attic retraction and myringitis among residents. AI-supported accuracy for attic retraction was higher in senior residents than that of the stand-alone AI, although the 95% CIs overlapped substantially, so the difference was not statistically significant. Conclusions: Our findings demonstrate that deep learning–based AI support substantially improves the diagnostic accuracy of middle ear diseases, particularly for less experienced clinicians. By narrowing the performance gap between trainees and experts, this system shows high potential as a valuable adjunct for clinical decision-making and medical education in otology.

  • Open Access

    Diagnostic Accuracy of a Deep Learning–Assisted System for Middle Ear Diseases Compared with Otolaryngologists: Prospective Observational Study (Preprint)

    by Se Eun Yi, Jihye Chae, Keunwoo Park, Yeon Joo Choi, Jihoon Kweon, Joong Ho Ahn

    BACKGROUND Accurate diagnosis of middle ear diseases remains challenging due to the specialized nature of otoscopic interpretation and the global shortage of otolaryngologists. OBJECTIVE We aimed to evaluate the impact of a deep learning-based artificial intelligence (AI) support system on the diagnostic performance of clinicians with varying levels of expertise, specifically focusing on its ability to bridge the gap between trainees and experts. METHODS A deep learning model (EfficientNet-B4) was developed using a large-scale dataset of 7,335 endoscopic images. The model was designed for multi-class classification, simultaneously identifying primary categories (Chronic Otitis Media, Otitis Media with Effusion, or Normal) and secondary features (attic retraction, myringitis, otomycosis, and ventilation tube). To assess clinical impact, diagnostic performance was compared among three otology professors (OP), five senior residents (SR), and five junior residents (JR) using a clinical assessment set of 100 images. Participants performed diagnoses both with and without AI-generated predictions, with a time constraint of one minute per image to simulate real-world clinical pressure. RESULTS AI support significantly enhanced diagnostic accuracy across all groups for primary classification, with improvements of 11.0% for OP, 28.8% for SR, and 32.2% for JR (p < 0.001). Notably, the accuracy gap between the most experienced (OP) and least experienced (JR) groups was reduced from 24.3% to 3.1%. For secondary categories, significant improvements were observed in diagnosing attic retraction and myringitis among residents, with some performances even surpassing the standalone AI accuracy. CONCLUSIONS Our findings demonstrate that deep learning-based AI support substantially improves the diagnostic accuracy of middle ear diseases, particularly for less experienced clinicians. By narrowing the performance gap between trainees and experts, this system shows high potential as a valuable adjunct for clinical decision-making and medical education in otology. CLINICALTRIAL This study complies with the Declaration of Helsinki and received research approval from the Institutional Review Board of the Asan Medical Center with a waiver of research consent (IRB no. 2021–0837).

  • Open Access

    CLUE: Controllable Latent space of Unprompted Embeddings for Diversity Management in Text-to-Image Synthesis

    by Keunwoo Park, Jihye Chae, Joong Ho Ahn, Jihoon Kweon

    Text-to-image synthesis models require the ability to generate diverse images while maintaining stability. To overcome this challenge, a number of methods have been proposed, including the collection of prompt-image datasets and the integration of additional data modalities during training. Although these methods have shown promising results in general domains, they face limitations when applied to specialized fields such as medicine, where only limited types and insufficient amounts of data are available. We present CLUE (Controllable Latent space of Unprompted Embeddings), a generative model framework that achieves diverse generation while maintaining stability through fixed-format prompts without requiring any additional data. Based on the Stable Diffusion architecture, CLUE employs a Style Encoder that processes images and prompts to generate style embeddings, which are subsequently fed into a new second attention layer of the U-Net architecture. Through Kullback-Leibler divergence, the latent space achieves continuous representation of image features within Gaussian regions, independent of prompts. Performance was assessed on otitis media dataset. CLUE reduced FID to 9.30 (vs. 46.81) and improved recall to 70.29% (vs. 49.60%). A classifier trained on synthetic-only data at 1000% scale achieved an F1 score of 83.21% (vs. 73.83%). Combining synthetic data with equal amounts of real data achieved an F1 score of 94.76%, higher than when using only real data. On an external dataset, synthetic-only training achieved an F1 score of 76.77% (vs. 60.61%) at 1000% scale. The combined approach achieved an F1 score of 85.78%, higher than when using only the internal dataset. These results demonstrate that CLUE enables diverse yet stable image generation from limited datasets and serves as an effective data augmentation method for domain-specific applications.

  • Open Access

    Balloon Eustachian Tuboplasty in Chronic Suppurative Otitis Media and Dilatory Eustachian Tube Dysfunction: A Randomized Controlled Trial

    by Hye Ah Joo, Sung-Min Park, Yehree Kim, Dong Kyu Lee, Yun Ji Lee, Yeonjoo Choi, Woo Seok Kang, Joong Ho Ahn, Jong Woo Chung, Won-Ho Chung, et al.

    Objectives. Balloon eustachian tuboplasty (BET) is an emerging treatment for dilatory eustachian tube (ET) dysfunction, yet its efficacy in patients with chronic suppurative otitis media (COM) combined with ET dysfunction remains unclear. This trial aimed to evaluate the efficacy and safety of BET versus medical management (MM) alone in adult patients with COM and chronic dilatory ET dysfunction, as defined by a failed Valsalva maneuver.Methods. This prospective, multicenter, randomized controlled trial enrolled 116 participants (121 ears) between January 2021 and June 2023. Participants were randomly assigned (1:1) to receive either BET combined with MM or MM alone (fluticasone furoate nasal steroid spray). The primary short-term outcome was the normalization of the Valsalva maneuver at 8 weeks, while secondary outcomes included improvements in Eustachian Tube Dysfunction Questionnaire-7 (ETDQ-7) scores and reductions in the air-bone gap (ABG) measured at 8 weeks. Adverse events were monitored in both groups throughout the follow-up period.Results. Of the 116 participants (121 ears), 60 participants (62 ears) received BET and 56 participants (59 ears) received MM alone. There were no significant differences in demographics, baseline characteristics, ETDQ-7 scores, bone conduction thresholds, or ABG between the groups. In the BET group, 46.8% of ears (29/62) achieved a successful Valsalva maneuver compared with 15.3% of ears (9/59) in the MM-only group (P<0.001). The BET group exhibited greater subjective symptom improvement, with an average ETDQ-7 score reduction of –6.2±9.4 compared to –2.6±8.6 in the MM-only group (P=0.028). Additionally, the ABG decreased more significantly in the BET group (–5.8±11.4 dB HL) than in the MM-only group (–1.2±10.5 dB HL) (P=0.023). No serious procedure- or device-related adverse events were observed during the 8-week follow-up.Conclusion. This trial demonstrates that BET combined with MM is superior to MM alone in treating dilatory ET dysfunction in patients with COM, while also maintaining a favorable safety profile.


  • Open Access

    Brain volumetric analysis of elderly patients based on hearing levels quantified using FreeSurfer

    by Yun Ji Lee, Eun-Jae Lee, Chong Hyun Suh, Hye Jung Youn, Jihye Chae, Keunwoo Park, Jihoon Kweon, Joong Ho Ahn

    Objective: This study investigated the effects of hearing loss on brain volume, particularly total gray matter, the cerebellum, and the limbic system, using MRI data analyzed with FreeSurfer software.Study Design: Data was obtained from 745 individuals aged 60 or older who underwent health screenings at Asan Medical Center between March and December 2019. Participants were categorized into three groups based on their pure-tone hearing thresholds: normal hearing (n = 328), mild hearing loss (20–40 dB, n = 336), and moderate-to-severe hearing loss (≥40 dB, n = 81).Results: Results showed significant differences among the groups in age, sex, and health conditions such as hypertension, diabetes, and cholesterol levels (p < 0.05). Increased hearing loss severity correlated with older age, male predominance, and higher prevalence of these conditions. Multiple regression analysis, controlling for confounders like age and comorbidities, revealed age-related brain atrophy across all groups. Notably, even mild hearing loss was associated with reduced total gray matter volume, while moderate-to-severe hearing loss led to significant hippocampal volume reduction.Conclusion: The findings suggest that hearing loss in individuals over the age of 60 may contribute to the atrophy of critical brain regions, particularly the hippocampus and total gray matter. The study emphasizes the importance of addressing even mild hearing loss, as active hearing rehabilitation could potentially prevent cortical volume loss and mitigate the risk of cognitive decline.


  • Open Access

    Impact of Hyperbaric Oxygen Therapy on Recovery in Patients With Sudden Sensorineural Hearing Loss: An Observational Study

    by Hye Ah Joo, So Hee Kang, Se Eun Lee, Yun Ji Lee, Min-Young Kwak, Woo Seok Kang, Joong Ho Ahn, Hong Ju Park, Jong Woo Chung

    Background and Objectives: We aimed to assess the efficacy of various treatments for patients experiencing sudden sensorineural hearing loss (SSNHL), with a focus on hyperbaric oxygen therapy (HBOT) and the effect of HBOT administration timing on hearing recovery.Subjects and Methods: In keeping with guideline recommendations, patients with SSNHL enrolled in this study received systemic corticosteroids, intratympanic dexamethasone, or HBOT in combination or as monotherapy. Hearing thresholds were determined using pure-tone audiometry (dB HL), and recovery was categorized as complete recovery (CR), partial recovery (PR), slight recovery (SR), or no recovery (NR). Post-treatment thresholds were assessed 2-3 months after symptom onset.Results: After treatment, we observed CR in 16 patients (15.2%), PR in 22 patients (21.0%), SR in 44 patients (42.0%), and NR in 23 patients (22.0%). Notably, among the 79 patients with severe hearing loss (>70 dB), those who received treatment within 2 weeks of onset showed better hearing outcomes than those who started treatment after 2 weeks. In the HBOT group, a shorter interval between symptom onset and HBOT initiation was associated with a higher likelihood of hearing improvement. Importantly, in patients with an initial hearing threshold of <70 dB, initiating HBOT within 2 weeks was associated with statistically significantly higher hearing recovery rates.Conclusions: The findings of our study support the premise that early initiation of HBOT within 2 weeks of symptom onset results in better hearing recovery in patients with SSNHL.


  • Open Access

    Temporal Bone Magnetic Resonance Imaging in Sudden Sensorineural Hearing Loss: Low Frequency Versus Other Types

    by Yun Ji Lee, Doyeon Kim, Joong Ho Ahn, Hong Ju Park, Jong Woo Chung, Woo Seok Kang

    Background: Gadolinium-enhanced magnetic resonance imaging (MRI) is the gold standard for diagnosing vestibular schwannoma (VS). Thisstudy aimed to compare the incidence of VS among patients with low-frequency, high-frequency, and flat-type sudden sensorineural hearingloss (SSNHL) and to assess the effectiveness of MRI for each SSNHL type.Methods: We analyzed 755 patients diagnosed with SSNHL at a single tertiary center between January 2014 and March 2020. All underwenttemporal bone MRI. Logistic regression was used to determine the associations of SSNHL types with VS. Additionally, we conducted a correlationanalysis to examine the relationship between word recognition scores (WRS) and the size of tumors measured on MRI at the time of diagnosis.Results: Magnetic resonance imaging identified VS in 23 of 755 cases (3.0%). Of the patients, 123 (16.3%) had low-frequency SSNHL; none ofthese patients had VS. The incidence of VS differed significantly between the low-frequency SSNHL group and other SSNHL types (P < .001). Asignificant negative correlation was found between initial tumor size and WRS (r2=0.2905, P=.008).Conclusion: The prevalence of VS was rare in patients with low-frequency SSNHL. The pattern of hearing loss may provide insights into thelikelihood of tumor presence, thereby aiding in more effective audiometric screening and decision-making regarding MRI for detecting VS.Cite this article as: Ji Lee Y, Kim D, Ho Ahn J, Ju Park H, Woo Chung J, Seok Kang W. Temporal bone magnetic resonance imaging in suddensensorineural hearing loss: Low frequency versus other types. J Int Adv Otol. 2025, 21(2), 1490, doi: 10.5152/iao.2025.241490.