Background: Tinnitus diagnosis poses a challenge in otolaryngology owing to an extremely complex pathogenesis, lack of effective objectification methods, and factor-affected diagnosis. There is currently a lack of explainable auxiliary diagnostic tools for tinnitus in clinical practice. Objective: This study aims to develop a diagnostic model using an explainable artificial intelligence (AI) method to address the issue of low accuracy in tinnitus diagnosis. Methods: In this study, a knowledge graph–based tinnitus diagnostic method was developed by combining clinical medical knowledge with electronic medical records. Electronic medical record data from 1267 patients were integrated with traditional Chinese clinical medical knowledge to construct a tinnitus knowledge graph. Subsequently, weights were introduced, which measured patient similarity in the knowledge graph based on mutual information values. Finally, a collaborative neighbor algorithm was proposed, which scored patient similarity to obtain the recommended diagnosis. We conducted 2 group experiments and 1 case derivation to explore the effectiveness of our models and compared the models with state-of-the-art graph algorithms and other explainable machine learning models. Results: The experimental results indicate that the method achieved 99.4% accuracy, 98.5% sensitivity, 99.6% specificity, 98.7% precision, 98.6% F1-score, and 99% area under the receiver operating characteristic curve for the inference of 5 tinnitus subtypes among 253 test patients. Additionally, it demonstrated good interpretability. The topological structure of knowledge graphs provides transparency that can explain the reasons for the similarity between patients. Conclusions: This method provides doctors with a reliable and explainable diagnostic tool that is expected to improve tinnitus diagnosis accuracy.

Objective: To provide evidence for medical management of tinnitus based on an assessment of the evidence concerning the effectiveness of acupuncture as a treatment for tinnitus using network meta-analysis (NMA). Methods: We conducted a systematic literature review by searching 8 national and international databases (inception to February 2023) for randomized controlled trials (RCTs) for tinnitus. Only RCTs that recruited participants aged over 18 and diagnosed with tinnitus, and that evaluated acupuncture or acupuncture in combination with conventional western medical therapy were included. We used response rate and tinnitus handicap inventory (THI) to examine efficacy. We conducted NMA with random effects, and the rate ratio or mean difference with its 95% credible interval was calculated. In addition, we ranked all treatments via their SUCRA and assessed the quality of evidence according to the GRADE criteria. Results: A total of 2575 patients were included in the study. The main findings of the current NMA were that acupoint injection combined with warm acupuncture was the most effective for response rate, followed by warm acupuncture and acupoint injection combined with western medical treatment. Acupuncture combined with western medical treatment was the most effective for THI, followed by electroacupuncture combined with warm acupuncture and acupuncture combined with moxibustion. Conclusion: Acupuncture seems to be a better trend treatment for tinnitus. Further rigorous RCT studies that include direct comparisons for different acupuncture-related treatments are encouraged to provide the most promising evidence for patients with tinnitus. Protocol registration: CRD42023398745.

BACKGROUND Tinnitus diagnosis poses a challenge in otolaryngology owing to an extremely complex pathogenesis, lack of effective objectification methods, and factor-affected diagnosis. There is currently a lack of explainable auxiliary diagnostic tools for tinnitus in clinical practice. OBJECTIVE This study aims to develop a diagnostic model using an explainable artificial intelligence (AI) method to address the issue of low accuracy in tinnitus diagnosis. METHODS In this study, a knowledge graph–based tinnitus diagnostic method was developed by combining clinical medical knowledge with electronic medical records. Electronic medical record data from 1267 patients were integrated with traditional Chinese clinical medical knowledge to construct a tinnitus knowledge graph. Subsequently, weights were introduced, which measured patient similarity in the knowledge graph based on mutual information values. Finally, a collaborative neighbor algorithm was proposed, which scored patient similarity to obtain the recommended diagnosis. We conducted 2 group experiments and 1 case derivation to explore the effectiveness of our models and compared the models with state-of-the-art graph algorithms and other explainable machine learning models. RESULTS The experimental results indicate that the method achieved 99.4% accuracy, 98.5% sensitivity, 99.6% specificity, 98.7% precision, 98.6% F1-score, and 99% area under the receiver operating characteristic curve for the inference of 5 tinnitus subtypes among 253 test patients. Additionally, it demonstrated good interpretability. The topological structure of knowledge graphs provides transparency that can explain the reasons for the similarity between patients. CONCLUSIONS This method provides doctors with a reliable and explainable diagnostic tool that is expected to improve tinnitus diagnosis accuracy.

The purpose of this study was to investigate the effect of Yupingfeng granules, a well-known traditional Chinese medication, combined with glucocorticoids on eosinophil (EOS) number and inflammatory factors in nasal secretions of patients with allergic rhinitis (AR). 76 patients with AR were randomly divided into controls (nasal inhaled corticosteroids alone) and experimental group (combined with Yupingfeng granules orally). The efficacy, symptoms, number of EOS and levels of inflammatory factors in nasal secretions were compared between the two groups after 2 months of treatment. The results showed that in the experimental group (94.73%, 0.48 ± 0.25 points), the effective rate of treatment was obviously higher and the physical sign score was clearly lower compared with controls (81.57%, 0.97 ± 0.36 points); the EOS count in the experimental group was clearly less than that in the controls; the tumor necrosis factor-α (TNF-α) (0.76 ± 0.44 ng/mL), interleukin-6 (IL-6) (11.83 ± 3.72 pg/mL), and interleukin-8 (IL-8) (17.45 ± 6.53 pg/mL) levels were evidently lower in the experimental group than in the controls (1.48 ± 0.72 ng/mL, 14.62 ± 5.19 pg/mL, 22.86 ± 9.35 pg/mL) (P < 0.05). In summary, compared with nasal inhalation of glucocorticoids alone, combined with oral administration of Yupingfeng granules, it can more significantly improve the symptoms of nasal obstruction and rhinorrhea in AR patients, reduce nasal mucosal edema, reduce the number of EOS in nasal secretions, and reduce the level of inflammatory factors, which is of positive significance for the clinical treatment of AR patients.