Xun Niu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Xun Niu has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- Binjun Chen
ENT institute and Department of Otorhinolaryngology, Eye & ENT Hospital, Fudan University, Shanghai, China
2 shared publications - Bo Liang
Department of Radiology, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
2 shared publications - Dongdong Ren
Fudan University
2 shared publications - Fang-Lu Chi
Department of Otorhinolaryngology, Eye and ENT Hospital, ENT Institute, Fudan University, Shanghai, People’s Republic of China
2 shared publications - Haojie Sun
ENT institute and Department of Otorhinolaryngology, Eye & ENT Hospital, Fudan University, Shanghai, China
2 shared publications - Jiajie Guo
Huazhong University of Science and Technology
2 shared publications
Latest Publications
- Open Access
- 8 August 2024
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 26, e51706
A 3D and Explainable Artificial Intelligence Model for Evaluation of Chronic Otitis Media Based on Temporal Bone Computed Tomography: Model Development, Validation, and Clinical Application
Background: Temporal bone computed tomography (CT) helps diagnose chronic otitis media (COM). However, its interpretation requires training and expertise. Artificial intelligence (AI) can help clinicians evaluate COM through CT scans, but existing models lack transparency and may not fully leverage multidimensional diagnostic information. Objective: We aimed to develop an explainable AI system based on 3D convolutional neural networks (CNNs) for automatic CT-based evaluation of COM. Methods: Temporal bone CT scans were retrospectively obtained from patients operated for COM between December 2015 and July 2021 at 2 independent institutes. A region of interest encompassing the middle ear was automatically segmented, and 3D CNNs were subsequently trained to identify pathological ears and cholesteatoma. An ablation study was performed to refine model architecture. Benchmark tests were conducted against a baseline 2D model and 7 clinical experts. Model performance was measured through cross-validation and external validation. Heat maps, generated using Gradient-Weighted Class Activation Mapping, were used to highlight critical decision-making regions. Finally, the AI system was assessed with a prospective cohort to aid clinicians in preoperative COM assessment. Results: Internal and external data sets contained 1661 and 108 patients (3153 and 211 eligible ears), respectively. The 3D model exhibited decent performance with mean areas under the receiver operating characteristic curves of 0.96 (SD 0.01) and 0.93 (SD 0.01), and mean accuracies of 0.878 (SD 0.017) and 0.843 (SD 0.015), respectively, for detecting pathological ears on the 2 data sets. Similar outcomes were observed for cholesteatoma identification (mean area under the receiver operating characteristic curve 0.85, SD 0.03 and 0.83, SD 0.05; mean accuracies 0.783, SD 0.04 and 0.813, SD 0.033, respectively). The proposed 3D model achieved a commendable balance between performance and network size relative to alternative models. It significantly outperformed the 2D approach in detecting COM (P≤.05) and exhibited a substantial gain in identifying cholesteatoma (P<.001). The model also demonstrated superior diagnostic capabilities over resident fellows and the attending otologist (P<.05), rivaling all senior clinicians in both tasks. The generated heat maps properly highlighted the middle ear and mastoid regions, aligning with human knowledge in interpreting temporal bone CT. The resulting AI system achieved an accuracy of 81.8% in generating preoperative diagnoses for 121 patients and contributed to clinical decision-making in 90.1% cases. Conclusions: We present a 3D CNN model trained to detect pathological changes and identify cholesteatoma via temporal bone CT scans. In both tasks, this model significantly outperforms the baseline 2D approach, achieving levels comparable with or surpassing those of human experts. The model also exhibits decent generalizability and enhanced comprehensibility. This AI system facilitates automatic COM assessment and shows promising viability in real-world clinical settings. These findings underscore AI’s potential as a valuable aid for clinicians in COM evaluation. Trial Registration: Chinese Clinical Trial Registry ChiCTR2000036300; https://www.chictr.org.cn/showprojEN.html?proj=58685
- 8 August 2024
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 26, e51706
Open AccessA 3D and Explainable Artificial Intelligence Model for Evaluation of Chronic Otitis Media Based on Temporal Bone Computed Tomography: Model Development, Validation, and Clinical Application
Background: Temporal bone computed tomography (CT) helps diagnose chronic otitis media (COM). However, its interpretation requires training and expertise. Artificial intelligence (AI) can help clinicians evaluate COM through CT scans, but existing models lack transparency and may not fully leverage multidimensional diagnostic information. Objective: We aimed to develop an explainable AI system based on 3D convolutional neural networks (CNNs) for automatic CT-based evaluation of COM. Methods: Temporal bone CT scans were retrospectively obtained from patients operated for COM between December 2015 and July 2021 at 2 independent institutes. A region of interest encompassing the middle ear was automatically segmented, and 3D CNNs were subsequently trained to identify pathological ears and cholesteatoma. An ablation study was performed to refine model architecture. Benchmark tests were conducted against a baseline 2D model and 7 clinical experts. Model performance was measured through cross-validation and external validation. Heat maps, generated using Gradient-Weighted Class Activation Mapping, were used to highlight critical decision-making regions. Finally, the AI system was assessed with a prospective cohort to aid clinicians in preoperative COM assessment. Results: Internal and external data sets contained 1661 and 108 patients (3153 and 211 eligible ears), respectively. The 3D model exhibited decent performance with mean areas under the receiver operating characteristic curves of 0.96 (SD 0.01) and 0.93 (SD 0.01), and mean accuracies of 0.878 (SD 0.017) and 0.843 (SD 0.015), respectively, for detecting pathological ears on the 2 data sets. Similar outcomes were observed for cholesteatoma identification (mean area under the receiver operating characteristic curve 0.85, SD 0.03 and 0.83, SD 0.05; mean accuracies 0.783, SD 0.04 and 0.813, SD 0.033, respectively). The proposed 3D model achieved a commendable balance between performance and network size relative to alternative models. It significantly outperformed the 2D approach in detecting COM (P≤.05) and exhibited a substantial gain in identifying cholesteatoma (P<.001). The model also demonstrated superior diagnostic capabilities over resident fellows and the attending otologist (P<.05), rivaling all senior clinicians in both tasks. The generated heat maps properly highlighted the middle ear and mastoid regions, aligning with human knowledge in interpreting temporal bone CT. The resulting AI system achieved an accuracy of 81.8% in generating preoperative diagnoses for 121 patients and contributed to clinical decision-making in 90.1% cases. Conclusions: We present a 3D CNN model trained to detect pathological changes and identify cholesteatoma via temporal bone CT scans. In both tasks, this model significantly outperforms the baseline 2D approach, achieving levels comparable with or surpassing those of human experts. The model also exhibits decent generalizability and enhanced comprehensibility. This AI system facilitates automatic COM assessment and shows promising viability in real-world clinical settings. These findings underscore AI’s potential as a valuable aid for clinicians in COM evaluation. Trial Registration: Chinese Clinical Trial Registry ChiCTR2000036300; https://www.chictr.org.cn/showprojEN.html?proj=58685
- 16 September 2023
- Published by Wiley in International Forum of Allergy & Rhinology
- Vol. 14 (3) , 679-694
Endoscopic vidian and vidian‐branch neurectomy for refractory allergic rhinitis: A systematic review
Background: Endoscopic vidian neurectomy is expected to provide good therapeutic relief in patients with allergic rhinitis (AR) being refractory to medication therapy or conservative surgery. However, the evidence bases for its benefit remain debatable. In this study, we conducted a systematic review and meta‐analysis to clarify the therapeutic role of various forms of vidian neurectomy in refractory AR.Method: Preferred Reporting Items for Systematic Reviews and Meta‐analyses guidelines were used to conduct a systematic review of primary studies that reported original patient data for endoscopic vidian neurectomy (EVN) and vidian‐branch neurectomy, which includes selective vidian neurectomy (SVN) and posterior nasal neurectomy (PNN). The primary outcome was patient‐reported outcome measures (PROMs), including the Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ) and Visual Analog Scale (VAS), to assess an improvement in nasal symptom severity and quality of patient's life. The incidence of surgical complications and other objective outcomes were considered secondary outcomes.Results: This review included 24 clinical studies involving 1677 patients with refractory AR, of which 510 patients in six studies had combined chronic rhinosinusitis with nasal polyps (CRSwNP) and 95 patients in one study had combined asthma. Postoperative PROMs were significantly better than preoperatively in almost all patients who underwent vidianp (RQLQ: standardized mean difference [SMD] = 2.66, 95% confidence interval [CI] = 2.40–2.92, p < 0.001; VAS: SMD = 5.15, 95% CI = 4.29–6.02, p < 0.001) or vidian‐branch neurectomy (RQLQ in PNN: SMD = 3.29, 95% CI = 2.45–4.13, p < 0.001; VAS in PNN: SMD = 4.38, 95% CI = 3.41–5.34, p < 0.001), and were generally better than in the conservative treatment group. Dividing with 18 months as the cutoff point, a subgroup analysis of the follow‐up period was conducted, and the results showed that both long‐term and short‐term postoperative patients had considerably reduced symptoms compared to the preoperative period. The two surgical procedures, SVN and PNN, attributed to vidian‐branch neurectomy have extremely few complications. However, EVN is more likely to cause dry eyes and palatal numbness, with no other serious complications. In patients with AR and CRSwNP, vidian or selective vidian neurectomy combined with functional endoscopic sinus surgery (FESS) is more effective than conventional FESS (RQLQ: SMD = 2.17, 95% CI = 1.66–2.69, p < 0.001; VAS: SMD = 6.42, 95% CI = 4.78–8.06, p < 0.001). For patients who have both AR and asthma, SVN with pharyngeal branch excision is a potential treatment option.Conclusion: EVN and vidian‐branch neurectomy (including SVN and PNN) are effective treatments, but the former has a higher risk of complications. Additionally, vidian‐branch neurectomy with FESS is beneficial for patients with mixed CRSwNP. SVN is a potential approach for patients with coexisting AR and asthma.
- 16 September 2023
- Published by Wiley in International Forum of Allergy & Rhinology
- Vol. 14 (3) , 679-694
Endoscopic vidian and vidian‐branch neurectomy for refractory allergic rhinitis: A systematic review
Background: Endoscopic vidian neurectomy is expected to provide good therapeutic relief in patients with allergic rhinitis (AR) being refractory to medication therapy or conservative surgery. However, the evidence bases for its benefit remain debatable. In this study, we conducted a systematic review and meta‐analysis to clarify the therapeutic role of various forms of vidian neurectomy in refractory AR.Method: Preferred Reporting Items for Systematic Reviews and Meta‐analyses guidelines were used to conduct a systematic review of primary studies that reported original patient data for endoscopic vidian neurectomy (EVN) and vidian‐branch neurectomy, which includes selective vidian neurectomy (SVN) and posterior nasal neurectomy (PNN). The primary outcome was patient‐reported outcome measures (PROMs), including the Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ) and Visual Analog Scale (VAS), to assess an improvement in nasal symptom severity and quality of patient's life. The incidence of surgical complications and other objective outcomes were considered secondary outcomes.Results: This review included 24 clinical studies involving 1677 patients with refractory AR, of which 510 patients in six studies had combined chronic rhinosinusitis with nasal polyps (CRSwNP) and 95 patients in one study had combined asthma. Postoperative PROMs were significantly better than preoperatively in almost all patients who underwent vidianp (RQLQ: standardized mean difference [SMD] = 2.66, 95% confidence interval [CI] = 2.40–2.92, p < 0.001; VAS: SMD = 5.15, 95% CI = 4.29–6.02, p < 0.001) or vidian‐branch neurectomy (RQLQ in PNN: SMD = 3.29, 95% CI = 2.45–4.13, p < 0.001; VAS in PNN: SMD = 4.38, 95% CI = 3.41–5.34, p < 0.001), and were generally better than in the conservative treatment group. Dividing with 18 months as the cutoff point, a subgroup analysis of the follow‐up period was conducted, and the results showed that both long‐term and short‐term postoperative patients had considerably reduced symptoms compared to the preoperative period. The two surgical procedures, SVN and PNN, attributed to vidian‐branch neurectomy have extremely few complications. However, EVN is more likely to cause dry eyes and palatal numbness, with no other serious complications. In patients with AR and CRSwNP, vidian or selective vidian neurectomy combined with functional endoscopic sinus surgery (FESS) is more effective than conventional FESS (RQLQ: SMD = 2.17, 95% CI = 1.66–2.69, p < 0.001; VAS: SMD = 6.42, 95% CI = 4.78–8.06, p < 0.001). For patients who have both AR and asthma, SVN with pharyngeal branch excision is a potential treatment option.Conclusion: EVN and vidian‐branch neurectomy (including SVN and PNN) are effective treatments, but the former has a higher risk of complications. Additionally, vidian‐branch neurectomy with FESS is beneficial for patients with mixed CRSwNP. SVN is a potential approach for patients with coexisting AR and asthma.
- Open Access
- 9 August 2023
- Published by JMIR Publications Inc. in JMIR Preprints
A 3D and Explainable Artificial Intelligence Model for Evaluation of Chronic Otitis Media Based on Temporal Bone Computed Tomography: Model Development, Validation, and Clinical Application (Preprint)
BACKGROUND Temporal bone computed tomography (CT) helps diagnose chronic otitis media (COM). However, its interpretation requires training and expertise. Artificial intelligence (AI) can help clinicians evaluate COM through CT scans, but existing models lack transparency and may not fully leverage multidimensional diagnostic information. OBJECTIVE We aimed to develop an explainable AI system based on 3D convolutional neural networks (CNNs) for automatic CT-based evaluation of COM. METHODS Temporal bone CT scans were retrospectively obtained from patients operated for COM between December 2015 and July 2021 at 2 independent institutes. A region of interest encompassing the middle ear was automatically segmented, and 3D CNNs were subsequently trained to identify pathological ears and cholesteatoma. An ablation study was performed to refine model architecture. Benchmark tests were conducted against a baseline 2D model and 7 clinical experts. Model performance was measured through cross-validation and external validation. Heat maps, generated using Gradient-Weighted Class Activation Mapping, were used to highlight critical decision-making regions. Finally, the AI system was assessed with a prospective cohort to aid clinicians in preoperative COM assessment. RESULTS Internal and external data sets contained 1661 and 108 patients (3153 and 211 eligible ears), respectively. The 3D model exhibited decent performance with mean areas under the receiver operating characteristic curves of 0.96 (SD 0.01) and 0.93 (SD 0.01), and mean accuracies of 0.878 (SD 0.017) and 0.843 (SD 0.015), respectively, for detecting pathological ears on the 2 data sets. Similar outcomes were observed for cholesteatoma identification (mean area under the receiver operating characteristic curve 0.85, SD 0.03 and 0.83, SD 0.05; mean accuracies 0.783, SD 0.04 and 0.813, SD 0.033, respectively). The proposed 3D model achieved a commendable balance between performance and network size relative to alternative models. It significantly outperformed the 2D approach in detecting COM (P≤.05) and exhibited a substantial gain in identifying cholesteatoma (P<.001). The model also demonstrated superior diagnostic capabilities over resident fellows and the attending otologist (P<.05), rivaling all senior clinicians in both tasks. The generated heat maps properly highlighted the middle ear and mastoid regions, aligning with human knowledge in interpreting temporal bone CT. The resulting AI system achieved an accuracy of 81.8% in generating preoperative diagnoses for 121 patients and contributed to clinical decision-making in 90.1% cases. CONCLUSIONS We present a 3D CNN model trained to detect pathological changes and identify cholesteatoma via temporal bone CT scans. In both tasks, this model significantly outperforms the baseline 2D approach, achieving levels comparable with or surpassing those of human experts. The model also exhibits decent generalizability and enhanced comprehensibility. This AI system facilitates automatic COM assessment and shows promising viability in real-world clinical settings. These findings underscore AI’s potential as a valuable aid for clinicians in COM evaluation. CLINICALTRIAL Chinese Clinical Trial Registry ChiCTR2000036300; https://www.chictr.org.cn/showprojEN.html?proj=58685
- 9 August 2023
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessA 3D and Explainable Artificial Intelligence Model for Evaluation of Chronic Otitis Media Based on Temporal Bone Computed Tomography: Model Development, Validation, and Clinical Application (Preprint)
BACKGROUND Temporal bone computed tomography (CT) helps diagnose chronic otitis media (COM). However, its interpretation requires training and expertise. Artificial intelligence (AI) can help clinicians evaluate COM through CT scans, but existing models lack transparency and may not fully leverage multidimensional diagnostic information. OBJECTIVE We aimed to develop an explainable AI system based on 3D convolutional neural networks (CNNs) for automatic CT-based evaluation of COM. METHODS Temporal bone CT scans were retrospectively obtained from patients operated for COM between December 2015 and July 2021 at 2 independent institutes. A region of interest encompassing the middle ear was automatically segmented, and 3D CNNs were subsequently trained to identify pathological ears and cholesteatoma. An ablation study was performed to refine model architecture. Benchmark tests were conducted against a baseline 2D model and 7 clinical experts. Model performance was measured through cross-validation and external validation. Heat maps, generated using Gradient-Weighted Class Activation Mapping, were used to highlight critical decision-making regions. Finally, the AI system was assessed with a prospective cohort to aid clinicians in preoperative COM assessment. RESULTS Internal and external data sets contained 1661 and 108 patients (3153 and 211 eligible ears), respectively. The 3D model exhibited decent performance with mean areas under the receiver operating characteristic curves of 0.96 (SD 0.01) and 0.93 (SD 0.01), and mean accuracies of 0.878 (SD 0.017) and 0.843 (SD 0.015), respectively, for detecting pathological ears on the 2 data sets. Similar outcomes were observed for cholesteatoma identification (mean area under the receiver operating characteristic curve 0.85, SD 0.03 and 0.83, SD 0.05; mean accuracies 0.783, SD 0.04 and 0.813, SD 0.033, respectively). The proposed 3D model achieved a commendable balance between performance and network size relative to alternative models. It significantly outperformed the 2D approach in detecting COM (P≤.05) and exhibited a substantial gain in identifying cholesteatoma (P<.001). The model also demonstrated superior diagnostic capabilities over resident fellows and the attending otologist (P<.05), rivaling all senior clinicians in both tasks. The generated heat maps properly highlighted the middle ear and mastoid regions, aligning with human knowledge in interpreting temporal bone CT. The resulting AI system achieved an accuracy of 81.8% in generating preoperative diagnoses for 121 patients and contributed to clinical decision-making in 90.1% cases. CONCLUSIONS We present a 3D CNN model trained to detect pathological changes and identify cholesteatoma via temporal bone CT scans. In both tasks, this model significantly outperforms the baseline 2D approach, achieving levels comparable with or surpassing those of human experts. The model also exhibits decent generalizability and enhanced comprehensibility. This AI system facilitates automatic COM assessment and shows promising viability in real-world clinical settings. These findings underscore AI’s potential as a valuable aid for clinicians in COM evaluation. CLINICALTRIAL Chinese Clinical Trial Registry ChiCTR2000036300; https://www.chictr.org.cn/showprojEN.html?proj=58685
- 19 January 2023
- Published by Springer Nature in European Archives of Oto-Rhino-Laryngology
- Vol. 280 (5) , 2091-2097
The relationship between serum homocysteine levels and sudden sensorineural hearing loss: a meta-analysis
Objective There was disagreement over the association between serum/plasma homocysteine (HCY) levels and sudden sensorineural hearing loss (SSNHL). Through the use of a meta-analysis, this study aims to determine whether there is a significant difference in serum homocysteine levels between the SSNHL group and the control group. Design The Cochrane Library, EMBASE, and PubMed databases were all thoroughly searched. The two independent reviewers thoroughly examined the initially searched articles. The data results were calculated by standard mean difference (SMD) or odds ratios (OR). Review Manager (version 5.3) was applied to statistical data. Study sample There were 766 participants in the 6 trials with continuous outcomes that were part of the meta-analysis A. In addition, meta-analysis B, which included 961 people, contained a total of 3 studies with dichotomous results. Results Both meta-analyses revealed the same conclusion that serum/plasma HCY levels in the SSNHL patients are higher than those in the controls (SMD 0.41, 95 % confidence interval (CI) 0.11 to 0.72, P < 0.01; OR 3.27, 95 % CI 2.16 to 4.94, P < 0.01). Conclusion This study demonstrated that the SSNHL patients' serum/plasma HCY levels were greater than those of the control group.
- 19 January 2023
- Published by Springer Nature in European Archives of Oto-Rhino-Laryngology
- Vol. 280 (5) , 2091-2097
The relationship between serum homocysteine levels and sudden sensorineural hearing loss: a meta-analysis
Objective There was disagreement over the association between serum/plasma homocysteine (HCY) levels and sudden sensorineural hearing loss (SSNHL). Through the use of a meta-analysis, this study aims to determine whether there is a significant difference in serum homocysteine levels between the SSNHL group and the control group. Design The Cochrane Library, EMBASE, and PubMed databases were all thoroughly searched. The two independent reviewers thoroughly examined the initially searched articles. The data results were calculated by standard mean difference (SMD) or odds ratios (OR). Review Manager (version 5.3) was applied to statistical data. Study sample There were 766 participants in the 6 trials with continuous outcomes that were part of the meta-analysis A. In addition, meta-analysis B, which included 961 people, contained a total of 3 studies with dichotomous results. Results Both meta-analyses revealed the same conclusion that serum/plasma HCY levels in the SSNHL patients are higher than those in the controls (SMD 0.41, 95 % confidence interval (CI) 0.11 to 0.72, P < 0.01; OR 3.27, 95 % CI 2.16 to 4.94, P < 0.01). Conclusion This study demonstrated that the SSNHL patients' serum/plasma HCY levels were greater than those of the control group.
- 28 December 2022
- Published by Springer Nature in European Archives of Oto-Rhino-Laryngology
- Vol. 280 (5) , 2065-2072
Can polysomnography be replaced by ApneaGraph200 for OSA diagnosis? A protocol for systematic review and meta-analysis
Background It is yet unknown if the ApneaGraph200 system can replace the Polysomnography system (PSG), which is now the gold standard for obstructive sleep apnea (OSA) diagnosis. We performed a meta-analysis to evaluate the difference in apnea hypopnea index and lowest oxygen saturation levels between the AG200 system and PSG system. Methods Utilizing PubMed, Elsevier, Wanfang, and China National Knowledge Internet (CNKI), a thorough literature search was carried out. Eligible studies contrasting the AG200 system and the PSG system were found by two separate researchers. Review Manager (version 5.3) was adopted for data synthesis. Results The meta-analysis A pooled the comparison of AHI between the AG and PSG groups, and included seven studies involving a total of 419 participants, which revealed that there was a significant difference in AHI between the above two diagnostic methods (standard mean difference (SMD) = − 0.19, 95% confidence interval (CI) = − 0.32 to − 0.05, P = 0.008). As a supplementary, meta-analysis B pooled the comparison of lowest oxygen saturation between the AG and PSG groups and included five studies involving a total of 224 participants. It revealed that there was no significant difference between the AG group and the PSG group (SMD = 0.09, 95% CI − 0.24 to 0.43, P = 0.58). Conclusion The ApneaGraph200 system can be used for the initial screening of OSA patients, but it is not a complete alternative to polysomnography for the diagnosis and severity of the OSA condition.
- 28 December 2022
- Published by Springer Nature in European Archives of Oto-Rhino-Laryngology
- Vol. 280 (5) , 2065-2072
Can polysomnography be replaced by ApneaGraph200 for OSA diagnosis? A protocol for systematic review and meta-analysis
Background It is yet unknown if the ApneaGraph200 system can replace the Polysomnography system (PSG), which is now the gold standard for obstructive sleep apnea (OSA) diagnosis. We performed a meta-analysis to evaluate the difference in apnea hypopnea index and lowest oxygen saturation levels between the AG200 system and PSG system. Methods Utilizing PubMed, Elsevier, Wanfang, and China National Knowledge Internet (CNKI), a thorough literature search was carried out. Eligible studies contrasting the AG200 system and the PSG system were found by two separate researchers. Review Manager (version 5.3) was adopted for data synthesis. Results The meta-analysis A pooled the comparison of AHI between the AG and PSG groups, and included seven studies involving a total of 419 participants, which revealed that there was a significant difference in AHI between the above two diagnostic methods (standard mean difference (SMD) = − 0.19, 95% confidence interval (CI) = − 0.32 to − 0.05, P = 0.008). As a supplementary, meta-analysis B pooled the comparison of lowest oxygen saturation between the AG and PSG groups and included five studies involving a total of 224 participants. It revealed that there was no significant difference between the AG group and the PSG group (SMD = 0.09, 95% CI − 0.24 to 0.43, P = 0.58). Conclusion The ApneaGraph200 system can be used for the initial screening of OSA patients, but it is not a complete alternative to polysomnography for the diagnosis and severity of the OSA condition.