DEVELOPMENT OF A MACHINE LEARNING MODEL FOR DIFFERENTIAL DIAGNOSIS OF SYNCOPAL AND NON-SYNCOPAL TRANSIENT LOSS OF CONSCIOUSNESS IN CHILDREN
DOI:
https://doi.org/10.11603/mie.1996-1960.2023.3-4.14471Keywords:
syncope, non-syncopal transient loss of consciousness, differential diagnosis, XGBoost model, childrenAbstract
Background. The occurrence of a transient loss of consciousness (TLOC) episode in childhood presents a complex diagnostic challenge for clinicians. This study aimed to develop an effective machine learning model for the differential diagnosis of syncopal and non-syncopal TLOC, providing pediatricians, pediatric cardiologists, rheumatologists, and neurologists with high diagnostic accuracy.
Materials and methods. A cohort of 140 patients aged 8-17 years with syncope and 58 with non-syncopal TLOC were examined. The XGBoost algorithm was employed to construct the machine learning model.
Results. The proposed machine learning model demonstrated high effectiveness, evidenced by metrics with 95 % confidence interval: accuracy (0.90; CI 0.80-0.99), precision (syncope: 0.92, CI 0.85-1.0; non-syncope: 0.83, CI 0.72-0.95), recall (syncope: 0.92, CI 0.85-1.0; non-syncope: 0.83, CI 0.72-0.95), f1-score (syncope: 0.92, CI 0.85-1.0; non-syncope: 0.83, CI 0.72-0.95), specificity (syncope: 0.83, CI 0.72-0.95; non-syncope:
0.92, CI 0.85-1.0), ROC AUC (0.96, CI 0.90-1.00) and PR AUC (0.86, CI 0.75-0.96). The most informative indicators of the model are the Modified Calgary Syncope Seizure Score, the rate of morning systolic blood pressure increase, cardiac index, rate of morning diastolic blood pressure increase, frequency of tachycardia episodes during 24-hour Holter monitoring, nocturnal diastolic blood pressure decrease, total peripheral vascular resistance, child's age, diastolic blood pressure variability, pNN50, LF/HF ratio, and percentile distribution relative to height, diastolic blood pressure, and body mass index.
Conclusions. The application of this machine learning model enables the differentiation between syncopal and non-syncopal TLOC in children. It can serve as an additional diagnostic tool for pediatricians, pediat-ric cardiologists, rheumatologists, and neurologists, complementing traditional diagnostic criteria for syncope (vasovagal syncope, syncope due to orthostatic hypotension and cardiac syncope) and non-syncopal TLOC (epilepsy, first unprovoked epileptic seizures, psychogenic pseudo-syncope, and psychogenic nonepileptic seizure) in the initial diagnostic phase.
References
Brignole, M., Moya, A., de Lange, F. J. et al. (2018). ESC Guidelines for the diagnosis and management of syncope. Eur Heart J., 39(21), 1883948. doi: 10.1093/eurheartj/ehy037.
Goldberger, Z. D., Petek, B. J., Brignole, M. et al. (2019). ACC/AHA/HRS Versus ESC Guidelines for the Diagnosis and Management of Syncope: JACC Guideline Comparison. J Am Coll Cardiol., 74 (19), 2410-2423. doi: 10.1016/j.jacc.2019.09.012.
Wardrope, A., Newberry, E., Reuber, M. (2018). Diagnostic criteria to aid the differential diagnosis of patients presenting with transient loss of consciousness: A systematic review. Seizure, 61, 139148. doi: 10.1016/j.seizure.2018.08.012.
Brody, E. I., Genuini, M., Auvin, S., Lode, N., Brunet, S. R. (2022). Prehospital capillary lactate in children differentiates epileptic seizure from febrile seizure, syncope, and psychogenic nonepileptic seizure. Epilepsy Behav., 127. doi: 10.1016/j.yebeh.2021.108551.
Leibetseder, A., Eisermann, M., LaFrance, W. C. Jr., Nobili, L., von Oertzen, T. J. (2020). How to distinguish seizures from non-epileptic manifestations. Epileptic Disord., 22 (6), 716-738. doi: 10.1684/epd.2020.1234.
Villafane, J., Miller, J. R., Glickstein, J. et al. (2021). Loss of Consciousness in the Young Child. Pediatr Cardiol., 42 (2), 234-254. doi: 10.1007/s00246-020-02498-6.
Masoumi, B., Mozafari, S., Golshani, K., Heydari, F., Nasr-Esfahani, M. (2022). Differential Diagnosis of Seizure and Syncope by the Means of Biochemical Markers in Emergency Department Patients. Int J Prev Med., 13, 58. doi: 10.4103/ijpvm.IJPVM_129_20.
Liao, Y., Du, J., Benditt, D. G., Jin, H. (2022). Vasovagal syncope or psychogenic pseudosyncope: a major issue in the differential diagnosis of apparent transient loss of consciousness in children. Sci Bull (Beijing), 67 (16), 1618-1620. doi: 10.1016/j.scib.2022.07.024.
Chen, M., Jamnadas-Khoda, J., Broadhurst, M. et al. (2019). Value of witness observations in the differential diagnosis of transient loss of consciousness. Neurology, 92 (9), e895-e904. doi: 10.1212/WNL.0000000000007017.
Talibi, S., Douglas, C., Pope, B. (2020). Cardiac Syncope with Anoxic Seizure Activity. Case Rep Emerg Med., 8. doi: 10.1155/2020/6749382.
Rivolta, I., Binda, A., Masi, A., DiFrancesco, J. C. (2020). Cardiac and neuronal HCN channelopathies. Pflugers Arch., 472 (7), 931-951. doi: 10.1007/s00424-020-02384-3.
Yu, C., Deng, X. J., Xu, D. (2023). Gene mutations in comorbidity of epilepsy and arrhythmia. J Neurol., 270 (3), 1229-1248. doi: 10.1007/s00415-022-11430-2.
Costagliola, G., Orsini, A., Coll, M., Brugada, R., Parisi, P., Striano, P. (). The brain-heart interaction in epilepsy: implications for diagnosis, therapy, and SUDEP prevention. Ann Clin Transl Neurol., 8 (7), 1557-1568. doi: 10.1002/acn3.51382.
Fisher, R. S., Acevedo, C., Arzimanoglou, A. et al. (2014). ILAE official report: a practical clinical definition of epilepsy. Epilepsia, 55 (4), 475-82. doi: 10.1111/epi.12550.
Zou, R., Wang, S., Zhu, L. et al. (2017). Calgary score and modified Calgary score in the differential diagnosis between neurally mediated syncope and epilepsy in children. Neurol Sci., 38 (1), 143-149. doi: 10.1007/s10072-016-2740-5.
Tanaka, H., Fujita, Y., Takenaka, Y. et al. (2009). Japanese clinical guidelines for juvenile orthostatic dysregulation version 1. Pediatr Int., 51 (1), 169-79. doi: 10.1111/j.1442-200X.2008.02783.x.
Kovalchuk, T. A., Luchyshyn, N. Yu. (2022). The level of functioning of adaptive mechanisms of the cardiovascular system in children with syncope of various genesis. Modern pediatrics. Ukraine, 3 (123), 16-26. doi 10.15574/SP.2022.123.16.
Ogunleye, A., Wang, Q. G. (2020). XGBoost Model for Chronic Kidney Disease Diagnosis. IEEE/ ACM Trans Comput Biol Bioinform., 17 (6), 21312140. doi: 10.1109/TCBB.2019.2911071.
Raihan, M. J., Khan, M. A., Kee, S. H., Nahid, A. A. (2023). Detection of the chronic kidney disease using XGBoost classifier and explaining the influence of the attributes on the model using SHAP. Sci Rep., 13 (1), 6263. doi: 10.1038/s41598-023-33525-0.
Zsom, A., Tsekhan, S., Hamid, T. et al. (2019). Ictal autonomic activity recorded via wearable-sensors plus machine learning can discriminate epileptic and psychogenic nonepileptic seizures. Annu Int Conf IEEE Eng Med Biol Soc., 3502-3506. doi: 10.1109/EMBC.2019.8857552.
Anzellotti, F., Dono, F., Evangelista, G. et al. (2020). Psychogenic Non-epileptic Seizures and Pseudo-Refractory Epilepsy, a Management Challenge. Front Neurol., 11, 461. doi: 10.3389/fneur.2020.00461.
Hou, N., Li, M., He, L. et al. (2020). Predicting 30-days mortality for MIMIC-III patients with sepsis-3: a machine learning approach using XGboost. J Transl Med., 18 (1), 462. doi: 10.1186/s12967-020-02620-5.
Yue, S., Li, S., Huang, X. et al. (2022). Machine learning for the prediction of acute kidney injury in patients with sepsis. J Transl Med., 20 (1), 215. doi: 10.1186/s12967-022-03364-0.
Shi, Y., Zou, Y., Liu, J. et al. (2022). Ultrasound-based radiomics XGBoost model to assess the risk of central cervical lymph node metastasis in patients with papillary thyroid carcinoma: Individual application of SHAP. Front Oncol., 12. doi: 10.3389/fonc.2022.897596.
Li, C., Zhang, Y., Liao, Y. et al. (2022). Differential Diagnosis Between Psychogenic Pseudosyncope and Vasovagal Syncope in Children: A Quantitative Scoring Model Based on Clinical Manifestations. Front Cardiovasc Med., 9. doi: 10.3389/fcvm.2022.839183.
Ouyang, C. S., Yang, R. C., Chiang, C. T., Wu, R. C., Lin, L. C. (2020). EEG autoregressive modeling analysis: A diagnostic tool for patients with epilepsy without epileptiform discharges. Clin Neurophysiol., 131 (8), 1902-1908. doi: 10.1016/j.clinph.2020.04.172.
Kovalchuk, T. A., Boyarchuk, O. R., Bogai, S. Ye. (2023). XGBoost machine learning algorithm for differential diagnosis of pediatric syncope. The Journal of V.N. Karazin Kharkiv National University. Series «Medicine», 47, 33-46. doi: 10.26565/2313-66932023-47-04. [In Ukrainian].
Downloads
Published
How to Cite
Issue
Section
License
Journal Medical Informatics and Engineering allows the author(s) to hold the copyright without registration
The majority of Medical Informatics and Engineering Open Access journals publish open access articles under the terms of the Creative Commons Attribution (CC BY) License which permits use, distribution and reproduction in any medium, provided the original work is properly cited. The remaining journals offer a choice of licenses.
This journal is available through Creative Commons (CC) License CC-BY 4.0
