Machine learning with SMOTE-NC on routine hematology parameters for early identification of Hb E trait
Article Sidebar
Main Article Content
Abstract
Background: The hemoglobin E (Hb E) trait, which is highly prevalent in Southeast Asia and may co-occur with α-thalassemia, often exhibits heterogeneous complete blood count profiles; notably, some individuals with the Hb E trait have near-normal MCV and MCH values, increasing the risk of misclassification during routine hematological screening.
Objective: This study aimed to develop and externally validate an interpretable machine-learning model to screen for Hb E trait using routine Complete Blood Count (CBC) data.
Materials and methods: We analyzed 5,787 standardized CBC records using two processing pipelines: a standard dataset and an SMOTE-NC oversampled dataset. We selected features using ReliefF and evaluated nine machinelearning algorithms. We interpreted the optimal model using SHAP values and validated it on an independent external cohort (n=625).
Results: Discrimination for Hb E trait was moderate-to-good and consistent between internal and external evaluation. Erythrocyte indices-Z-MCV, Z-MCH, and Z-RBC-were the most discriminative features. Among nine algorithms, a support vector machine gave the best combination of sensitivity and discrimination (internal AUC 0.881, sensitivity 70.66%, specificity 87.92%; external AUC 0.879, sensitivity 72.38%, specificity 87.39%). With the decision threshold tuned for screening, external sensitivity rose to approximately 91% with negative predictive value above 94%. SHAP analysis confirmed the model’s reliance on biologically plausible drivers (low MCV/MCH with a relatively preserved RBC count).
Conclusion: An explainable machine-learning model utilizing routine CBC parameters offers a viable, low-cost pre-screening tool for Hb E trait. Internal and external performance are concordant, and with local calibration and threshold tuning, the model triages patients for confirmatory testing and definitive genotyping rather than replacing them, supporting optimized diagnostic workflows in endemic regions.
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Personal views expressed by the contributors in their articles are not necessarily those of the Journal of Associated Medical Sciences, Faculty of Associated Medical Sciences, Chiang Mai University.
References
Pornprasert S, Tookjai M, Punyamung M, Pongpunyayuen P. HbE level and red cell parameters in heterozygous HbE with and without -thalassemia trait. Indian J Hematol Blood Transfus. 2018; 34(4): 662-5. doi.org/10.1007/s12288-018-0947-8.
Paiboonsukwong K, Jopang Y, Winichagoon P, Fucharoen S. Thalassemia in Thailand. Hemoglobin. 2022; 46(1): 53-7. doi.org/10.1080/03630269.2022.2025824.
Nuntakarn L, Fucharoen S, Fucharoen G, Sanchaisuriya K, Jetsrisuparb A, Wiangnon S. Molecular, hematological and clinical aspects of thalassemia major and thalassemia intermedia associated with Hb E--thalassemia in Northeast Thailand. Blood Cells Mol Dis. 2009; 42(1): 32-5. doi.org/10.1016/j.bcmd.2008.09.002.
Yamsri S, Singha K, Prajantasen T, Fucharoen G, Sanchaisuriya K, Fucharoen S. A large cohort of -thalassemia in Thailand: molecular, hematological and diagnostic considerations. Blood Cells Mol Dis. 2015; 54(2): 164-9. doi.org/10.1016/j.bcmd.2014.11.008.
Laengsri V, Shoombuatong W, Adirojananon W, Nantasenamart C, Prachayasittikul V, Nuchnoi P. ThalPred: a web-based prediction tool for discriminating thalassemia trait and iron deficiency anemia. BMC Med Inform Decis Mak. 2019; 19(1): 212. doi.org/10.1186/s12911-019-0929-2.
Appiahene P, Asare JW, Donkoh ET, Dimauro G, Maglietta R. Detection of iron deficiency anemia by medical images: a comparative study of machine learning algorithms. BioData Min. 2023; 16(1): 18. doi.org/10.1186/s13040-023-00319-z.
Plengsuree S, Punyamung M, Yanola J, Suwankesawong W, Pornprasert S. Red cell indices and formulas used in differentiation of -thalassemia trait from iron deficiency in Thai adults. Hemoglobin. 2015; 39(4): 235-9. doi.org/10.3109/03630269.2015.1048352.
Jahangiri M, Rahim F, Malehi AS. Diagnostic performance of hematological discrimination indices to discriminate between -thalassemia trait and iron deficiency anemia and using cluster analysis: introducing two new indices tested in an Iranian population. Sci Rep. 2019; 9(1): 18607. doi.org/10.1038/s41598-019-54575-3.
Ayyıldız H, Arslan Tuncer S. Determination of the effect of red blood cell parameters in the discrimination of iron deficiency anemia and beta thalassemia via neighborhood component analysis feature selection-based machine learning. Chemom Intell Lab Syst. 2020; 196: 103886. doi.org/10.1016/j.chemolab.2019.103886.
Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP. SMOTE: synthetic minority over-sampling technique. J Artif Intell Res. 2002; 16: 321-57. doi.org/10.1613/jair.953. [11] Pradipta GA, Wardoyo R, Musdholifah A, Sanjaya INH, Ismail M. SMOTE for handling imbalanced data problem: a review. In: Proceedings of the 6th International Conference on Information and Computing (ICIC); 2021 Nov 18-19; Semarang, Indonesia. Piscataway (NJ): IEEE; 2021. p. 1-6. doi.org/10.1109/ICIC54025.2021.9632912.
Elreedy D, Atiya AF, Kamalov F. A theoretical distribution analysis of synthetic minority oversampling technique (SMOTE) for imbalanced learning. Mach Learn. 2024; 113(7): 4903-23. doi.org/10.1007/s10994-022-06296-4.
Islahulhaq I, Wibowo W, Ratih ID. Classification of non-performing financing using logistic regression and SMOTE-NC. Int J Adv Soft Comput Appl. 2021; 13(3): 115-28. doi.org/10.15849/ijasca.211128.09.
Nasir MU, Naseem MT, Ghazal TM, Akram MU, Ashraf I. A comprehensive case study of deep learning on the detection of alpha and beta thalassemia using public and private datasets. Sci Rep. 2025; 15(1): 8540. doi.org/10.1038/s41598-025-97353-0.
Saleem M, Aslam W, Lali MIU, Rauf HT, Nasr EA. Predicting thalassemia using feature selection techniques: a comparative analysis. Diagnostics (Basel). 2023; 13(22): 3441. doi.org/10.3390/diagnostics13223441.
Das R, Saleh S, Nielsen I, Sadiq R, Datta C, Chakrabarti P, et al. Performance analysis of machine learning algorithms and screening formulae for -thalassemia trait screening of Indian antenatal women. Int J Med Inform. 2022; 167: 104866. doi.org/10.1016/j.ijmedinf.2022.104866.
Long Y, Bai W. Constructing a novel clinical indicator model to predict the occurrence of thalassemia in pregnancy through machine learning algorithm. Front Hematol. 2024; 3: 1341225. doi.org/10.3389/frhem.2024.1341225.
Munkongdee T, Tongsima S, Ngamphiw C, Sonsilphong S, Tanjarern S, Winichagoon P, et al. Predictive SNPs for -thalassemia/HbE disease severity. Sci Rep. 2021; 11(1): 10229. doi.org/10.1038/s41598-021-89641-2.
Fucharoen S, Weatherall DJ. The hemoglobin E thalassemias. Cold Spring Harb Perspect Med. 2012; 2(8): a011734. doi.org/10.1101/cshperspect.a011734.
Polprasert C, Wongprachar P, Suksusut A, Kanjanabuch T, Vaneesorn J, Tanakanya S, et al. Comprehensive screening for coexisting heterozygous-thalassemia in hemoglobin E trait. Hematology. 2020; 25(1): 276-9. doi.org/10.1080/16078454.2020.1786972.
Tepakhan W, Srisintorn W, Penglong T, Saelue P. Machine learning approach for differentiating iron deficiency anemia and thalassemia using random forest and gradient boosting algorithms. Sci Rep. 2025; 15(1): 4585. doi.org/10.1038/s41598-025-01458-5.
Lachover-Roth I, Peretz S, Zoabi H, Shani-Gigi N, Abu-Kaf H, Koren A, et al. Support vector machine-based formula for detecting suspected-thalassemia carriers. Int J Mol Sci. 2024; 25(12): 6446. doi.org/10.3390/ijms25126446.
Christensen F, Kılıç DK, Nielsen IE, Vinding MS, Farooq F, Winther O, et al. Classification of -thalassemia data using machine learning models. Comput Methods Programs Biomed. 2025; 260: 108581. doi.org/10.1016/j.cmpb.2024.108581.
Mukherjee M, Khushi M. SMOTE-ENC: a novel SMOTE-based method to generate synthetic data for nominal and continuous features. Appl Syst Innov. 2021; 4(1): 18. doi.org/10.3390/asi4010018.
Rustam F, Ashraf I, Jabbar S, Ullah S, Sultana R, Jamil F, et al. Prediction of -thalassemia carriers using complete blood count features. Sci Rep. 2022; 12(1): 17616. doi.org/10.1038/s41598-022-AC22011-8.
