Comparing conventional correction formulas and machine learning-based prediction of ionized calcium
Clinica Chimica Acta, cilt.592, 2027 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 592
- Basım Tarihi: 2027
- Doi Numarası: 10.1016/j.cca.2026.121205
- Dergi Adı: Clinica Chimica Acta
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, BIOSIS, Chemical Abstracts Core, Chimica, EMBASE, MEDLINE, Academic Search Ultimate (EBSCO), Engineering Source (EBSCO)
- Anahtar Kelimeler: Albumin, Hypocalcemia, Ionized calcium, Machine learning, Total calcium
- Ankara Yıldırım Beyazıt Üniversitesi Adresli: Evet
Özet
Background and aim Accurate assessment of ionized calcium (Ca++) is critical in clinical settings but remains technically and logistically challenging in many healthcare facilities. This study aimed to evaluate the performance of machine learning (ML) models in predicting Ca++ levels measured by blood gas analysis, using routinely available biochemical parameters—total calcium (TotCa), total protein, and albumin—and to compare them with values obtained through direct measurement and established correction formulas. Materials and methods A retrospective analysis was conducted on 84,410 patients aged 20–70 years (43,863 men, 40,547 women). Whole-blood Ca++, serum TotCa, albumin, and total protein levels were retrieved from hospital records. Three ML algorithms—Support Vector Machine (SVM), Random Forest (RF), and Gradient Boosting (GB)—were trained and validated using 5-fold cross-validation. Their performance was benchmarked against three conventional correction formulas: Hanna, Zeisler, and Butler. Results Among the conventional formulas, the Hanna method showed the highest mean absolute error (MAE = 0.3626), while Zeisler (MAE = 0.0719) and Butler (MAE = 0.0988) performed more closely to measured Ca++. The ML models outperformed all formula-based methods, with GB (R2 = 0.6742), SVM (R2 = 0.6732), and RF (R2 = 0.6730) achieving the highest explained variance. In contrast, Butler and Zeisler yielded R2 values of 0.2684 and 0.4879, respectively. Conclusions ML models demonstrate superior predictive accuracy for Ca++ compared with conventional correction formulas when using routine biochemical parameters. These findings support the potential integration of ML-based tools into clinical decision support systems. Future research should address model interpretability, pH incorporation, and prospective external validation.