Vol. 24 No. 3 (2025): Mapana Journal of Sciences
Research Articles

Tri-Class Exponential ROC Space

PDF
  • G. Kumarapandiyan,
  • Sahana T S
G. Kumarapandiyan
Department of Statistics, Madras Christian College, Chennai, India
Sahana T S
Department of Statistics, Madras Christian College, Chennai, Tamil Nadu, India
DOI: https://doi.org/10.12723/mjs.74.10

Published 2025-10-14

Keywords

  • Biomarker,
  • Volume Under the ROC,
  • Asymptotic Variance and Exponential Distribution

Copyright (c) 2025

Creative Commons License

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Abstract

Biomarkers play a vital role in detecting the presence of disease or medical condition of interest. The challenging tasks in clinical diagnosis are to interpret the performance of biomarkers.  To evaluate the biomarkers efficiency, the most advantageous tool used is Receiver Operating Characteristic (ROC) Curve.  For two class problems (abnormal and normal), various models and alternatives are developed to find the biomarker’s performance. In this study, the two class problem has been further extended to the three-class problem i.e. (normal, suspicious and abnormal). The three-class exponential ROC model, Volume Under the ROC Surface (VUS), asymptotic variance and Confidence Interval (CI) for VUS have been derived. The model has been validated using a simulated data and for the real-life dataset from the underlying distribution.

PDF

References

  1. Mossman, D. (1999). Three-way rocs. Medical Decision Making, 19(1), 78-89.
  2. https://doi.org/10.1177/0272989X9901900110
  3. Krzanowski, W. J., & Hand, D. J. (2009). ROC curves for continuous data. CRC Press.
  4. https://doi.org/10.1201/9781439800225
  5. Pundir, Sudesh, and R. Amala. "Evaluation of biomarker using two parameter bi-exponential ROC curve." Pakistan Journal of Statistics and Operation Research (2015): 481-496 https://doi.org/10.18187/pjsor.v11i4.992
  6. Heckerling, P. S. (2001). Parametric three-way receiver operating characteristic surface analysis using mathematica. Medical Decision Making, 21(5), 409-417.
  7. https://doi.org/10.1177/0272989X0102100507
  8. Everson, Richard M., and Jonathan E. Fieldsend. ”Multi-class ROC analysis from a multi-objective optimization perspective.” Pattern Recognition Letters 27.8 (2006): 918-927. https://doi.org/10.1016/j.patrec.2005.10.016
  9. Yang, H., and Carlin, D. (2000), “ROC Surface: A Generalization of ROC Curve Analysis,” Journal of Biopharmaceutical Statistics, 10, https://doi.org/10.1081/BIP-100101021
  10. Fieldsend, Jonathan E., and Richard M. Everson. ”Visualisation of multiclass ROC surfaces.” (2005). http://hdl.handle.net/10871/11691
  11. Nakas, Christos T. “Developments in ROC surface analysis and assessment of diagnostic markers in three-class classification problems.” REVSTAT-Statistical Journal 12.1 (2014): 43-65. https://doi.org/10.57805/revstat.v12i1.143
  12. Kang, Le, and Lili Tian. ”Estimation of the volume under the ROC surface with three ordinal diagnostic categories.” Computational Statistics Data Analysis 62 (2013): 39-51. https://doi.org/10.1016/j.csda.2013.01.004