Requirements Specification of a Machine Learning-based Medical Software System

• Bruno Garcia de Oliveira Breda USP
• Elisa Yumi Nakagawa USP

DOI: https://doi.org/10.5753/sbcas_estendido.2026.21721

Resumo

Medicine software systems have increasingly incorporated Machine Learning (ML) models, posing significant challenges for Software Engineering (SE), particularly in dealing with their requirements that could ensure qualities such as reliability and security. In this scenario, the main problem is the difficulty of specifying requirements that often change continuously, including in the context of an large research project in which ML-based medical systems have been developed and this work is part of. Additionally, there is no consensus on better techniques for this. Hence, the main objective of this work was to define techniques for requirements specification and apply them in a realworld ML-based medical system. To this end, we scrutinized the literature to select techniques for specifying requirements; following this, we applied them to a medical system for predicting patient’s respiratory health conditions in the context of the large project. As a result, the specified requirements were the key element to the successful development of the system. Hence, these techniques could be helpful for the development of other ML-based medical systems.

Referências

Barrera, J. M., Reina-Reina, A., Lavalle, A., Maté, A., and Trujillo, J. (2024). An extension of iStar for machine learning requirements by following the prise methodology. Computer Standards & Interfaces, 88:1–15.

Breda, B., Valle, P., Tamburri, D., and Nakagawa, E. Y. (2026). Towards requirements specification for machine learning-based software systems. In 29th Ibero-American Conference on Software Engineering (CibSE), pages 1–15.

Dick, J., Hull, E., and Jackson, K. (2017). Requirements Engineering. Springer International Publishing, 4 edition.

Kirikova, M. (2017). Continuous requirements engineering. In 18th International Conference on Computer Systems and Technologies (CompSysTech), pages 1–10.

Kitchenham, B. A., Budgen, D., and Brereton, P. (2015). Evidence-based software engineering and systematic reviews, volume 4. CRC press.

Michailidis, K., Strazdina, R., and Kirikova, M. (2021). Continuous requirements engineering for digital transformation. In CEUR Workshop Proceedings, volume 2991, pages 26–40.

Nalchigar, S., Yu, E., and Keshavjee, K. (2021). Modeling machine learning requirements from three perspectives: a case report from the healthcare domain. Requirements Engineering, 26:237–254.

O’Regan, G. (2017). Requirements engineering. In Undergraduate Topics in Computer Science, pages 47–60. Springer International Publishing.

Uysal, M. P. (2023). Requirements modeling: A use case approach to machine learning. In Integrating Machine Learning Into HPC-Based Simulations and Analytics, chapter 19, pages 261–275. IGI Global.

Vogelsang, A. and Borg, M. (2019). Requirements engineering for machine learning: Perspectives from data scientists. In IEEE 27th International Requirements Engineering Conference Workshops (REW), pages 245–251.

Yang, Y., Zeng, B., and Gao, J. (2024). RM4ML: requirements model for machine learning-enabled software systems. Requirements Engineering, 30:1–33.