# Use of What-if Scenarios to Help Explain Artificial Intelligence Models for Neonatal Health

arXiv: https://arxiv.org/abs/2410.09635 · HTML: https://mlsanity.com/paper/2410.09635/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 71 references that are in the map.
- First posted on arXiv 12 Oct 2024.

## Builds on (most cited first, 17 of 17)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Explaining and Harnessing Adversarial Examples](https://mlsanity.com/paper/1412.6572/) (2014)
- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- [mixup: Beyond Empirical Risk Minimization](https://mlsanity.com/paper/1710.09412/) (2017)
- [XGBoost: A Scalable Tree Boosting System](https://mlsanity.com/paper/1603.02754/) (2016)
- [Towards A Rigorous Science of Interpretable Machine Learning](https://mlsanity.com/paper/1702.08608/) (2017)
- [Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning](https://mlsanity.com/paper/1704.03976/) (2017)
- [Modeling Tabular data using Conditional GAN](https://mlsanity.com/paper/1907.00503/) (2019)
- [TabNet: Attentive Interpretable Tabular Learning](https://mlsanity.com/paper/1908.07442/) (2019)
- [Explaining Machine Learning Classifiers through Diverse Counterfactual Explanations](https://mlsanity.com/paper/1905.07697/) (2019)
- [Pay Attention to MLPs](https://mlsanity.com/paper/2105.08050/) (2021)
- [Statistical Aspects of Wasserstein Distances](https://mlsanity.com/paper/1806.05500/) (2018)
- [If Only We Had Better Counterfactual Explanations: Five Key Deficits to Rectify in the Evaluation of Counterfactual XAI Techniques](https://mlsanity.com/paper/2103.01035/) (2021)
- [NICE: An Algorithm for Nearest Instance Counterfactual Explanations](https://mlsanity.com/paper/2104.07411/) (2021)
- [Scaling MLPs: A Tale of Inductive Bias](https://mlsanity.com/paper/2306.13575/) (2023)
- [Explaining Models: An Empirical Study of How Explanations Impact Fairness Judgment](https://mlsanity.com/paper/1901.07694/) (2019)
- [Network On Network for Tabular Data Classification in Real-world Applications](https://mlsanity.com/paper/2005.10114/) (2020)

## Built on by (most cited first, 5 of 5)

- [NutriGen: Personalized Meal Plan Generator Leveraging Large Language Models to Enhance Dietary and Nutritional Adherence](https://mlsanity.com/paper/2502.20601/) (2025)
- [GlyTwin: Digital Twin for Glucose Control in Type 1 Diabetes Through Optimal Behavioral Modifications Using Patient-Centric Counterfactuals](https://mlsanity.com/paper/2504.09846/) (2025)
- [LLM-Powered Prediction of Hyperglycemia and Discovery of Behavioral Treatment Pathways from Wearables and Diet](https://mlsanity.com/paper/2503.03935/) (2025)
- [Enhancing Metabolic Syndrome Prediction with Hybrid Data Balancing and Counterfactuals](https://mlsanity.com/paper/2504.06987/) (2025)
- [Trustworthy AI in Digital Health: A Comprehensive Review of Robustness and Explainability](https://mlsanity.com/paper/2608.02238/) (2026)

## Abstract (arXiv)

Early detection of intrapartum risks enables timely interventions to prevent or mitigate adverse labor outcomes such as cerebral palsy. However, accurate automated systems to support clinical decision-making during delivery are currently lacking. To address this gap, we propose Artificial Intelligence for Modeling and Explaining Neonatal Health (AIMEN), a deep learning framework that predicts adverse labor outcomes from maternal, fetal, obstetrical, and intrapartum factors while providing interpretable reasoning behind its predictions. AIMEN reveals how specific modifications to input variables could alter predicted outcomes, enhancing clinical insight. To address class imbalance and limited sample size, AIMEN employs Conditional Tabular GAN (CTGAN) for data augmentation. This process includes synthetic data generation, and we investigate in detail properties such as relaxing feature bounds for a subset of training points to explore slightly out-of-range physiological values, and applying silhouette-score-based filtering to increase the separability of synthetic samples. AIMEN uses an ensemble of fully connected neural networks for classification and outperforms state-of-the-art models such as XGBoost, TabNet, DANet, and LightGBM, achieving an average F1 score of 0.784 in predicting high-risk deliveries. Moreover, AIMEN generates counterfactual explanations that identify actionable changes involving only two to three attributes on average. Resources: https://github.com/ab9mamun/AIMEN.