Cuffless blood pressure screening based on easily acquired photoplethysmography (PPG) signals offers a practical pathway toward scalable cardiovascular health assessment. Despite rapid progress, existing PPG-based blood pressure estimation models have not consistently achieved the established clinical numerical limits such as AAMI/ISO 81060-2, and prior evaluations often lack the rigorous experimental controls necessary for valid clinical assessment. Moreover, the publicly available datasets commonly used are heterogeneous and lack physiologically controlled conditions for fair benchmarking. To enable fair benchmarking under physiologically controlled conditions, we created a standardized benchmarking subset NBPDB comprising 101,453 high-quality PPG segments from 1,103 healthy adults, derived from MIMIC-III and VitalDB. Using this dataset, we systematically benchmarked several state-of-the-art PPG-based models. The results showed that none of the evaluated models met the AAMI/ISO 81060-2 accuracy requirements (mean error $<$ 5 mmHg and standard deviation $<$ 8 mmHg). To improve model accuracy, we modified these models and added patient demographic data such as age, sex, and body mass index as additional inputs. Our modifications consistently improved performance across all models. In particular, the MInception model reduced error by 23\% after adding the demographic data and yielded mean absolute errors of 4.75 mmHg (SBP) and 2.90 mmHg (DBP), achieves accuracy comparable to the numerical limits defined by AAMI/ISO accuracy standards. Our results show that existing PPG-based BP estimation models lack clinical practicality under standardized conditions, while incorporating demographic information markedly improves their accuracy and physiological validity.
The role of artificial intelligence is growing in healthcare and disease prediction. Because of its potential impact and demographic disparities that have been identified in machine learning models for disease prediction, there are growing concerns about transparency, accountability and fairness of these predictive models. However, very little research has investigated methods for improving model fairness in disease prediction, particularly when the sensitive attribute is multivariate and when the distribution of sensitive attribute groups is highly skewed. In this work, we explore algorithmic fairness when predicting heart disease and Alzheimer's Disease and Related Dementias (ADRD). We propose a fine tuning approach to improve model fairness that takes advantage of observations from the majority groups to build a pre-trained model and uses observations from each underrepresented subgroup to fine tune the pre-trained model, thereby incorporating additional specific knowledge about each subgroup. We find that our fine tuning approach performs better than other algorithmic fairness fixing methods across all subgroups even if the subgroup distribution is very imbalanced and some subgroups are very small. This is an important step toward understanding approaches for improving fairness for healthcare and disease prediction.
