CORTEXA
← Browse
crossrefFrontiers in Digital Health2026-07-17Cited by 0

Adversarial debiasing for age-equitable diabetes prediction: performance–fairness trade-offs and partition dependency in machine learning

Vinod Kumar Yata, Sravanthi Jena, Meera Indracanti, Shivaprasad Chitta, Narasaiah Kolliputi

Background Machine learning models used for diabetes risk prediction may encode age-related biases that reduce diagnostic accuracy for specific demographic groups. Adversarial debiasing with a gradient reversal layer (GRL) offers a theoretically principled approach to learning representations that are invariant to a protected attribute; however, its practical effectiveness under realistic conditions of subgroup imbalance in healthcare datasets has not been fully characterised. Research question Does adversarial debiasing with a GRL improve age-equitable diabetes prediction, and how do its fairness effects vary across different data partitions? Methods Adversarial debiasing was evaluated for age-bias mitigation in diabetes prediction using the publicly available Pima Indians Diabetes Database ( n = 768). All eight dataset predictors were used; three age groups (<30, 30–50, and >50 years) were derived from age for fairness evaluation. An adversarial neural model with a gradient reversal layer was compared against a logistic regression baseline. Features were standardised using a scaler fitted on training data only. The train–test split was stratified by diabetes outcome. Overall performance metrics (accuracy, recall, ROC-AUC) and the recall parity gap across age groups were computed on a primary labelled test partition ( n = 154); robustness was assessed across five independent random seeds (0–4). Results On the primary test partition, the adversarial model improved recall for the smallest age group [>50 years: 0.5556 → 0.7778, +22.22 percentage points (pp)] while maintaining comparable overall discrimination (ROC-AUC: 0.7852 → 0.7896, +0.45 pp). However, the recall parity gap increased from 0.0996 to 0.2153 (+11.57 pp), reflecting a concurrent decline in recall for the <30-year group (−6.25 pp). Across five random seeds, the mean recall parity gap showed a modest mean reduction (0.3282 → 0.3033, −2.49 pp), but with high variability (SD > 0.27) exceeding the mean difference. The adversarial model reduced the fairness gap in three of five seeds, increased it in one, and produced no change in one. Conclusion Adversarial debiasing can improve predictive recall for underrepresented demographic subgroups but does not guarantee consistent fairness improvements across data partitions, particularly when subgroup sample sizes are small. Multi-seed evaluation is essential for reliable fairness assessment; single train–test splits are insufficient.

View free PDFSource page

Related papers

openalexFrontiers in Digital Health2026-07-23

CancerStop.dev: an interactive web platform integrating prognostic data, clinical trials, and genomic resources for patient empowerment

Vedanth Ramji, Baladithya Muralitharan, Ganeshram Janakiraman, Natarajan Ganesan

Patients facing cancer diagnoses must navigate fragmented information spanning prognosis, clinical research opportunities, and genomics-informed therapy. CancerStop.dev is a React-based web platform that consolidates trusted public resources into a single patient-centered interfa…

View free PDFSource page
openalexFrontiers in Digital Health2026-07-23

Artificial intelligence in HIV research: a structured review and task-oriented clinical framework

Ruben E. Munoz-Cabrera, Joaquin Bravo-Urbieta, Raquel Martinez-España, Sergio Alemán Belando, José Miguel Gómez Verdú, Enrique Bernal, et al.

Background Human Immunodeficiency Virus (HIV) poses a global health challenge despite the success of the Antiretroviral Treatment (ART), which allows the disease to be potentially controlled. The growing availability of heterogeneous data has impulsed the use of Artificial Intell…

View free PDFSource page
openalexFrontiers in Digital Health2026-07-23

Coping to move again: the psychological dimension of tele-supported rehabilitation after knee arthroplasty—A randomized controlled trial

Teresa Paolucci, Laura Rizzo, Alice Cichelli, Federica Bressi, Andrea Bernetti, Giacomo Farì, et al.

Introduction Persistent pain and functional limitations after total knee arthroplasty (TKA) remain clinical challenges, influenced heavily by psychological factors like coping strategies and treatment expectancy. Tele-rehabilitation offers a promising approach to enhance continui…

View free PDFSource page
crossrefFrontiers in Digital Health2026-07-21

Application of machine learning algorithms to predict heat-sensitive angina (HSA) attacks: a multicentric observational cohort study

Jincheng Wang, Conghui Zhou, Yue Zhao, Jingqing Hu

Background The association between air temperature and angina has been confirmed by several studies, which show that some patients with cardiovascular disease are “heat sensitive” and experience angina more frequently in high-temperature environments. Although several predictive…

View free PDFSource page
crossrefFrontiers in Digital Health2026-07-15

Contextual recommendation modeling in eCoaching with machine learning, X-AI, and semantic ontology

Ayan Chatterjee, Nurilla Avazov

Physical activities can be divided into indoor and outdoor activities. While outdoor activities offer enjoyable fitness opportunities, they are often limited by weather conditions. Unfavorable weather conditions such as cold, rain, fog, or snow can significantly re­duce physical…

View free PDFSource page
crossrefFrontiers in Digital Health2026-07-13

Association between clinical characteristics within 6 h of ICU admission and 30-day mortality risk in immunocompromised sepsis patients: development and validation of a machine learning model based on the MIMIC-IV database

Zhipeng Cheng, Xiuqing Ma, Weiying Duan, Zeyu Mou, Zhixin Liang

Objective To develop and validate a machine learning model for predicting 30-day mortality in immunocompromised sepsis patients using clinical data within 6 h of ICU admission. Methods This retrospective cohort study utilized data from the MIMIC-IV and eICU databases. Adult immun…

View free PDFSource page