xAI

21 gain · 22 problem · 0 high-evidence

1. gainHealthModerate evidenceImpact 79

   At three Indian tertiary cancer centres, oncologists using OneRx.AI reached clinical recommendations far faster, with high expert-rated accuracy, safety and guideline concordance.

   Researchers tested OneRx.AI, a retrieval-augmented platform grounded in international and national oncology guidelines, at three tertiary oncology centres in India. Three oncologists independently reviewed de-identified clinical queries covering five cancer types, and 110 of 124 submitted queries were eligible.

   Oct 15, 2026

   1 source

   • International Journal of Medical InformaticsPeer-reviewed
2. gainClimateModerate evidenceImpact 79

   Explainable AI-Supported Cyber-Physical Collaboration for Sustainable Manufacturing in Industry 5.0: Experimental findings show that the proposed CNN-Transformer architecture achieves 98.5% accuracy and outperforms existing CPHS systems while maintaining low latency.

   In this study, we present an innovative approach to the sustainable manufacturing of industrial parts using an explainable artificial intelligence (XAI)- based cyber-physical collaboration system for Industry 5.0. Current cyber-physical human systems (CPHSs) have been found to integrate AI only to a limited extent and often lack explainability.

   Sep 3, 2026

   1 source

   • Journal of Visualized ExperimentsPeer-reviewed
3. gainCrimeModerate evidenceImpact 74

   Peer-reviewed synthesis finds AI-enabled cybersecurity defenses reduce successful breaches by up to 30% in real-world deployment contexts.

   On 2025-09-08, a peer-reviewed review in Computers synthesized 256 publications to bridge computer science and cybersecurity advances in explainable AI, automated development, and privacy-preserving learning. It reports AI embedded across the software lifecycle for productivity and testing, and AI-driven real-time threat detection in security operations.

   Oct 7, 2026

   1 source

   • ComputersPeer-reviewed
4. gainHealthModerate evidenceImpact 74

   FedMediFormer-XAI combined federated learning, multimodal transformers, diffusion augmentation and GNN drug recommendation to achieve 94.2% accuracy for diabetes prediction and NDCG 0.91 for personalized drug recommendation while enabling privacy-preserving training without sharing raw patient data.

   Researchers described FedMediFormer-XAI, a framework that combines federated learning, multimodal transformers, diffusion-based data augmentation, graph neural networks for drug recommendation, and explainable AI to address fragmented diabetes data, privacy concerns, and lack of personalized guidance. It was tested on diverse inputs including clinical records, glucose monitoring, retinal fundus images, wearable sensors, and pharmacological information.

   Sep 10, 2026

   1 source

   • Journal of Visualized ExperimentsPeer-reviewed
5. problemHealthModerate evidenceImpact 74

   Four machine learning (ML) algorithms (Random Survival Forest, XGBoost, Elastic Net-regularized Cox, Support Vector Machine) were trained (80%) and tested (20%) to predict overall survival (OS).

   Background Enfortumab vedotin (EV) has transformed treatment for advanced urothelial carcinoma (aUC), but outcomes vary. Machine learning (ML) with explainable artificial intelligence (XAI) may improve survival prediction.

   Sep 4, 2026

   1 source

   • Expert Opinion on Biological TherapyPeer-reviewed
6. gainCrimeHigh evidenceImpact 70

   XAI-driven data mining applied to IoT ecosystems can detect anomalies and support automated security decisions through transparent and interpretable reasoning for smart city infrastructure.

   Published February 20, 2026, this peer-reviewed survey in Cognitive Computation reviews XAI-driven data mining for self-defending IoT systems. It describes how IoT expansion in smart cities, healthcare, and industrial automation creates need for real-time, scalable security, and how XAI methods aim to detect anomalies and support automated decisions with transparent reasoning.

   Jul 13, 2026

   5 sources

   • International Journal of Innovative Science and Research TechnologyPeer-reviewed
   • Cognitive ComputationPeer-reviewed
   • Journal of Cybersecurity and PrivacyPeer-reviewed
   • International Journal of Scientific Research in Science and TechnologyPeer-reviewed
   • Communications MaterialsPeer-reviewed
7. problemCrimeHigh evidenceImpact 70

   Traditional and opaque AI security mechanisms are inadequate for protecting interconnected urban IoT infrastructures, facing challenges of data privacy, scalability, computational constraints, and limited interpretability.

   Published February 20, 2026, this peer-reviewed survey in Cognitive Computation reviews XAI-driven data mining for self-defending IoT systems. It describes how IoT expansion in smart cities, healthcare, and industrial automation creates need for real-time, scalable security, and how XAI methods aim to detect anomalies and support automated decisions with transparent reasoning.

   Jul 13, 2026

   5 sources

   • International Journal of Innovative Science and Research TechnologyPeer-reviewed
   • Cognitive ComputationPeer-reviewed
   • Journal of Cybersecurity and PrivacyPeer-reviewed
   • International Journal of Scientific Research in Science and TechnologyPeer-reviewed
   • Communications MaterialsPeer-reviewed
8. gainHealthModerate evidenceImpact 69

   Grok AI, ChatGPT-5 Plus and Gemini 2.5 Flash gave highly rated, statistically comparable answers to complex kidney transplantation questions, suggesting a possible adjunct decision-support role under physician oversight.

   Researchers gave Grok AI, ChatGPT-5 Plus and Gemini 2.5 Flash ten open-ended scenarios drawn from kidney transplantation practice, covering surgical technique, complications, immunosuppression and perioperative care. Two experienced physicians scored the replies for accuracy, comprehensiveness, clarity and clinical relevance, and also applied a global quality score.

   Oct 17, 2026

   1 source

   • World Journal of UrologyPeer-reviewed
9. gainHealthModerate evidenceImpact 69

   In 260 cancer patients, explainable ML identified four pain phenotypes and showed that catastrophizing and distress predicted group membership more than pain intensity, supporting a framework for individualized care.

   Researchers analyzed data from 260 cancer patients with machine learning to group them by pain profile rather than by pain intensity alone. The analysis produced four phenotypes: High-Burden Pain, Catastrophizing-Dominant, Stress-Related Pain, and Low-Risk Adaptive. Of three boosting models tested, CatBoost performed best.

   Oct 16, 2026

   1 source

   • Pain Management NursingPeer-reviewed
10. gainHealthModerate evidenceImpact 69

    When the AI classifier was correct, showing its prediction probability raised geneticists' diagnostic accuracy and confidence in recognizing genetic conditions from facial images.

    Researchers ran a pilot experiment in which 44 medical geneticists, split into an AI-only group and an XAI-supported group, judged 18 facial images of people with or without genetic conditions. Accuracy and confidence were recorded before and after the geneticists saw a classifier's prediction probability, with or without saliency-based explanations.

    Oct 16, 2026

    1 source

    • Clinical DysmorphologyPeer-reviewed
11. problemHealthModerate evidenceImpact 69

    When the AI classifier was incorrect, geneticists' accuracy and confidence fell, and saliency-based explanations did not improve decisions, so the explanations were not integrated.

    Researchers ran a pilot experiment in which 44 medical geneticists, split into an AI-only group and an XAI-supported group, judged 18 facial images of people with or without genetic conditions. Accuracy and confidence were recorded before and after the geneticists saw a classifier's prediction probability, with or without saliency-based explanations.

    Oct 16, 2026

    1 source

    • Clinical DysmorphologyPeer-reviewed
12. gainPolicyModerate evidenceImpact 69

    A four-layer governance framework with mandatory explainability and bias monitoring offers a roadmap for transparent, trustworthy AI risk prediction in Iran's health insurance sector.

    Researchers in Iran developed a data governance framework called EDG-AI4H to support explainable AI in health insurance risk prediction. They combined a narrative review of international articles and Iranian policy documents with interviews of five experts, analyzed thematically in 2024.

    Oct 16, 2026

    1 source

    • Dialogues in HealthPeer-reviewed