Deployer-side XAI instrumentation for regulated AI: a clinical case study in ICL sizing

Vista sencilla de ítem
dc.contributor.authorUrrutia Oñate, Dorleta
dc.contributor.authorOnieva Caracuel, Enrique
dc.contributor.authorPerallos Ruiz, Asier
dc.date.accessioned2026-08-04T08:50:53Z
dc.date.available2026-08-04T08:50:53Z
dc.date.issued2026
dc.date.updated2026-08-04T08:50:53Z
dc.description.abstractRegulated AI creates a monitoring problem for deployers who must organise human oversight, log-retention and post-market surveillance while often having access only to the prediction interface. This study specifies a deployer-side XAI instrumentation protocol for the output→action boundary, where a model output becomes a reason for action. The protocol reorganises KernelSHAP, nearest-neighbour envelope checks, bounded perturbation, and rank-order stability into a per-decision evidence record computed from predict() calls. We instantiate the protocol in a clinical case of phakic Implantable Collamer Lens sizing, using a 55-eye held-out cohort and an Extra Trees regressor for post-operative vault prediction. The record contains five signals: score_margin, constraint_enforcement, envelope_validity, decision_robustness, and record_integrity, plus two cohort-level oversight aggregates. The case study shows how the same record can support decision-time human oversight, later audit and post-market surveillance under the EU AI Act and the Medical Devices Regulation.en
dc.description.sponsorshipThis research has been supported by the Ministerio de Ciencia, Innovación y Universidades, Gobierno de España, under research grant PID2024-158490OB-C33 (ECEAAS)en
dc.identifier.citationUrrutia-Onate, D., Onieva, E., & Perallos, A. (2026). Deployer-side XAI instrumentation for regulated AI: a clinical case study in ICL sizing. International Journal of Advanced Computer Science and Applications, 17(6), 45-61. https://doi.org/10.14569/IJACSA.2026.0170605
dc.identifier.doi10.14569/IJACSA.2026.0170605
dc.identifier.eissn2156-5570
dc.identifier.issn2158-107X
dc.identifier.urihttps://hdl.handle.net/20.500.14454/6463
dc.language.isoeng
dc.publisherScience and Information Organization
dc.subject.otherEU AI Act
dc.subject.otherExplainable AI
dc.subject.otherHuman oversight
dc.subject.otherICL sizing
dc.subject.otherMedical Device Software
dc.subject.otherXAI instrumentation
dc.titleDeployer-side XAI instrumentation for regulated AI: a clinical case study in ICL sizingen
dc.typejournal article
dcterms.accessRightsopen access
oaire.citation.endPage61
oaire.citation.issue6
oaire.citation.startPage45
oaire.citation.titleInternational Journal of Advanced Computer Science and Applications
oaire.citation.volume17
oaire.licenseConditionhttps://creativecommons.org/licenses/by/4.0/
oaire.versionVoR
Archivos
Bloque original
Mostrando 1 - 1 de 1
Miniatura
Nombre:
urrutia_deployer_2026.pdf
Tamaño:
1.84 MB
Formato:
Adobe Portable Document Format
Descargar
Colecciones
Artículos