Yapay zeka (YZ), sağlık hizmetlerinde devrim yaratma potansiyeli taşımakta ve Yapay Zeka Tabanlı Klinik Karar Destek Sistemleri (YZ-KKDS), bu dönüşümün merkezinde yer almaktadır. Bu sistemler, tanı doğruluğunu artırma, tedavi planlarını kişiselleştirme ve operasyonel verimliliği yükseltme gibi önemli vaatler sunmaktadır (Bulut, 2023). Ancak, YZ modellerinin geliştirilmesi, bu teknolojinin başarılı bir şekilde benimsenmesi için gereken sürecin yalnızca bir parçasıdır. Türkiye sağlık sisteminin kendine özgü dinamikleri; parçalı Hastane Bilgi Yönetim Sistemleri (HBYS) yapısı, Kişisel Verilerin Korunması Kanunu (KVKK) gibi katı yasal düzenlemeler, hekimlerin iş akışları ve teknolojiye yönelik tutumları gibi faktörler, entegrasyon sürecini karmaşık bir sosyo-teknik probleme dönüştürmektedir (Dizman, 2018; Önder, 2022). Bu makale, YZ-KKDS’nin Türkiye’de başarılı ve sürdürülebilir entegrasyonu için bütüncül bir yol haritası sunmayı amaçlamaktadır. Literatür taraması, Türkiye’deki mevcut durum analizi ve sistem tasarımı prensiplerini birleştiren bir metodoloji kullanılarak, teknolojik gelişimin ötesine geçen ve insan, organizasyon ve politika faktörlerini merkeze alan çok boyutlu bir çerçeve modeli önerilmektedir. Önerilen model; (1) Yönetişim ve Politika, (2) Teknik Altyapı ve Birlikte Çalışabilirlik, (3) Klinik Entegrasyon ve Kullanıcı Odaklılık, (4) Yasal ve Etik Uyum, (5) Eğitim ve Kapasite Geliştirme olmak üzere beş temel sütun üzerine inşa edilmiştir. Bu çerçeve, YZ-KKDS entegrasyonunun önündeki sosyo-teknik engelleri aşmak için politika yapıcılara, sağlık yöneticilerine ve teknoloji geliştiricilere stratejik bir rehber sunmaktadır.
Yapay Zeka Klinik Karar Destek Sistemleri Sosyo-Teknik Sistemler Teknoloji Entegrasyonu Sağlık Bilişimi Hastane Bilgi Yönetim Sistemleri (HBYS) Türkiye Sağlık Sistemi Teknoloji Kabul Modeli (TAM) KVKK.
Artificial Intelligence (AI) holds transformative potential in healthcare, with AI-Based Clinical Decision Support Systems (AI-CDSS) positioned at the core of this transformation. These systems offer significant promises such as improving diagnostic accuracy, personalizing treatment plans, and enhancing operational efficiency (Bulut, 2023). However, model development constitutes only one part of the process necessary for the successful adoption of this technology. The unique dynamics of the Turkish healthcare system - such as the fragmented structure of Hospital Information Management Systems (HIMS), strict legal regulations like the Personal Data Protection Law (PDPL), physicians' workflows, and their attitudes toward technology - turn the integration process into a complex socio-technical challenge (Dizman, 2018; Önder, 2022). This article aims to present a comprehensive roadmap for the successful and sustainable integration of AI-CDSS in Turkey. Employing a methodology that combines literature review, an analysis of the current state in Turkey, and system design principles, the study proposes a multidimensional framework model that goes beyond technological development and centers on human, organizational, and policy factors. The proposed model is built upon five core pillars: (1) Governance and Policy, (2) Technical Infrastructure and Interoperability, (3) Clinical Integration and User-Centeredness, (4) Legal and Ethical Compliance, and (5) Education and Capacity Building. This framework offers strategic guidance to policymakers, healthcare administrators, and technology developers in overcoming the socio-technical barriers to AI-CDSS integration.
Artificial Intelligence Clinical Decision Support Systems (CDSS) Socio-Technical Systems Technology Integration Health Informatics Hospital Information Management Systems (HIMS) Turkish Healthcare System Technology Acceptance Model (TAM) Personal Data Protection Law (PDPL).
Primary Language | Turkish |
---|---|
Subjects | Health Services and Systems (Other) |
Journal Section | Research Article |
Authors | |
Early Pub Date | July 19, 2025 |
Publication Date | July 20, 2025 |
Submission Date | March 30, 2025 |
Acceptance Date | June 25, 2025 |
Published in Issue | Year 2025 Volume: 9 Issue: 2 |