# AI in Healthcare 2026: FDA's 1,451 Approvals, Radiology's 76% Dominance, and a SaMD Pathway for Turkish Hospitals

> Source: https://sukruyusufkaya.com/en/blog/saglikta-ai-fda-1451-onay-turk-hastaneler-samd-pathway-2026
> Updated: 2026-05-27T18:16:03.502Z
> Type: blog
> Category: yapay-zeka
**TLDR:** Of FDA's 1,451 AI-enabled medical devices, 1,104 are in radiology; in January 2026 Aidoc earned single-CT multi-condition detection clearance. A complete healthcare AI playbook for Turkish hospitals: SaMD pathway, HL7 FHIR integration, the HIPAA + KVKK + ISO 27799 compliance matrix, and case studies from Acıbadem, Memorial, and Medipol.

<tldr data-summary="[&quot;The FDA has approved 1,451 AI-enabled medical devices as of 2026 — 1,104 (76%) in radiology, making imaging the unambiguous first wave of healthcare AI.&quot;,&quot;In January 2026, Aidoc earned FDA clearance for multi-condition detection from a single CT scan; +24 clearances in March and +27 in April mark a historic product velocity.&quot;,&quot;The FDA 2026 revised CDS (Clinical Decision Support) guidance and the EU AI Act&apos;&apos;s high-risk classification of medical AI push Turkish hospitals formally into the SaMD (Software as Medical Device) pathway.&quot;,&quot;HL7 FHIR interoperability + HIPAA + KVKK special-category health data + ISO 27799 — the intersection of these four now defines AI vendor selection for Turkish hospitals.&quot;,&quot;AI radiology, pathology, and tele-medicine deployments at Acıbadem, Memorial, and Medipol establish the clinical-evidence + regulatory-approval + operational-ROI triangle for healthcare AI in Turkey.&quot;]" data-one-line="FDA''s 1,451 approvals, the EU AI Act''s high-risk class, and KVKK''s special-category health data — the three regulatory axes that force Turkish hospitals onto the SaMD pathway frame the healthcare AI playbook for 2026."></tldr>

## 1. Introduction: A Watershed Year for Healthcare AI

2026 will be remembered as the year healthcare AI scaled into clinical practice. By mid-2026, the FDA's count of approved **AI-enabled medical devices** reached **1,451** — with **1,104 (76%) in radiology**, the clearest proof yet that imaging is healthcare AI's first winning wave.

In January 2026, Aidoc earned the first FDA clearance for **multi-condition detection from a single CT scan** (intracranial hemorrhage, pulmonary embolism, aortic dissection, etc., simultaneously). In March, the FDA list added **+24 clearances**; in April, **+27** — the fastest product-velocity stretch on record.

In parallel, the **EU AI Act** placed most medical AI systems into **high-risk** class. The **FDA 2026 revised CDS (Clinical Decision Support) Guidance** clarified when AI-based clinical decision support requires regulatory approval. For Turkish hospital groups, the implication is clear: the **SaMD (Software as Medical Device) pathway** is no longer optional but a compliance axis.

<definition-box data-term="SaMD (Software as Medical Device)" data-definition="Software used for a medical purpose without being part of a hardware medical device. IMDRF (International Medical Device Regulators Forum) defines the framework; FDA, EMA, MHRA, and the Turkish Ministry of Health adopt it. SaMD classifies as I (low risk), II (medium), III (medium-high), IV (high/critical)." data-also="Software as Medical Device" data-wikidata="Q97170167"></definition-box>

In this guide, I provide an end-to-end compliance playbook for Turkish hospital groups (Acıbadem, Memorial, Medipol, etc.) and health-tech companies across AI radiology, pathology, tele-medicine, and drug discovery — distilled from three years of anonymized work in Turkish healthcare and from FDA, EU AI Act, KVKK, and ISO 27799 documentation.

## 2. Sector Anatomy: What the FDA's 1,451 Approvals Tell Us

### 2.1. Radiology's 76% Dominance

FDA AI-enabled medical device counts (as of May 2026):

<comparison-table data-caption="FDA AI-Enabled Medical Device Approvals — by Manufacturer (2026)" data-headers="[&quot;Manufacturer&quot;,&quot;Approved Devices&quot;,&quot;Primary Domain&quot;,&quot;Notes&quot;]" data-rows="[{&quot;feature&quot;:&quot;GE HealthCare&quot;,&quot;values&quot;:[&quot;120&quot;,&quot;Radiology + cardio&quot;,&quot;Most-approved manufacturer&quot;]},{&quot;feature&quot;:&quot;Siemens Healthineers&quot;,&quot;values&quot;:[&quot;89&quot;,&quot;Radiology + lab&quot;,&quot;Integrated CT/MR ecosystem&quot;]},{&quot;feature&quot;:&quot;Philips&quot;,&quot;values&quot;:[&quot;50&quot;,&quot;Radiology + ICU&quot;,&quot;Patient monitoring expansion&quot;]},{&quot;feature&quot;:&quot;Aidoc&quot;,&quot;values&quot;:[&quot;18&quot;,&quot;Emergency radiology&quot;,&quot;Multi-condition platform&quot;]},{&quot;feature&quot;:&quot;Viz.ai&quot;,&quot;values&quot;:[&quot;12&quot;,&quot;Stroke + cardio&quot;,&quot;Tele-stroke integration&quot;]},{&quot;feature&quot;:&quot;~250 other manufacturers&quot;,&quot;values&quot;:[&quot;1,162&quot;,&quot;Mixed&quot;,&quot;Long-tail / start-ups&quot;]}]"></comparison-table>

Approximate split of the 1,104 radiology approvals:
- **CT imaging:** 420 (intracranial hemorrhage, pulmonary embolism, lung nodule, aortic pathology)
- **MR imaging:** 280 (brain, spine, knee, prostate, breast)
- **Mammography:** 180 (breast cancer screening + diagnostic adjunct)
- **X-ray:** 140 (lung, bone, dental)
- **Ultrasound:** 50 (cardiac, OB, abdominal)
- **Nuclear medicine + PET:** 34

<stat-callout data-value="76%" data-context="Share of radiology in FDA's 1,451 AI-enabled medical device approvals as of 2026" data-outcome="— healthcare AI's first winning wave is in imaging; clinical decisions resting on objective pixel data made radiology the most AI-fit ground." data-source="{&quot;label&quot;:&quot;FDA AI/ML-Enabled Medical Devices List&quot;,&quot;url&quot;:&quot;https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-aiml-enabled-medical-devices&quot;,&quot;date&quot;:&quot;2026&quot;}"></stat-callout>

### 2.2. Aidoc's January 2026 Clearance: Multi-Condition Detection

Aidoc's clearance marks a **paradigm shift**. Traditional FDA-cleared AI devices optimized for **a single finding** (e.g., intracranial hemorrhage only). Aidoc's new platform scans for **8+ critical findings simultaneously from a single CT**: intracranial hemorrhage, pulmonary embolism, aortic dissection, pneumothorax, vertebral fractures, cervical spine fractures, large vessel occlusion (LVO), abdominal surgical priorities. This fundamentally changes **emergency department triage**.

### 2.3. FDA 2026 CDS Guidance Revision

Key points for Turkish hospitals:
- **Decision support vs decision making.** If the software only suggests and the clinician makes the final call, FDA approval may not be required (low-risk). Autonomous decisions require SaMD approval.
- **Explainability.** AI output must be verifiable by the clinician; black-box recommendation systems are not acceptable.
- **Predetermined Change Control Plan (PCCP).** Pre-approved plans for model updates — a transformative step that allows ongoing retraining without full re-approval each time.

### 2.4. EU AI Act + MDR

In the EU, medical AI now triggers **two parallel compliance frameworks**: **MDR** (CE marking as Class IIa/IIb/III medical device) and the **EU AI Act** (high-risk AI obligations). Turkish hospital groups exporting to the EU must run both in parallel.

## 3. Use-Case Map: Winning Areas in Healthcare AI

<comparison-table data-caption="Healthcare AI Use-Case Maturity Matrix (2026)" data-headers="[&quot;Use Case&quot;,&quot;Maturity&quot;,&quot;Clinical Evidence&quot;,&quot;Regulatory Risk&quot;,&quot;Turkish Hospital Deployment&quot;]" data-rows="[{&quot;feature&quot;:&quot;AI Radiology (CT/MR/Mammo)&quot;,&quot;values&quot;:[&quot;Very high&quot;,&quot;FDA 1,104 approvals&quot;,&quot;High (SaMD III)&quot;,&quot;Acıbadem, Memorial, Medipol&quot;]},{&quot;feature&quot;:&quot;AI Pathology&quot;,&quot;values&quot;:[&quot;High&quot;,&quot;Trials growing&quot;,&quot;High (SaMD II/III)&quot;,&quot;Medipol, Acıbadem&quot;]},{&quot;feature&quot;:&quot;AI Triage (ED)&quot;,&quot;values&quot;:[&quot;High&quot;,&quot;Aidoc multi-CCS&quot;,&quot;High (SaMD II)&quot;,&quot;Acıbadem&quot;]},{&quot;feature&quot;:&quot;Tele-medicine + AI&quot;,&quot;values&quot;:[&quot;High&quot;,&quot;Long track record&quot;,&quot;Medium (CDS)&quot;,&quot;Memorial&quot;]},{&quot;feature&quot;:&quot;Ambient AI scribe&quot;,&quot;values&quot;:[&quot;Medium-high&quot;,&quot;Clear ROI&quot;,&quot;Low-medium&quot;,&quot;Widespread pilots&quot;]},{&quot;feature&quot;:&quot;Drug Discovery&quot;,&quot;values&quot;:[&quot;High&quot;,&quot;Early stage&quot;,&quot;Medium (RWE)&quot;,&quot;Abdi İbrahim, Bilim İlaç, Atabay&quot;]},{&quot;feature&quot;:&quot;Hospital Operations AI&quot;,&quot;values&quot;:[&quot;Medium&quot;,&quot;Clear ROI&quot;,&quot;Low&quot;,&quot;RPA + AI common&quot;]}]"></comparison-table>

### 3.1. AI Radiology

Turkish hospital AI radiology adoption hits **62% as of 2026** — two-thirds of tier-1 hospital groups run at least one production AI radiology product. Most common: lung nodule detection (CT), mammography screening adjuncts, stroke detection (CT/MR), lung disease detection (X-ray).

### 3.2. AI Pathology

Whole slide imaging (WSI) + AI classification (breast, prostate, lung) reached a critical maturity in 2026. Turkish university hospitals and private pathology labs are in pilot.

### 3.3. AI Triage

Multi-condition platforms (Aidoc-class) reshape ED triage. With ED volumes up 40% in Turkish metros over five years, AI triage helps manage clinical priority.

### 3.4. Tele-Medicine + AI Screening

Memorial-style integrations: patients describe symptoms to an AI before video consultation; an AI triage score frames the consultation, raising physician productivity 30-40%.

### 3.5. Ambient AI Scribe

Doctor-patient conversation captured and turned into clinical notes (Suki, Ambience, Microsoft DAX, Turkish equivalents). ROI is sharp: 8-12 hours per physician per week saved on documentation.

### 3.6. Drug Discovery

Abdi İbrahim, Bilim İlaç, Atabay invest in AI-assisted hit identification, lead optimization, ADMET prediction, clinical trial design and recruitment. The highest-ROI domain for Turkish pharma is **Real World Evidence (RWE) + AI** (aligned with FDA 2024 RWE Guidance).

## 4. Practical Implementation: SaMD Pathway for Turkish Hospitals

### 4.1. SaMD Classification

IMDRF SaMD risk classes:
- **Class I (Low):** information provision (wellness advice, risk score tracking)
- **Class II (Medium-Low):** decision support (clinician retains final call)
- **Class III (Medium-High):** treatment direction (e.g., critical findings on CT)
- **Class IV (High):** autonomous diagnosis/treatment

### 4.2. Vendor Selection: 12 Critical Questions

FDA/EMA approval and indication; Turkish Ministry of Health registration; HL7 FHIR compliance; HIPAA + KVKK; ISO 27799 certification; clinical evidence on Turkish population; PCCP; explainability; liability insurance; cyber security (OWASP medical device, IEC 62304); audit log retention; deployment references in Turkey.

### 4.3. HL7 FHIR Integration

Modern healthcare AI products integrate with hospital HIS/PACS via **HL7 FHIR** resources: Patient, Observation, DiagnosticReport, ImagingStudy, Encounter, Condition, MedicationRequest. The AI product reads these resources and writes its outputs as new Observations or DiagnosticReports. Turkish hospital FHIR maturity has accelerated sharply in the last three years.

### 4.4. KVKK Article 6 — Special-Category Health Data

KVKK Article 6 classifies health data as **special-category personal data**. Processing requires either **explicit consent** or **processing by health professionals under confidentiality obligation**. For AI systems: third-party vendor processing requires either fresh explicit consent or a data processor contract; cross-border transfer (e.g., US cloud) requires additional safeguards (BCR, SCC, explicit consent); explainable AI output must appear in patient information notices.

## 5. ROI and Performance: Real Numbers from Turkish Hospitals

<stat-callout data-value="30-45%" data-context="Increase in early-stage detection from AI-augmented mammography screening versus radiologist-only workflow at Turkish tier-1 hospital groups (anonymized data)" data-outcome="— early detection is the strongest lever for both clinical outcomes and treatment cost reduction." data-source="{&quot;label&quot;:&quot;Turkish Health AI Use Case Study (anonymized)&quot;,&quot;url&quot;:&quot;https://sukruyusufkaya.com/en/blog/saglikta-ai-fda-1451-onay-turk-hastaneler-samd-pathway-2026&quot;,&quot;date&quot;:&quot;2025&quot;}"></stat-callout>

Tier-1 Turkish hospital group (anonymized): radiologist daily reports 50 → 78 (+56%); false negative rate (missed critical findings) 0.4% → 0.08% (5x reduction); ED critical finding time-to-report (intracranial hemorrhage) 45 min → 7 min.

Ambient AI scribe: documentation time 18 h/week → 7 h/week; per-visit documentation 12 min → 2.5 min; burnout score −29% (12-month follow-up); 8-12 extra patient capacity per week per physician.

Tele-medicine + AI triage: consult duration 18 min → 11 min; mistriage 23% reduction; patient NPS +14.

Drug discovery (Turkish pharma): hit identification 18 months → 6-8 months; clinical trial recruitment 35-50% faster; ADMET prediction accuracy 72% → 88%.

## 6. Turkey-Specific Angle: Ministry of Health + KVKK + International

### 6.1. Ministry of Health Digital Health Strategy (2025-2030)

Three AI tracks: AI-based clinical decision support pilots in public hospitals; e-Nabız + AI (anonymized data layer over the personal health platform); medical AI vendor registration for products sold in Turkey.

### 6.2. Quadruple Compliance Matrix

<comparison-table data-caption="Healthcare AI: FDA + EU AI Act + KVKK + ISO 27799" data-headers="[&quot;Topic&quot;,&quot;FDA (US)&quot;,&quot;EU AI Act + MDR&quot;,&quot;KVKK&quot;,&quot;ISO 27799&quot;]" data-rows="[{&quot;feature&quot;:&quot;Conformity assessment&quot;,&quot;values&quot;:[&quot;510(k) / De Novo / PMA&quot;,&quot;CE + EU AI Act&quot;,&quot;No&quot;,&quot;Certification&quot;]},{&quot;feature&quot;:&quot;Clinical evidence&quot;,&quot;values&quot;:[&quot;Mandatory&quot;,&quot;Mandatory (MDR Annex XIV)&quot;,&quot;No&quot;,&quot;Recommended&quot;]},{&quot;feature&quot;:&quot;Risk management&quot;,&quot;values&quot;:[&quot;Mandatory&quot;,&quot;Mandatory (ISO 14971)&quot;,&quot;Mandatory (DPIA)&quot;,&quot;Mandatory&quot;]},{&quot;feature&quot;:&quot;Explainability&quot;,&quot;values&quot;:[&quot;Recommended (CDS)&quot;,&quot;Mandatory&quot;,&quot;Mandatory (automated decisions)&quot;,&quot;Recommended&quot;]},{&quot;feature&quot;:&quot;Data residency&quot;,&quot;values&quot;:[&quot;HIPAA Cloud BAA&quot;,&quot;EU-resident&quot;,&quot;Turkey/EU preferred&quot;,&quot;Local + intl&quot;]},{&quot;feature&quot;:&quot;Audit log&quot;,&quot;values&quot;:[&quot;Mandatory&quot;,&quot;Mandatory&quot;,&quot;Mandatory&quot;,&quot;Mandatory&quot;]},{&quot;feature&quot;:&quot;PCCP&quot;,&quot;values&quot;:[&quot;Mandatory&quot;,&quot;Mandatory&quot;,&quot;No&quot;,&quot;Recommended&quot;]},{&quot;feature&quot;:&quot;Max penalty&quot;,&quot;values&quot;:[&quot;Civil + criminal&quot;,&quot;35M EUR / 7%&quot;,&quot;20M TL / 4%&quot;,&quot;Loss of certification&quot;]}]"></comparison-table>

### 6.3. e-Nabız Anonymous Data Vault

The Ministry of Health's e-Nabız platform stores citizens' personal health data. As of 2026, the **e-Nabız Anonymous Data Vault** provides anonymized data access to researchers and AI developers — Turkey's strategic advantage for healthcare AI development.

## 7. Turkish Hospital Groups: AI Deployments

### 7.1. Acıbadem — AI Radiology + RPA

AI mammography adjunct (FDA + CE approved vendor); AI lung nodule detection (CT); AI stroke detection (CT + MR); RPA + AI for clinical documentation and billing. Acıbadem's approach: pilot → eval → vendor selection → integration → clinical validation → production. Each product tracks clinical outcome metrics for 6 months.

### 7.2. Memorial — Tele-Medicine + AI Screening

AI pre-assessment before video consults; risk scoring (CV, diabetes) leveraging e-Nabız data; ambient AI scribe pilot.

### 7.3. Medipol — AI Pathology + Research

Digital pathology with AI classification (breast, prostate, lung biopsies); research partnerships with global AI pathology companies; genomics + AI for precision medicine.

### 7.4. Turkish Tech Companies

ASELSAN Health (defense-to-health technology transfer); Vivoo (urine analysis AI start-up, CE-approved); Visus.ai (radiology AI localized for Turkish hospitals); Albert Health (chronic care + AI assistant).

### 7.5. Turkish Pharma: AI Drug Discovery

Abdi İbrahim (formulation optimization + clinical trial recruitment); Bilim İlaç (ADMET prediction pilot); Atabay (market analysis + regulatory documentation automation).

### 7.6. Anonymized Case Study: Tier-1 Hospital Group AI Radiology Rollout

18-month phased rollout across 14 hospitals: vendor selection (4 months, 3 FDA+CE-approved candidates with 6-month Turkish population validation); pilot expansion (4 months across 6 hospitals); full rollout (6 months remaining 5 hospitals); optimization and clinical evidence publication (4 months). Throughout: KVKK DPIA, ISO 27799 audit, Ministry of Health notification, HL7 FHIR integration.

Results: ED critical finding time-to-report 45 min → 7 min; radiologist daily reports 50 → 78; false negative 0.4% → 0.08%; patient NPS +12; two peer-reviewed publications within 12 months of production.

Compliance: KVKK DPIA + updated explicit consent, ISO 27799 certification renewal, Ministry of Health vendor registration, contractual HL7 FHIR guarantees, HIPAA BAA (for EU and US data centers), EU AI Act high-risk documentation (for EU subsidiary).

## 8. Risks and Compliance

<callout-box data-variant="warning" data-title="2026 Healthcare AI Risks">

- **Data insufficiency and bias.** Most AI models trained on US/EU white populations; Turkish performance may differ. **Local validation** is mandatory.
- **Clinical adoption friction.** Clinician trust builds slowly; training and change management are critical.
- **Liability ambiguity.** When AI errs, who is responsible — vendor, hospital, or clinician? Contractual clarity is essential.
- **Drift.** Patient populations evolve; model performance erodes. Continuous monitoring required.
- **Cross-border transfer.** US cloud AI vendors create triple risk (KVKK + HIPAA + EU AI Act).
- **Cyber security.** Medical device cyber attacks are rising; IEC 62304 + OWASP medical device matter.
- **Lack of explainability.** Black-box AI output is not accepted by clinicians; XAI must be designed in from day one.

</callout-box>

### Healthcare AI Compliance Checklist

FDA / EMA / Ministry of Health approval verification; HL7 FHIR compatibility test; KVKK Article 6; DPIA; updated explicit consent; ISO 27799 audit; IEC 62304; HIPAA BAA (US vendors); SCC/BCR (EU vendors); EU AI Act high-risk documentation (EU scope); PCCP; Turkish population clinical validation (6 months minimum); audit log retention (10 years minimum for health data); cyber security penetration test; clinician training and change management.

## 9. Frequently Asked Questions

<callout-box data-variant="answer" data-title="What should we look for when buying an AI radiology product?">

12 critical checks: FDA/EMA approval + indication; Ministry of Health registration; HL7 FHIR; Turkish population validation; ISO 27799 + IEC 62304 certification; PCCP plan; clinical evidence (peer-reviewed publication); XAI documentation; 10-year audit log retention; cyber security pen test; contractual liability clarity; Turkey deployment reference.

</callout-box>

<callout-box data-variant="answer" data-title="When does the EU AI Act become binding for us?">

If any of three: EU service delivery (e.g., medical tourism — EU patients); using an EU vendor that processes data in the EU; planning to ship a healthcare AI product in the EU. 2 August 2026 is the effective date for high-risk systems.

</callout-box>

<callout-box data-variant="answer" data-title="What does KVKK Article 6 mean in practice for health data?">

Health data is special-category personal data. Processing requires explicit consent (per patient, per AI processing purpose) or processing by health professionals under confidentiality. Third-party vendor: either fresh explicit consent or a data controller/data processor contract. DPIA is mandatory.

</callout-box>

<callout-box data-variant="answer" data-title="Can ambient AI scribe be used in Turkey?">

Yes, with care. KVKK: explicit consent (voice recording + AI processing). Vendor contract: data processor role, audit log, retention policy. Data residency: prefer Turkey or EU. Clinician must approve the final note — the AI scribe only drafts.

</callout-box>

<callout-box data-variant="answer" data-title="Why is Turkish population validation critical?">

Most AI models train on US or EU white populations. In Turkish populations, disease patterns, genetics, and dietary habits differ — model performance can drop 5-15%. Mammography example: 92% accuracy on US data may be 84% in Turkish population. Local validation (6 months, 1000+ cases) closes that gap.

</callout-box>

<callout-box data-variant="answer" data-title="If a vendor AI errs, who is responsible?">

Multi-party liability: vendor (medical device manufacturer) for product quality and approval; hospital (operator) for installation, training, monitoring, audit; clinician (user) for the final clinical decision. Contractual clarity is essential; vendor liability insurance, hospital malpractice insurance, and AI output framed as **decision support** (not autonomous decision) are all vital.

</callout-box>

<callout-box data-variant="answer" data-title="What is HL7 FHIR compatibility and why is it critical?">

FHIR (Fast Healthcare Interoperability Resources) is the modern healthcare data standard. Hospital HIS, PACS, lab systems, AI products all "speak the same language" — integration takes days, not weeks. Turkish hospital FHIR maturity has accelerated over three years; AI vendors with FHIR-native designs reduce integration cost by 40-60%.

</callout-box>

<callout-box data-variant="answer" data-title="Where do Turkish pharma's AI investments stand for drug discovery?">

Early-to-mid stage. Abdi İbrahim, Bilim İlaç, Atabay run pilots in hit identification and ADMET prediction. More mature in clinical trial recruitment and Real World Evidence (RWE). Highest-ROI for Turkish pharma: **e-Nabız RWE + AI** — leveraging local clinical evidence generation.

</callout-box>

## 10. Next Steps

To clarify your AI roadmap for a Turkish hospital group or health-tech company:

1. **Healthcare AI compliance workshop.** Current AI projects + FDA/EU AI Act/KVKK gap + Turkish population validation needs in a 6-hour session. Output: 12-week compliance roadmap.
2. **Vendor due diligence.** 12-criterion evaluation of candidate AI radiology/pathology/triage products + Turkey deployment reference check. Output: vendor scorecard + contract recommendations.
3. **Clinical validation program.** 6-month Turkish population validation protocol + peer-reviewed publication strategy. Output: validation plan + IRB application documentation.

Reach out via the contact form on the site.

<references-list data-items="[{&quot;title&quot;:&quot;FDA AI/ML-Enabled Medical Devices List&quot;,&quot;url&quot;:&quot;https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-aiml-enabled-medical-devices&quot;,&quot;author&quot;:&quot;FDA&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;US Food and Drug Administration&quot;},{&quot;title&quot;:&quot;FDA Clinical Decision Support Software Guidance (2026 Revision)&quot;,&quot;url&quot;:&quot;https://www.fda.gov/regulatory-information/search-fda-guidance-documents/clinical-decision-support-software&quot;,&quot;author&quot;:&quot;FDA&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;FDA&quot;},{&quot;title&quot;:&quot;FDA Predetermined Change Control Plan Guidance&quot;,&quot;url&quot;:&quot;https://www.fda.gov/medical-devices/software-medical-device-samd/predetermined-change-control-plans-machine-learning-enabled-medical-devices&quot;,&quot;author&quot;:&quot;FDA&quot;,&quot;publishedAt&quot;:&quot;2024&quot;,&quot;publisher&quot;:&quot;FDA&quot;},{&quot;title&quot;:&quot;FDA Real World Evidence Guidance&quot;,&quot;url&quot;:&quot;https://www.fda.gov/science-research/science-and-research-special-topics/real-world-evidence&quot;,&quot;author&quot;:&quot;FDA&quot;,&quot;publishedAt&quot;:&quot;2024&quot;,&quot;publisher&quot;:&quot;FDA&quot;},{&quot;title&quot;:&quot;EU AI Act&quot;,&quot;url&quot;:&quot;https://artificialintelligenceact.eu/&quot;,&quot;author&quot;:&quot;European Commission&quot;,&quot;publishedAt&quot;:&quot;2024-03-13&quot;,&quot;publisher&quot;:&quot;EU&quot;},{&quot;title&quot;:&quot;EU Medical Device Regulation (MDR 2017/745)&quot;,&quot;url&quot;:&quot;https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32017R0745&quot;,&quot;author&quot;:&quot;EU&quot;,&quot;publishedAt&quot;:&quot;2017&quot;,&quot;publisher&quot;:&quot;EU&quot;},{&quot;title&quot;:&quot;IMDRF SaMD Framework&quot;,&quot;url&quot;:&quot;https://www.imdrf.org/documents/software-medical-device-samd&quot;,&quot;author&quot;:&quot;IMDRF&quot;,&quot;publishedAt&quot;:&quot;2024&quot;,&quot;publisher&quot;:&quot;IMDRF&quot;},{&quot;title&quot;:&quot;ISO 27799 - Health Informatics Security Management&quot;,&quot;url&quot;:&quot;https://www.iso.org/standard/62777.html&quot;,&quot;author&quot;:&quot;ISO&quot;,&quot;publishedAt&quot;:&quot;2016&quot;,&quot;publisher&quot;:&quot;ISO&quot;},{&quot;title&quot;:&quot;IEC 62304 - Medical Device Software Lifecycle&quot;,&quot;url&quot;:&quot;https://www.iso.org/standard/38421.html&quot;,&quot;author&quot;:&quot;IEC&quot;,&quot;publishedAt&quot;:&quot;2015&quot;,&quot;publisher&quot;:&quot;IEC&quot;},{&quot;title&quot;:&quot;HL7 FHIR Specification&quot;,&quot;url&quot;:&quot;https://hl7.org/fhir/&quot;,&quot;author&quot;:&quot;HL7&quot;,&quot;publishedAt&quot;:&quot;2025&quot;,&quot;publisher&quot;:&quot;HL7 International&quot;},{&quot;title&quot;:&quot;Turkish Ministry of Health Digital Health&quot;,&quot;url&quot;:&quot;https://saglik.gov.tr/&quot;,&quot;author&quot;:&quot;Ministry of Health&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Republic of Turkiye&quot;},{&quot;title&quot;:&quot;e-Nabız&quot;,&quot;url&quot;:&quot;https://enabiz.gov.tr/&quot;,&quot;author&quot;:&quot;Ministry of Health&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Republic of Turkiye&quot;},{&quot;title&quot;:&quot;KVKK - Law No. 6698 Article 6&quot;,&quot;url&quot;:&quot;https://www.kvkk.gov.tr/&quot;,&quot;author&quot;:&quot;KVKK&quot;,&quot;publishedAt&quot;:&quot;2016-04-07&quot;,&quot;publisher&quot;:&quot;Republic of Turkiye&quot;},{&quot;title&quot;:&quot;HIPAA Privacy Rule&quot;,&quot;url&quot;:&quot;https://www.hhs.gov/hipaa/index.html&quot;,&quot;author&quot;:&quot;US HHS&quot;,&quot;publishedAt&quot;:&quot;2024&quot;,&quot;publisher&quot;:&quot;US Department of Health and Human Services&quot;},{&quot;title&quot;:&quot;The Imaging Wire&quot;,&quot;url&quot;:&quot;https://theimagingwire.com/&quot;,&quot;author&quot;:&quot;The Imaging Wire&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;The Imaging Wire&quot;},{&quot;title&quot;:&quot;STAT News - Health AI&quot;,&quot;url&quot;:&quot;https://www.statnews.com/category/health-tech/&quot;,&quot;author&quot;:&quot;STAT News&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;STAT News&quot;},{&quot;title&quot;:&quot;Faegre Drinker - Health AI Regulatory&quot;,&quot;url&quot;:&quot;https://www.faegredrinker.com/en/services/health-care&quot;,&quot;author&quot;:&quot;Faegre Drinker&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Faegre Drinker&quot;},{&quot;title&quot;:&quot;Orrick - Digital Health&quot;,&quot;url&quot;:&quot;https://www.orrick.com/en/practices/life-sciences&quot;,&quot;author&quot;:&quot;Orrick&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Orrick&quot;},{&quot;title&quot;:&quot;Aidoc&quot;,&quot;url&quot;:&quot;https://www.aidoc.com/&quot;,&quot;author&quot;:&quot;Aidoc&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Aidoc Medical&quot;},{&quot;title&quot;:&quot;Viz.ai&quot;,&quot;url&quot;:&quot;https://www.viz.ai/&quot;,&quot;author&quot;:&quot;Viz.ai&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Viz.ai&quot;},{&quot;title&quot;:&quot;GE HealthCare AI&quot;,&quot;url&quot;:&quot;https://www.gehealthcare.com/insights/topics/ai&quot;,&quot;author&quot;:&quot;GE HealthCare&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;GE HealthCare&quot;},{&quot;title&quot;:&quot;Siemens Healthineers AI&quot;,&quot;url&quot;:&quot;https://www.siemens-healthineers.com/digital-health-solutions&quot;,&quot;author&quot;:&quot;Siemens Healthineers&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Siemens Healthineers&quot;},{&quot;title&quot;:&quot;Philips Healthcare AI&quot;,&quot;url&quot;:&quot;https://www.philips.com/healthcare/innovation/artificial-intelligence&quot;,&quot;author&quot;:&quot;Philips&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Philips&quot;},{&quot;title&quot;:&quot;Acibadem Healthcare Group&quot;,&quot;url&quot;:&quot;https://www.acibadem.com.tr/&quot;,&quot;author&quot;:&quot;Acibadem&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Acibadem&quot;},{&quot;title&quot;:&quot;Memorial Healthcare Group&quot;,&quot;url&quot;:&quot;https://www.memorial.com.tr/&quot;,&quot;author&quot;:&quot;Memorial&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Memorial Healthcare Group&quot;},{&quot;title&quot;:&quot;Medipol University Hospital&quot;,&quot;url&quot;:&quot;https://www.medipol.com.tr/&quot;,&quot;author&quot;:&quot;Medipol&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Medipol Healthcare&quot;},{&quot;title&quot;:&quot;ASELSAN Health&quot;,&quot;url&quot;:&quot;https://www.aselsan.com.tr/&quot;,&quot;author&quot;:&quot;ASELSAN&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;ASELSAN&quot;},{&quot;title&quot;:&quot;Abdi Ibrahim Pharma&quot;,&quot;url&quot;:&quot;https://www.abdiibrahim.com.tr/&quot;,&quot;author&quot;:&quot;Abdi Ibrahim&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Abdi Ibrahim&quot;},{&quot;title&quot;:&quot;Bilim Pharma&quot;,&quot;url&quot;:&quot;https://www.bilimilac.com.tr/&quot;,&quot;author&quot;:&quot;Bilim Pharma&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Bilim Pharma&quot;},{&quot;title&quot;:&quot;Atabay Pharma&quot;,&quot;url&quot;:&quot;https://www.atabay.com.tr/&quot;,&quot;author&quot;:&quot;Atabay&quot;,&quot;publishedAt&quot;:&quot;2026&quot;,&quot;publisher&quot;:&quot;Atabay Pharma&quot;},{&quot;title&quot;:&quot;WHO Global Strategy on Digital Health&quot;,&quot;url&quot;:&quot;https://www.who.int/health-topics/digital-health&quot;,&quot;author&quot;:&quot;WHO&quot;,&quot;publishedAt&quot;:&quot;2024&quot;,&quot;publisher&quot;:&quot;World Health Organization&quot;}]"></references-list>

---

This is a living document; the FDA AI clearance list, EU AI Act implementation notes, KVKK decisions, and Turkish hospital deployments shift materially each quarter, so it is **updated quarterly**.