← TheraPetic.AI

Blog

Latest articles and insights

Retrieval-Augmented Generation Over Clinical Guidelines: Practical Architecture for DSM-5-TR, APA and NICE

How to build RAG systems over DSM-5-TR, APA and NICE guidelines using clinical chunking strategies, grounding scores and citation fidelity architecture.

Dr. Patrick Fisher, PhD, LPC, NCC · May 30, 2026
Read Article →

NIST AI Risk Management Framework for Nonprofit Clinical Deployment: Board Oversight, Audit Logging and Model Cards on a Constrained Budget

How nonprofit healthcare organizations can implement the NIST AI Risk Management Framework with board-level oversight, audit logging and model cards without enterprise budgets.

Dr. Patrick Fisher, PhD, LPC, NCC · May 27, 2026
Read Article →

Consent for AI-Assisted Clinical Evaluation: Beyond the Checkbox

Informed consent for AI-assisted clinical evaluation demands far more than a checkbox. Explore disclosure requirements, comprehension validation, and IRB obligations for nonprofit AI screening in 2026.

Dr. Patrick Fisher, PhD, LPC, NCC · May 23, 2026
Read Article →

HANK AI and the Purpose-Specific Assistant Pattern: Why Narrow-Scope LLM Deployment Outperforms General Chatbots in Healthcare Support

HANK AI demonstrates why narrow-scope LLM deployment outperforms general chatbots in healthcare intake through closed RAG, hard guardrails and clinical outcome metrics.

Dr. Patrick Fisher, PhD, LPC, NCC · May 20, 2026
Read Article →

Large Language Models in Clinical Intake: Where the Evidence Actually Stands

A systematic review of validated evidence on LLM-assisted clinical intake, from triage accuracy to decision support limits, drawing on JAMA, NEJM AI and Lancet Digital Health.

Dr. Patrick Fisher, PhD, LPC, NCC · May 16, 2026
Read Article →

Algorithmic Bias in Mental Health Triage: The Subgroup Validation Gap

Training set demographics rarely match deployment populations in mental health AI. Here is how subgroup validation, equalized odds and calibration testing close the gap.

Dr. Patrick Fisher, PhD, LPC, NCC · May 13, 2026
Read Article →

Biometric Authentication for Handler-Dog Team Documentation: Privacy-Preserving Verification at the Edge of ADA Law

Privacy-preserving biometric verification for handler-dog teams: template storage, liveness detection, ADA documentation non-requirement doctrine and fair deployment.

Dr. Patrick Fisher, PhD, LPC, NCC · May 9, 2026
Read Article →

FHIR R4 Integration for Nonprofit Clinical Workflows Without EHR Partnership

FHIR R4 integration for nonprofits without EHR partnerships: SMART on FHIR authorization, Patient/Observation/Questionnaire modeling and real interop pitfalls.

Dr. Patrick Fisher, PhD, LPC, NCC · May 6, 2026
Read Article →

Clinical Decision Support vs Clinical Decision Making: Where FDA SaMD Classification Draws the Line

FDA's SaMD framework determines whether AI clinical tools need 510(k) clearance or qualify as exempt CDS. Here is where the line actually sits in 2026.

Dr. Patrick Fisher, PhD, LPC, NCC · May 2, 2026
Read Article →

Computer Vision for Service Dog Public Access Assessment: A Technical Review of Canine Pose Estimation Models

A technical review of DeepLabCut and SLEAP canine pose estimation models applied to service dog public access evaluation, including behavioral markers and single-camera limits.

Dr. Patrick Fisher, PhD, LPC, NCC · April 29, 2026
Read Article →

ADA Two-Question Rule and the Case Against AI Documentation Checks

Why AI verification of service dog documentation violates ADA intent despite technical feasibility, exploring legal analysis and disability rights perspectives.

Dr. Patrick Fisher, PhD, LPC, NCC · April 25, 2026
Read Article →

HIPAA-Compliant Screening Infrastructure: What the BAA Chain Actually Requires

Building HIPAA-compliant AI screening infrastructure requires cascading Business Associate Agreements through every component, from LLM vendors to vector databases, with technical controls that implement minimum necessary standards in prompt design.

Dr. Patrick Fisher, PhD, LPC, NCC · April 22, 2026
Read Article →

Retrieval-Augmented Generation Over Clinical Guidelines: Practical Architecture for Mental Health AI Systems

Implementing RAG systems over clinical guidelines requires specialized chunking strategies, grounding scores, and hallucination mitigation to support Licensed Clinical Doctors in diagnostic workflows.

Dr. Patrick Fisher, PhD, LPC, NCC · April 22, 2026
Read Article →

HIPAA Audit Logging Requirements for Clinical AI Systems Under the 21st Century Cures Act

Clinical AI systems under HIPAA require comprehensive audit logging that satisfies OCR breach response expectations and 21st Century Cures Act information blocking provisions.

Dr. Patrick Fisher, PhD, LPC, NCC · April 22, 2026
Read Article →