Bridging WorldsOrder

Originally published in Forbes

Articles

All articles on Forbes by Frank Naeymi-Rad, PhD, MBA, FACMI. All articles on Forbes

Nov 06, 2026

The Feature Dictionary: Defining Disease Through Clinical Terminology

In the early 1980s, Frank Naeymi-Rad joined a NASA-funded project to adapt the Medical Emergency Decision Assistance System for the space shuttle. The author sought to expand MEDAS to support physicians across every clinical specialty using a structured representation called the Feature Dictionary. A major problem arose when physicians described the same clinical finding using different words, abbreviations, and synonyms. To solve this, the team connected these variants to shared, explicitly defined clinical concepts while preserving diagnostic relationships. This approach became known as interface terminology, which links the language physicians use to standardized medical concepts. Today, this foundational principle remains critical as new generations build AI for healthcare and require accurate clinical knowledge.

Oct 06, 2026

An Immigrant, An American Dream, And A Life In Healthcare Innovation

Frank and Theresa recently celebrated fifty years of marriage after first meeting in 1970 while washing dishes in a dorm cafeteria at Illinois State University. Theresa arrived as the American-born daughter of Slovenian refugees, while Frank immigrated from Iran in 1969. Theresa pursued medicine and became an OB/GYN, treating patients across VA and DOD health systems, while Frank became a computer scientist. Theresa's academic research on rat brain cycles influenced Frank's thinking about the importance of timing in healthcare data. Additionally, her struggles with incompatible electronic medical record coding systems between the VA and DOD pushed Frank to focus on ontology and interface terminology. Fifty years later, they reflect on their careers in healthcare innovation and the American Dream as artificial intelligence enters the exam room.

Sep 01, 2026

What A 1968 Paper Asks Of Healthcare AI

Frank Naeymi-Rad credits Warner Slack for inspiring a patient-centric approach, noting Slack's observation that medical records often failed patients. This issue was addressed by Lawrence Weed's 1968 introduction of the Problem-Oriented Medical Record (POMR) and SOAP notes, which organized patient charts by specific problems, a significant improvement. However, the original intent of POMR—to capture clinical reasoning—was often lost, with problem lists becoming tools for billing rather than genuine patient understanding. The article warns that healthcare AI risks repeating this historical error by optimizing for reimbursement over clinical insight. Inspired by Slack, the author emphasizes the critical need for AI to preserve clinical intent and patient meaning, ensuring technology truly serves patients.

Aug 18, 2026

Clinical Data Accuracy: The First Rule Of Healthcare AI

Frank Naeymi-Rad, a doctor and computer scientist, recounts a pivotal healthcare AI project at a VA hospital, emphasizing that clinical data accuracy is paramount. Their team developed an AI to accelerate the Acute Physiology and Chronic Health Evaluation (APACHE) score, which traditionally had a 12-24 hour lag. Collaborating with physicians, they identified 1,500 clinical indicators and meticulously digitized patient data. This allowed them to train a pattern-recognition AI that successfully predicted the 24-hour APACHE score within just four hours. This breakthrough was directly attributed to the precise capture of doctors’ judgment. The author stresses that this foundational principle of accurate clinical data remains critical for contemporary AI.