Research

The problem

Neck, back, shoulder and knee pain are the most common reasons to visit a Korean medicine hospital, yet imaging often fails to explain the pain and some patients become chronic. I want measurable answers to who becomes chronic and which patients benefit, and how much, from Korean medicine treatments including acupuncture, pharmacopuncture and herbal medicine.

Data and methods

• Multicenter randomized controlled trials and single-center pilot trials (painful diabetic neuropathy, intractable neuropathic pain, chronic neck pain)
• Inpatient and outpatient cohorts at the Spine & Joint Center and EHR-based retrospective studies (acute neck and low back pain after traffic accidents)
• Cross-sectional analyses of public big data such as KNHANES
• Characterizing nociplastic pain and exploring predictors of chronification

The problem

Acupuncture, pharmacopuncture and acupotomy have relied on surface anatomy and tactile feedback. Visualizing the target with ultrasound can improve precision and safety, but standard protocols and safety and efficacy evidence tailored to Korean medicine procedures are still limited.

Data and methods

• Region-specific protocols (cervical, lumbar, shoulder, hand) for ultrasound-guided procedures and accumulated clinical cases
• Case reports and clinical studies collecting safety and outcome data (e.g., ultrasound-guided vagus nerve-targeted pharmacopuncture)
• A deep-learning system for automatic ultrasound image recognition and procedure assistance in lumbar acupotomy
• Dissemination through continuing education and congress hands-on sessions

The problem

Range of motion and posture are basic measures in musculoskeletal care, but goniometry is slow and varies between raters. Quantifying movement with a single camera would allow clinic and research to share the same yardstick.

Data and methods

• Reliability and validity of single-camera markerless 3D motion analysis (RGB-D sensors) against goniometry in healthy adults and patients
• Clinical evidence for listing video-based muscle performance assessment as a new health technology
• Clinical validation within a project developing an image-analysis device for posture, range of motion and spinal alignment in Chuna therapy
• Systematic review summarizing the performance of existing markerless systems

The problem

About 85% of chronic low back pain is non-specific, with no cause identified by imaging or biomarkers, and it is a multidimensional condition in which structure, function and emotion intertwine. Treating it under a single label lowers response and promotes chronicity. The goal is to digitize the multidimensional information once handled by the four examinations, identify subtypes, and predict prognosis.

Data and methods

• Multimodal datasets combining images (posture, facial expression), voice, questionnaires, palpation and biosignals such as pulse waves and heart-rate variability
• Representation learning (deep autoencoders, clustering) to derive clinical subtypes, with explainable AI to expose the reasoning
• Chronification and prognosis models using EHR and real-world data
• Shared data and models with adjacent projects on facial palsy assessment and lumbar ultrasound image recognition

This is an ongoing National Research Foundation project (2026–2031); the subtype and prognosis models are still under development.

The problem

Individual studies do not change practice unless they are organized into a usable form. For chronic intractable pain that medication alone controls poorly, such as postherpetic neuralgia and fibromyalgia, I systematically appraise the evidence for Korean medicine treatment and turn it into recommendations.

Data and methods

• Systematic reviews and meta-analyses, GRADE evidence and recommendation grading, Evidence to Decision frameworks
• Surveys of clinical practice, expert consensus (Delphi) and review by related societies
• Clinical pathways developed alongside the guideline and piloted in practice
• Large language models as assistants for literature extraction and evidence tables, with human verification

Published: Korean Medicine CPG for Herpes Zoster and Postherpetic Neuralgia (2026)

The guideline presents 32 recommendations, separately for herpes zoster and postherpetic neuralgia, covering acupuncture, electroacupuncture, fire needling, moxibustion, bloodletting cupping and herbal medicine as single or combined Korean medicine treatments and as integrative treatment alongside conventional care. The clinical algorithm includes differential diagnoses and criteria for referral when high-risk complications are suspected. Development followed systematic literature review, a 2024 survey of 1,122 Korean medicine doctors, expert review, GRADE-based grading and review by related societies.

• Full text, leaflet and infographics: National Clearinghouse for Korean Medicine
• Introductory article: Akom News, Sep 2026

In development: Korean Medicine CPG and Clinical Pathway for Fibromyalgia (2026–2027)

Consolidating the evidence for Korean medicine treatment of fibromyalgia and developing a clinical pathway that includes newer interventions such as ultrasound-guided pharmacopuncture.