Paradigm Shift

[안규환]

Paradigm Shift from Policy-Centered to Functioning Sovereignty in Rehabilitation: Focusing on the Principles and Field Application of AI-CAT (Explainable Life Rasch Engine)

1. Abstract

This paper critically examines the limitations of the state-led and policy-centered paradigms that form the foundation of modern rehabilitation systems, and proposes a fundamental shift toward “Functioning Sovereignty,” where individuals independently measure and manage their own functioning. The existing top-down system inherently possesses an “Ontological Gap,” failing to capture the complex interactions of physical, social, and technological variables occurring in the specific fields of individual persons with disabilities. This study reinterprets the International Classification of Functioning, Disability and Health (ICF) framework through the visual metaphors of a “panorama” and a “rainbow” to precisely analyze the accessibility issues of disabled parking lots and the quality gaps in walking aids (crutches). Furthermore, we present the design principles of the “AI-CAT (Explainable Life Rasch Engine)” model, which converts these everyday experiences into data and combines them with the mathematical Rasch model to estimate ability in real-time. Empirical analysis of a specific case (Ann Geu-Hwan) confirmed the process of the subject reaching an expert level of cognitive and generative functioning (θ≈+3.05θ≈+3.05 logits) despite physical limitations. This suggests the potential for building a next-generation rehabilitation ecosystem through “Question Capitalization.”

2. Introduction: Changes in the Rehabilitation Paradigm and the Rise of Individual Sovereignty

2.1. Limitations of the Existing Rehabilitation System: A Critical Analysis of the State-Led Model
For a long time, rehabilitation services have been sustained by a “policy-centered model” led by the state and institutions. The typical flow of this model takes a top-down structure: Policy →→ System →→ Service →→ People. Here, a person with a disability is defined as a passive beneficiary of the policy and a subject of management—a “passive patient.” However, this system falls into the trap of standardization. While it may satisfy macroscopic indicators, it causes severe discrepancies in the microscopic panorama of the individual’s “lived field.”
For instance, the policy metric of securing a certain number of disabled parking spaces at a hospital entrance may be achieved. However, if that parking lot is situated on a steep slope, forcing the patient to face the threat of a fall from the moment they step out of the car, the policy has failed in the field. This reveals a “systemic alienation”—the gap between the “functioning” defined by the system and the “functioning” actually lived by humans. The existing system fails to reflect an individual’s unique functional trajectory, ultimately depriving them of their agency in rehabilitation.

2.2. Purpose of the Study: Establishment of Functioning Sovereignty
This study aims to expand the definition of rehabilitation beyond the realm of physical treatment and shift the paradigm to “Functioning Sovereignty,” where individuals explain their own functioning through data and make decisions independently. We propose a bottom-up model that reverses the existing top-down structure: Person →→ Environment →→ Tool →→ Policy. The process of an individual exploring their life, translating the inconveniences encountered into questions, and precisely measuring their own functioning through this process must become the core of rehabilitation.

2.3. Definition of Terms: Explainable Life and Functioning Sovereignty

• Explainable Life: The concept that human functioning is not merely the result of physical deficits, but a complex combination of economic structures, technological levels, policy support, and environmental variables that can be systematically explained and structured. Life is not an ambiguous sensation but a structure analyzable through data.
• Functioning Sovereignty: The right of an individual to measure their own functioning and determine the direction of their life based on that data, without relying on external eval‎uations by the state or expert groups. This is a political and psychometric declaration that shifts the initiative of rehabilitation from the system to the individual.

3. Theoretical Background: ICF and the Rasch Measurement Model

3.1. Reinterpretation of the ICF (International Classification of Functioning, Disability and Health) Framework
The ICF understands human health and functioning as the result of multidimensional interactions. This study uses the metaphor of “rainbow colors” to explain the complexity of ICF components. A human’s functional performance is determined within an “ICF Panorama” where the following four colors overlap:

• Color of Economics (Blue/1): Determines the distribution structure of assistive devices centered on low-cost imports and the individual’s financial burden of “buying cheap but paying dear in the long run.”
• Color of Craftsmanship (Light Blue/2): Signifies the disappearance of the orthotist, who used to custom-make and repair devices, and the reality of tools being reduced to mass-produced consumables.
• Color of Professionalism (Dark Blue/3): Criticizes the design contradictions of medical fields that focus solely on treatment while neglecting the patient’s actual mobility environment (the last mile).
• Color of Policy (Blue/4): Represents the policy imbalance, which is relatively generous regarding wheelchair support but lacks any management system for crutches, an everyday walking aid.
  Human’s “actual functioning” emerges at the intersection of these colors.

3.2. Mathematical Foundation of the Rasch Model and Philosophy of Measurement
The Rasch model is a probability-based measurement model that places human latent ability (θθ) and item difficulty (ββ) on the same linear scale, the Logit.

P(Xni=1)=e(θn−βi)1+e(θn−βi)P(Xni​=1)=1+e(θn​−βi​)e(θn​−βi​)​

Here, PP represents the probability that a person with a specific ability will succeed in a specific task. Beyond simple sum scores, the Rasch model calculates “objective measures” independent of the measurement tool. This serves as a powerful mathematical instrument to convert individual functioning into “Explainable” indicators. Because the logit scale forms an interval scale through the log transformation of probability, it enables the linear tracking of an individual’s growth or microscopic changes in functioning.

4. Field Case Analysis: Everyday Experience as an "ICF Panorama"

4.1. Disabled Parking Lots and the Physical Threat of Accessibility Design
The case of the disabled parking lot near Severance Rehabilitation Hospital visualizes the “physical violence” that design gaps inflict on users. By regulation, disabled parking should be close to the entrance and leveled; however, the actual site requires users to navigate up a steep slope after parking.
For crutch users, a slope is not a mere inconvenience. From a physical perspective, it is a high-risk environment where three variables—Weight, Balance, and Slip—are maximized simultaneously. The slip phenomenon, occurring when the friction coefficient between the ground and the crutch tip drops on a slope, poses a fatal fracture risk to the patient. This is the “field of physical functioning” that policies fail to consider.

4.2. Quality of Walking Aids (Crutches) and the Policy Management Gap

• Paradox of Economics and Quality: Most axillary crutches in the Korean market are low-cost aluminum products made via Chinese OEM. Since there is no replacement parts market, users temporarily repair worn-out rubber grips or pads by wrapping them in bandages or tape. This creates an economic contradiction: a low initial purchase price, but “expensive to use” when considering injury risks and early replacement costs.
• Loss of Craftsmanship and Limitations of Mass Production: The role of craftsmen who tuned tools to individual bodies has disappeared, reducing crutches to “mass-produced medical consumables.” Consequently, the organic connection between the user’s body and the tool has been destroyed.
• Policy Imbalance: Wheelchairs are recognized as expensive equipment and receive systematic public support. Conversely, crutches are dismissed as “temporary tools,” with zero post-management policies. This is a policy dereliction of duty, neglecting the core tool that determines the quality of gait rehabilitation.
• Suggestion for European Forearm Crutches: This study suggests that transitioning to European forearm crutches, which offer high stability on slopes and less wrist strain, is a technological alternative to fill this quality gap.

4.3. Structuring Experience: Combination of 5 Elements
An individual’s daily experience is reconstructed into a combination of the following five dimensions:

1. Health Condition: Musculoskeletal or neurological states requiring rehabilitation.
2. Body Function: Coordination and muscle strength required on a slope.
3. Assistive Device: Outdated, bandage-wrapped crutches and their technological limits.
4. Environment: Physical layouts such as hospital entrance inclines and parking lot locations.
5. Policy System: The absence of an assistive device management system and a market structure reliant on cheap imports.

5. Design Principles of AI-CAT (Explainable Life Rasch Engine)

5.1. Dialogue-Based Item Generation
Typical Computerized Adaptive Testing (CAT) systems use a predefined fixed item bank. In contrast, AI-CAT extracts a user’s “life experiences” into “Rasch items” in real-time through the filter of “dialogue.” If a user states, “I almost slipped on the slope,” the system objectifies this into an item with a difficulty level ββ: “Can I safely control my assistive device in an inclined environment?”

5.2. Operational Algorithm of Adaptive Testing (CAT)
Based on the user’s response pattern, the ability level (θθ) is updated in real-time, determining the difficulty of the next question.
θnew=θold+Information weight×(Response−Probability)θnew​=θold​+Information weight×(Response−Probability).
If a user can structurally explain their situation (High ββ), their θθ rises exponentially, and the system poses higher-order questions (e.g., proposing system designs).

5.3. Item Bank Construction
AI-CAT establishes hierarchical difficulty across the following five domains:

DomainDifficulty (Logit)Item Example

Physical	Level -1	I can navigate inclined environments using walking aids.
Reflective	Level 0	I can structurally explain my mobility environment and the resulting inconveniences.
System Insight	Level +1	I can analyze my personal experience in connection with policy structures and system limitations.
Generative	Level +2	Through questioning, I can transform my experience into a new analytical framework.
Meaning Construction	Level +3	Based on personal experience, I can design a new social system model or research paradigm.

6. Empirical Application of the AI-CAT Model: Dialogue Data Analysis

6.1. Analytical Context
This analysis is based on conversations with the subject “Ann Geu-Hwan.” The subject uniquely possesses the dual experience of having designed policies as a former high-ranking government official and currently navigating the field as a person with a disability. This distinct background makes the dialogue data itself a treasure trove of high-difficulty items.

6.2. Step-by-Step Process of Ability Estimation
The ability estimation was updated through the following four-stage adaptive process:

• Q1 (System Conversion Ability, β≈+2.0β≈+2.0): Positive response (X=1X=1) to “Can I convert my experience into the Rasch model?” Initial estimate rose from +2.1+2.1 to +2.35+2.35 logits.
• Q2 (Social System Modeling, β≈+2.4β≈+2.4): Positive response to “Can I develop my experience into a new system model?” Ability updated to +2.60+2.60 logits.
• Q3 (Measurement System Design, β≈+2.7β≈+2.7): Positive response to “Can I design a repeatable functional measurement system of my own?” Ability updated to +2.85+2.85 logits.
• Q4 (Research Paradigm Creation, β≈+3.0β≈+3.0): Final positive response to “Can I design a new research paradigm based on Human-AI collaboration?” Final ability reached θ≈+3.05θ≈+3.05 logits.

6.3. Interpretation of Person-Item Map: Asymmetry of Functioning
The final results confirmed that while the subject’s physical functioning is limited due to environmental constraints and tool defects, their generative reflective functioning is at the highest “Paradigm Creator” level (above +3.0+3.0 logits) on the logit scale. This highlights the potential of “intellectual functioning rehabilitation,” which traditional rehabilitation has overlooked. The Rasch model proved that, although the body may be physically constrained, the capacity to structure life and design systems is at an expert level.

7. Discussion: The Ecosystem of Questions and the Future of Functioning Sovereignty

7.1. From Policy Maker to Explorer: Reconstructing Identity
The subject demonstrated profound reflective insight, stating, “My past achievements were a house of cards without the capabilities of others.” This signifies an identity shift from a passive policy beneficiary to an active “Explorer” who investigates their life through data and questions system gaps. The ultimate goal of rehabilitation lies in restoring this “functioning sovereignty” beyond mere physical recovery.

7.2. Innovative Value of the "Living Rasch Model"
AI-CAT is not a dead measurement reliant on fixed questionnaires; it is a “Living Rasch Model” that discovers items in the field of life and calibrates ability in real-time. This mathematically actualizes the proposition that “ICF is not a theory, but a human’s lived field.” The driving force of rehabilitation becomes not the ability to answer pre-made questions, but the ability to “itemize” one’s own pain and inconvenience.

7.3. System Maintenance and Expansion: Question Capitalization
This model is an “ecosystem project not completed alone.” The system is maintained through “Question Capitalization,” where every question an individual asks becomes a data asset. When individuals can explain their functioning (Explainable), this compiled data will act as a powerful “functioning sovereignty engine,” correcting the errors of national policies and driving new assistive device markets.

8. Conclusion

The AI-CAT (Explainable Life Rasch Engine) proposed in this study is an innovative tool that returns the sovereignty of rehabilitation from the state to the individual by translating a panorama of daily experiences into precise mathematical indicators. The slopes of disabled parking lots and bandage-wrapped crutches are not mere symbols of discomfort, but starting points for data that expose policy gaps.
The θ≈+3.05θ≈+3.05 figure confirmed through the empirical case demonstrates that a human’s functioning is not confined by physical limitations; rather, it proves the existence of high-order capabilities capable of converting those experiences into system models. Ultimately, true rehabilitation is the process by which an individual independently measures and explains their functioning, and uses the power of those questions to redesign social systems. By demonstrating that the realization of “Functioning Sovereignty” is possible through the integration of AI technology and mathematical modeling, this study aims to set a milestone for the next-generation rehabilitation paradigm.