Skip to main content

Table 1 Signals for Example #1.

From: A Dynamic Neuro-Fuzzy Model Providing Bio-State Estimation and Prognosis Prediction for Wearable Intelligent Assistants

Female client with stroke-induced disability a large-scale model with 16 types of potential input events, 12 states to estimate, 5 outputs, and 3 outcomes.

Problem statement: An older woman presents with stroke-induced disability (4 months post-stroke) that includes mild functional limitations to gait and posture, and significant impairment of the right arm and hand and of speech production. She also presents with mild osteoarthritis that affects her hips and knees. Released from outpatient care and living alone, her current "prescriptions" include three types of medication doses (for general joint and skeletal health, for pain from arthritis, and for spasticity), and three types of activities suggested by her former therapist (walking/cycling, hand operation, and oral communication). She also has two important weekly events: a visit most Sundays from her daughter (who is a nurse), and a visit most Tuesday's to the local community center (transportation is provided). She regularly uses a PDA-cellphone and a desktop computer (both set up by her other daughter who is an engineer, but lives in another state), and prefers to use an IP videoconferencing package to tele-visit with either of her daughters. Thus she is a good candidate for an assistive WIA.

Inputs (and MF example)

States (and MF example)

Outputs

Outcomes

Facts:

- Age (is old)

- Initial Stroke (is severe)

- Osteoarthritis (is mild)

Contexts:

- VisitDaughter (is full)

- VisitCommCenter (is full)

- LocationByGPS (is outside)

- TeleVisitDaught (is active)

- TimeOfDay (is morning)

- NovelEvent (is negative)

Interventions (Meds or Activity)

- PillsOsteo (is right-dose)

- PillsPain (is high-dose; conc)

- PillsSpast (is 2-pills; conc)

- Walking (is good)

- Cycling (is good-quality)

- Speech (is good-duration)

- Keyboard (is good-session)

Degree of Impairment:

- Gait (is faster)

- Balance (is better)

- RightArm (is worse)

- RightHand (is better)

- Speech (is improved)

Physiologic:

- RestingHR (is higher)

- RestingBP (is higher)

- BoneJointHealth (is low )

Other ("Degree of ..."):

- Pain (is high)

- RiskFalling (is high)

- Motivation (is high)

- SleepAtNight (is restful)

Communication [Φ(Speech, Pain)]

HandROM [Φ(Hand)]

FIM [Φ(Arm, Hand, Balance, Speech, Pain)]

RiskFracture [Φ(BJ-Health, Risk-Falling)]

Adherence [Φ(Motivation, Pain, Sleep)]

GenHealth [Φ(all impairment \physiologic states)]

Participation [Φ(Communication., Gait]

QualityLife [Φ(Weekly-Pain, FIM, Speech, Gait, Adherence, Hand-ROM)]

  1. Notes: while one MF value is shown for each input or state, typically there are additional ones. Use hedges such as "not" or "very" or "more-or-less" can lower the number of MFs (and thus parameters) associated with a linguistic variable.
  2. Key abbreviations: MF: membership function; GPS: Global Positioning System; HR: heart rate; BP: blood pressure; FIM: Functional Independence Measure [21].