Hospital systems

Integrations

Every SoleGuard scan flows into the patient's chart and the care team's workflow — no manual data entry.

EH

Epic Hyperspace

EHR · St. Mary's instance

Connected
Patient chart updated
2 seconds ago
Plantar pressure observation
2 seconds ago
Ulcer-risk score
2 seconds ago
Insole order written to flowsheet
just now
Last full sync · today 09:41:18 UTC

Also configured: Cerner Millennium · Athenahealth · MEDITECH Expanse

Referral & alerts

High-risk patients routed to podiatry automatically

High-risk alert sent to Dr. Mensah

Patient: Elena Ramírez · Score 72 · refer to podiatry within 72h

Workflow status

  1. Drafted

    09:41:20

  2. Sent to podiatry queue

    09:41:22

  3. Acknowledged by Dr. Mensah

    09:41:48

  4. 4

    Appointment scheduled

    pending

Standards-based exchange

FHIR Observation resource

Every scan is serialized as a FHIR R4 Observation with the pressure peak as valueQuantity and ulcer-risk + hotspot as components. Compatible with any FHIR-enabled EHR and the SMART on FHIR app launch.

  • LOINC 80394-6 · plantar pressure peak
  • Bound to Patient and Device references
  • HL7 v2.5 ORU^R01 also supported
  • Signed with OAuth 2.0 client credentials
FHIR R4HL7SMART on FHIR
POST /fhir/R4/Observation
FHIR R4
{
  "resourceType": "Observation",
  "id": "soleguard-obs-2087421",
  "status": "final",
  "category": [
    {
      "coding": [
        {
          "system": "http://terminology.hl7.org/CodeSystem/observation-category",
          "code": "exam",
          "display": "Exam"
        }
      ]
    }
  ],
  "code": {
    "coding": [
      {
        "system": "http://loinc.org",
        "code": "80394-6",
        "display": "Plantar pressure peak"
      }
    ]
  },
  "subject": {
    "reference": "Patient/elena-ramirez",
    "display": "Elena Ramírez"
  },
  "effectiveDateTime": "2026-06-29T14:32:11Z",
  "performer": [
    {
      "reference": "Device/soleguard-pad-A7"
    }
  ],
  "valueQuantity": {
    "value": 287,
    "unit": "kPa",
    "system": "http://unitsofmeasure.org",
    "code": "kPa"
  },
  "interpretation": [
    {
      "coding": [
        {
          "system": "http://terminology.hl7.org/CodeSystem/v3-ObservationInterpretation",
          "code": "H",
          "display": "High"
        }
      ]
    }
  ],
  "component": [
    {
      "code": {
        "text": "Ulcer risk score (SoleGuard AI)"
      },
      "valueQuantity": {
        "value": 72,
        "unit": "score"
      }
    },
    {
      "code": {
        "text": "Pressure-time integral"
      },
      "valueQuantity": {
        "value": 142,
        "unit": "kPa·s"
      }
    },
    {
      "code": {
        "text": "Hotspot location"
      },
      "valueString": "2nd metatarsal head, right foot"
    }
  ]
}
Observation · plantar pressure peak · LOINC 80394-6

Model provenance

Foot-type classifier · v0.3

Trained on Takata Y. (2020) — DOI 10.17632/vv3zgwwpk3.1

Research-grade

Subjects

39

Features

24

Folds

5

LR · AUC

0.74

RF · AUC

0.88

GBM · AUC

0.90

Top 5 features (standardized coefficient)

  • Contact area · Midfoot+1.22
  • Peak pressure · Midfoot+0.68
  • Peak pressure · Heel lateral-0.59
  • Peak pressure · Meta3+0.55
  • Peak pressure · Toe1+0.44

Inference runs entirely on-device using a 24-coefficient logistic regression exported from scikit-learn. The classifier predicts flatfoot vs normal arch — a contributing biomechanical signal, not a substitute for clinical diagnosis. Cross-validated on a small single-sex cohort; broader validation is on the roadmap.

All payloads on this page use mock data for demonstration. No PHI is transmitted.