Pittsburgh|USA, Delhi|India, Nairobi|Kenya, Cape Town|South Africa, Dubai|UAE, Saudi Arabia, Jordan, Nepal

Rijuven CardioSleeve® — Clinical Validation Dashboard

Rijuven CardioSleeve® Clinical Validation Dashboard — Andhra Pradesh Medtech Hackathon 2026

n = 373 patients · GGH Kurnool · April 2026

373

Total Patients

GGH Kurnool, Apr 2026

EF < 40% (HFrEF)

20.4% prevalence

297

EF ≥ 40% (Normal)

79.6% of cohort

0.942

Best AUC (PEP)

Excellent discrimination

90.8%

Best Sensitivity

MPI OR PEP OR PEP%

97.4%

Best NPV

MPI OR PEP OR PEP%

Dataset Composition

AUC by Parameter (Area Under ROC Curve)

Sensitivity vs Specificity at Standard Thresholds

Study Design: Prospective observational validation at GGH Kurnool (Government General Hospital, Andhra Pradesh). All patients underwent simultaneous Rijuven CardioSleeve® STI measurement and echocardiographic assessment. EF < 40% defined HFrEF (positive case). The 373-patient dataset supersedes the earlier 259-patient interim analysis; all statistics reflect the complete cohort.

Literature Cross-Validation

Study	n	Population	Key STI Finding	CS Alignment
Ramesh Hospitals, India 2019	258	Indian mixed cardiology	Spec 98.79%, Sens 70%	✔ Confirmed
Roos et al. (JCF)	—	LV systolic dysfunction	STIs predict LVSD non-invasively	✔ Confirmed
Kosmicki et al. (CHF)	—	CHF screening	PEP/LVET ratio predicts EF	✔ Confirmed
Dillier et al.	—	AF patients	STIs valid even in AF	✔ Confirmed
Chao et al. (ADHF outcomes)	—	Acute decompensated HF	EMAT predicts hospital discharge	✔ Aligned
Zuber et al. (ANE / AJC 2007)	—	General cardiology	MPI as myocardial performance index	✔ Confirmed
Efstratiadis & Michaels	—	Echo-validated STI	Non-invasive STI vs echo correlation	✔ Confirmed

Pearson Correlation with Echo EF (continuous)

PEP

r = −0.655

AUC = 0.942 ★ Best

p < 0.001

SPI

r = −0.628

AUC = 0.918

p < 0.001

PEP%

r = −0.551

AUC = 0.871

p < 0.001

MPI

r = −0.080

AUC = 0.890

p = 0.125 (NS linear)

MPI Note: MPI Pearson r is non-significant (p=0.125) because MPI behaves as a threshold-type discriminator rather than a continuous linear predictor. Despite low linear correlation, MPI achieves AUC = 0.890 — demonstrating strong non-linear discriminatory power between EF <40% and ≥40% groups. MPI is most useful as a flag/gate parameter rather than for continuous EF estimation.

Performance at Standard Thresholds

Parameter	Standard Threshold	Optimal Threshold	Sensitivity	Specificity	AUC	Youden Index
PEP	≤ 120 ms	114.5 ms	78.9%	94.6%	0.942	0.736
SPI	≤ 0.30	0.38	90.8%	77.1%	0.918	0.679
MPI	≤ 0.45	0.40	65.8%	93.3%	0.890	0.591
PEP%	≤ 15%	14.3%	64.5%	96.0%	0.871	0.604

AUC by Parameter (Horizontal)

Youden Index by Parameter

Sensitivity vs Specificity Trade-off

Twelve parameter combinations were evaluated using OR (any flag = positive), AND (all flags = positive), and ANY2 (≥2 of N flags = positive) logic, ranked by Youden Index. MPI OR PEP OR PEP% is the recommended combination, achieving Sensitivity 90.8% / Specificity 88.9% / NPV 97.4% / Youden 0.797.

All Combination Rankings

#	Combination	Logic	Sens	Spec	PPV	NPV	Youden	TP/FP/TN/FN	Tag
1	MPI OR PEP OR PEP%	OR (3)	90.8%	88.9%	67.6%	97.4%	0.797	69/33/264/7	★ RECOMMENDED
2	SPI OR MPI OR PEP OR PEP%	OR (4)	90.8%	83.5%	58.5%	97.2%	0.743	69/49/248/7	Good sens
3	PEP OR PEP%	OR (2)	84.2%	91.2%	70.3%	96.1%	0.754	64/28/269/12
4	MPI OR PEP	OR (2)	85.5%	88.2%	65.0%	96.3%	0.737	65/35/262/11
5	PEP alone	Single	78.9%	94.6%	78.9%	94.6%	0.736	60/16/281/16
6	SPI OR MPI OR PEP% (ANY2)	ANY2	78.9%	91.2%	68.1%	95.7%	0.701	60/28/269/16
7	SPI alone	Single	90.8%	77.1%	49.3%	97.3%	0.679	69/77/220/7
8	PEP% alone	Single	64.5%	96.0%	74.6%	93.8%	0.604	49/12/285/27
9	MPI alone	Single	65.8%	93.3%	65.8%	93.3%	0.591	50/20/277/26
10	MPI AND PEP AND PEP%	AND (3)	63.2%	97.3%	85.7%	91.4%	0.605	48/8/289/28	High spec
11	SPI AND MPI AND PEP	AND (3)	61.8%	97.0%	84.0%	91.1%	0.588	47/9/288/29
12	PEP OR SPI	OR (2)	90.8%	74.1%	49.3%	97.3%	0.649	69/77/220/7

Youden Index — Top 8 Combinations

PPV vs NPV — Top Combinations

4-Zone Traffic Light System (Abnormal Flags: 0 – 4)

FLAGS

1.4%

Mean EF = 55.3%

n = 221 patients

FLAG

12.7%

Mean EF = 51.8%

n = 63 patients

FLAGS

36.4%

Mean EF = 45.0%

n = 22 patients

FLAGS

77.4%

Mean EF = 36.3%

n = 31 patients

FLAGS

91.7%

Mean EF = 33.9%

n = 36 patients

HF Risk % by Number of Abnormal Flags

Mean Echo EF% by Zone

Patient Distribution Across Zones

Clinical Action by Zone

Flags	Zone	HF Risk	Mean EF	n (%)	Recommended Clinical Action
0	■ GREEN	1.4%	55.3%	221 (59.2%)	Reassure; routine 12-month STI follow-up
1	■ YELLOW-GREEN	12.7%	51.8%	63 (16.9%)	Lifestyle counselling; repeat STI in 6 months
2	■ YELLOW	36.4%	45.0%	22 (5.9%)	Cardiology referral; echocardiogram within 4 weeks
3	■ ORANGE	77.4%	36.3%	31 (8.3%)	Urgent cardiology review; echo within 1 week
4	■ RED	91.7%	33.9%	36 (9.7%)	Immediate HF management; same-day specialist review

Two-Stage CardioSleeve Screening Protocol

Stage 1 — Gate

SPI OR MPI Screen

Any STI flag triggers Stage 2 analysis. Designed as a high-sensitivity first pass.

Sensitivity 96.1%

NPV 98.6% — very few true HF cases missed at gate

Stage 2 — Triage

Count All 4 STI Flags

Total abnormal flags (SPI, MPI, PEP, PEP%) tallied. Best combo: MPI OR PEP OR PEP%.

NPV 97.4%

Youden 0.797 — highest combined discriminatory power

Stage 3 — Action

Zone-Based Clinical Decision

Zone determines urgency: GREEN = routine, YELLOW = referral, ORANGE = urgent, RED = immediate.

Specificity 88.9%

PPV 67.6% — 2 in 3 RED patients confirmed HFrEF on echo

PPV / NPV Across Clinical Settings

Clinical Setting	HF Prevalence	Sensitivity	Specificity	PPV	NPV	Use Case
Primary care / rural PHC	5%	90.8%	88.9%	31.0%	99.4%	Mass population screening
Community clinic	10%	90.8%	88.9%	47.2%	98.8%	Symptomatic walk-in patients
GGH Kurnool (this study)	20.4%	90.8%	88.9%	67.6%	97.4%	Mixed cardiology OPD
Cardiology OPD	35%	90.8%	88.9%	81.3%	94.3%	High-risk referrals
CCU / ICU	60%	90.8%	88.9%	92.6%	84.3%	Acute decompensated HF

PPV & NPV Across Prevalence Settings

Benefit-Risk: True/False Positives per 1,000 Screened

Optimal Thresholds (Youden-Maximizing via ROC)

Parameter	Standard Threshold	Optimal Threshold	Sens at Optimal	Spec at Optimal	Youden Improvement
PEP	≤ 120 ms	≤ 114.5 ms	84.2%	93.9%	+0.083 vs standard
SPI	≤ 0.30	≤ 0.38	90.8%	81.5%	+0.123 vs standard
PEP%	≤ 15%	≤ 14.3%	68.4%	94.9%	+0.029 vs standard
MPI	≤ 0.45	≤ 0.40	78.9%	91.9%	+0.118 vs standard

ABDM (Ayushman Bharat Digital Mission) — India's national digital health infrastructure. CardioSleeve integration stores and shares STI-derived cardiac measurements via ABHA (Ayushman Bharat Health Account), HIE-CM (Health Information Exchange & Consent Manager), and FHIR R4 APIs, enabling interoperability with HMIS, HIS, and EHR systems across Andhra Pradesh.

FHIR R4

Data Standard

HL7 FHIR Release 4

ABHA

Patient Identity

14-digit Health ID

HIE-CM

Data Exchange

Consent-based sharing

UHI

Health Interface

Unified Health Interface

API Endpoints (Click to Expand)

POST /v1/abdm/patients/register ABHA Register patient with ABHA ID ▼

{
  "abhaId": "91-1234-5678-9012",
  "name": { "family": "Reddy", "given": ["Venkata", "Ramana"] },
  "gender": "male",
  "birthDate": "1972-04-15",
  "telecom": [{ "system": "phone", "value": "+91-9876543210" }],
  "address": {
    "district": "Kurnool",
    "state": "Andhra Pradesh",
    "country": "IN"
  },
  "identifier": [{
    "system": "https://abdm.gov.in/abha",
    "value": "91-1234-5678-9012"
  }]
}

abhaId — 14-digit Ayushman Bharat Health Account number (format: XX-XXXX-XXXX-XXXX)
Response — Returns FHIR Patient resource with server-assigned id for use in subsequent STI observations

POST /v1/fhir/r4/Observation FHIR R4 Submit CardioSleeve STI measurement ▼

{
  "resourceType": "Observation",
  "status": "final",
  "category": [{ "coding": [{
    "system": "http://terminology.hl7.org/CodeSystem/observation-category",
    "code": "vital-signs"
  }]}],
  "code": {
    "coding": [{
      "system": "http://snomed.info/sct",
      "code": "251289006",
      "display": "Systolic Time Interval measurement"
    }],
    "text": "CardioSleeve STI Panel"
  },
  "subject": { "reference": "Patient/abha-91-1234-5678-9012" },
  "effectiveDateTime": "2026-04-04T09:30:00+05:30",
  "device": { "display": "Rijuven CardioSleeve v2.1" },
  "component": [
    {
      "code": { "coding": [{ "code": "59010-6", "system": "http://loinc.org",
                             "display": "Pre-Ejection Period" }] },
      "valueQuantity": { "value": 105, "unit": "ms" },
      "interpretation": [{ "coding": [{ "code": "N", "display": "Normal" }] }],
      "referenceRange": [{ "high": { "value": 120, "unit": "ms" } }]
    },
    {
      "code": { "coding": [{ "code": "cs-spi", "system": "https://rijuven.com/codes",
                             "display": "Systolic Performance Index" }] },
      "valueQuantity": { "value": 0.27, "unit": "ratio" },
      "interpretation": [{ "coding": [{ "code": "N" }] }],
      "referenceRange": [{ "high": { "value": 0.30 } }]
    },
    {
      "code": { "coding": [{ "code": "59408-2", "system": "http://loinc.org",
                             "display": "Myocardial Performance Index (Tei)" }] },
      "valueQuantity": { "value": 0.42, "unit": "ratio" },
      "interpretation": [{ "coding": [{ "code": "N" }] }],
      "referenceRange": [{ "high": { "value": 0.45 } }]
    },
    {
      "code": { "coding": [{ "code": "cs-pep-pct", "system": "https://rijuven.com/codes",
                             "display": "Heart-Rate Corrected PEP%" }] },
      "valueQuantity": { "value": 13.2, "unit": "%" },
      "interpretation": [{ "coding": [{ "code": "N" }] }],
      "referenceRange": [{ "high": { "value": 15 } }]
    }
  ],
  "extension": [
    { "url": "https://rijuven.com/fhir/StructureDefinition/cs-zone",
      "valueString": "GREEN" },
    { "url": "https://rijuven.com/fhir/StructureDefinition/cs-hf-risk",
      "valueDecimal": 0.014 }
  ]
}

component[PEP] — LOINC 59010-6 · threshold ≤ 120 ms
component[SPI] — Rijuven proprietary code · threshold ≤ 0.30
component[MPI] — LOINC 59408-2 (Tei Index) · threshold ≤ 0.45
component[PEP%] — Rijuven proprietary code · threshold ≤ 15%
extension[cs-zone] — GREEN / YELLOW / ORANGE / RED traffic light assignment
extension[cs-hf-risk] — HF probability (0.0–1.0) based on flag count

POST /v0.5/consent-requests/init HIE-CM Request patient consent for data sharing ▼

{
  "requestId": "uuid-consent-20260404-001",
  "timestamp": "2026-04-04T09:31:00.000Z",
  "consent": {
    "purpose": {
      "text": "Care Management",
      "code": "CAREMGT",
      "system": "http://terminology.hl7.org/CodeSystem/v3-ActReason"
    },
    "patient": { "id": "91-1234-5678-9012@sbx.abdm.gov.in" },
    "hiu": { "id": "GGH-KURNOOL-HIU-001" },
    "requester": {
      "name": "GGH Kurnool Cardiology Department",
      "identifier": { "type": "REGNO", "value": "MCI-AP-001234" }
    },
    "hiTypes": ["DiagnosticReport", "Observation", "Condition"],
    "permission": {
      "accessMode": "VIEW",
      "dateRange": {
        "from": "2024-01-01T00:00:00.000Z",
        "to": "2026-12-31T00:00:00.000Z"
      },
      "dataEraseAt": "2027-12-31T00:00:00.000Z",
      "frequency": { "unit": "HOUR", "value": 1, "repeats": 0 }
    }
  }
}

purpose.code — CAREMGT = Care Management (required ABDM consent type)
hiu.id — Health Information User ID registered with NHA
hiTypes — FHIR resource types requested (must match data types submitted)
dataEraseAt — Mandatory data deletion timestamp per ABDM privacy policy

GET /v1/fhir/r4/Observation?patient={abhaId}&code=cs-sti-panel&_sort=-date FHIR R4 Retrieve historical STI readings for longitudinal tracking ▼

// Query parameters
patient  = "91-1234-5678-9012"   // ABHA ID
code     = "cs-sti-panel"        // CardioSleeve panel code
_sort    = "-date"               // Newest first
_count   = 10                    // Max records returned
category = "vital-signs"

// Example Response Bundle
{
  "resourceType": "Bundle",
  "type": "searchset",
  "total": 3,
  "entry": [{
    "resource": {
      "resourceType": "Observation",
      "id": "obs-20260404-001",
      "status": "final",
      "effectiveDateTime": "2026-04-04T09:30:00+05:30",
      "component": [ /* STI values */ ],
      "extension": [{ "url": "cs-zone", "valueString": "GREEN" }]
    }
  }]
}

Returns — FHIR Bundle with all matching Observations (newest first)
Trending — Enables longitudinal STI monitoring per patient across visits
Authorization — Requires valid Bearer token from HIE-CM after patient consent is granted

POST /v1/fhir/r4/DiagnosticReport FHIR R4 Submit full CardioSleeve diagnostic report to HMIS ▼

{
  "resourceType": "DiagnosticReport",
  "status": "final",
  "category": [{ "coding": [{ "system": "http://loinc.org",
                               "code": "LP29684-5", "display": "Cardiology" }]}],
  "code": {
    "coding": [{ "system": "https://rijuven.com/codes",
                 "code": "cs-cardiac-report",
                 "display": "CardioSleeve Cardiac Function Report" }]
  },
  "subject": { "reference": "Patient/abha-91-1234-5678-9012" },
  "effectiveDateTime": "2026-04-04T09:30:00+05:30",
  "result": [{ "reference": "Observation/obs-20260404-001" }],
  "conclusion": "0 of 4 parameters abnormal. Zone: GREEN. HF risk: 1.4%. Routine follow-up advised.",
  "conclusionCode": [{ "coding": [{
    "system": "http://snomed.info/sct",
    "code": "57774000",
    "display": "Normal cardiac function"
  }]}],
  "presentedForm": [{
    "contentType": "application/pdf",
    "title": "CardioSleeve_Report_20260404.pdf",
    "url": "https://hmis.gghkurnool.ap.gov.in/reports/cs-20260404-001.pdf"
  }]
}

conclusion — Human-readable summary for HMIS/EHR display
conclusionCode — SNOMED CT code for structured downstream HIS processing
presentedForm — Link to PDF report stored in hospital HMIS document repository

POST /api/uhi/v0/search UHI Discover CardioSleeve service providers in Andhra Pradesh ▼

{
  "context": {
    "domain": "UHI:Teleconsultation",
    "country": "IND",
    "city": "Kurnool",
    "action": "search",
    "timestamp": "2026-04-04T09:00:00.000Z"
  },
  "message": {
    "intent": {
      "fulfillment": { "type": "CardioSleeve-STI-Screening" },
      "item": {
        "descriptor": {
          "name": "Systolic Time Interval Cardiac Screening",
          "code": "cs-sti-screening"
        }
      },
      "provider": {
        "locations": [{
          "gps": "15.8281,78.0373",
          "district": "Kurnool",
          "state": "Andhra Pradesh"
        }]
      }
    }
  }
}

UHI — Unified Health Interface enables discovery of CardioSleeve-equipped facilities across the AP HMIS network
Response — Returns list of facilities offering CS screening with availability and teleconsult booking slots

Integration Architecture Summary

Component	Standard	CardioSleeve Role	Data Exchanged
ABHA (Health ID)	NHA / ABDM	Patient identification at point of care	14-digit ABHA ID, demographic linking
HIE-CM	ABDM v0.5	Consent gateway for STI data sharing	STI Observations, DiagnosticReports
FHIR R4 API	HL7 FHIR R4	Standardized STI data format and exchange	Observation, DiagnosticReport, Patient, Bundle
SNOMED CT	WHO / NLM	Clinical terminology for STI findings	Condition codes, interpretation codes
LOINC	Regenstrief Institute	Observation codes for PEP, MPI	59010-6 (PEP), 59408-2 (MPI / Tei Index)
UHI	NHA UHI v0	CS service discovery & booking	Provider search, appointment scheduling
NHA Sandbox	sandbox.abdm.gov.in	Development & testing environment	All above with mock ABHA IDs for testing

Implementation Note: All API calls require OAuth 2.0 Bearer token authentication via the ABDM Gateway. HIE-CM patient consent must be obtained before fetching historical records. The CardioSleeve device must be registered in the NHA Medical Device Registry prior to production deployment in any ABDM-connected HMIS.

(412) 404-6292 | Info@Rijuven.com

Rijuven is a health innovation company and believes that access to care should be convenient and ubiquitous.

APDash

Dataset Composition

AUC by Parameter (Area Under ROC Curve)

Sensitivity vs Specificity at Standard Thresholds

AUC by Parameter (Horizontal)

Youden Index by Parameter

Sensitivity vs Specificity Trade-off

Youden Index — Top 8 Combinations

PPV vs NPV — Top Combinations

HF Risk % by Number of Abnormal Flags

Mean Echo EF% by Zone

Patient Distribution Across Zones

SPI OR MPI Screen

Count All 4 STI Flags

Zone-Based Clinical Decision

PPV & NPV Across Prevalence Settings

Benefit-Risk: True/False Positives per 1,000 Screened