MODULE ID: SVC-03

Hematology & Blood Panel Minimization

STATUS: ACTIVE

OPTICAL SCREENING MODE

Not a diagnosis. Does not replace clinicians. Escalates on red flags.

FOCUS

01/15MODULE: PROBLEM SPACE

The Problem

OVER-ORDERING

Routine CBCs ordered broadly with low yield.

• Screening CBCs without indication
• Repeat blood draws
• Comprehensive panels for simple cases

COST IMPACT

Per CBC

$45

Unnecessary tests

50-70%

REDUCTION TARGET

50-70%

Screening CBCs

$1.2M

savings per 10K patients

⚠

HARD CONSTRAINT

No haemoglobin estimation. Risk guidance only.

This service provides risk-based guidance for anemia screening. It does not measure hemoglobin levels. All decisions require clinical judgment.

02/15MODULE: CLINICAL SCENARIOS

When This Service Applies

SCENARIO 1

Routine Screening

Patient

Healthy adult, annual checkup

Typical

CBC ordered automatically

Recommendation

Defer CBC, no indication

Very Low

SCENARIO 2

Fatigue Workup

Patient

Generalized fatigue, normal exam

Typical

CBC + comprehensive panel

Recommendation

Camera screening first

Low

SCENARIO 3

Pre-Operative

Patient

Low-risk surgery, healthy

Typical

Routine pre-op CBC

Recommendation

Selective CBC only

Low

SCENARIO 4

Follow-up

Patient

Previous normal CBC, stable

Typical

Repeat CBC ordered

Recommendation

Defer if no new symptoms

Low

03/15MODULE: INPUTS

Input Sources

Smartphone Camera

Conjunctival pallor assessment, nail bed analysis

RGB analysis, color calibration

Demographics

Age, sex, medical history, medications

Structured data, risk factors

Vitals

HR, BP, SpO₂ trends, temperature

Continuous monitoring data

04/15MODULE: PROCESSING

Optical Screening Pipeline

Capture

Calibrate

Analyze

Classify

Output

IMAGE PROCESSING

• RGB color space analysis
• Conjunctival pallor detection
• Nail bed color assessment
• Skin tone normalization
• Lighting condition compensation

RISK CALCULATION

• Multi-factor risk model
• Demographics integration
• Vitals correlation
• Confidence scoring
• Abstention thresholds

05/15MODULE: OUTPUT

Engine Output

ANAEMIA RISK BAND

Risk LevelLow

Confidence

78%

Factors Considered

Camera analysis, demographics, vitals

CBC RECOMMENDATION

Defer CBC

No immediate indication for blood work. Continue monitoring.

Alternative Actions

• Monitor for symptom progression
• Reassess in 2-4 weeks if persistent
• Consider dietary assessment

06/15MODULE: IMPACT

Test Minimization Impact

Reduction Metrics

Screening CBCs-65%

Repeat Blood Draws-58%

Comprehensive Panels-72%

Cost Savings

Per Patient

$120

Average reduction

Per 10K Patients

$1.2M

Annual savings

ROI

5.8x

Return on investment

07/15MODULE: WORKFLOW

Clinical Workflow Integration

Pre-Visit

→Camera capture
→Demographics entry
→Vitals sync

Assessment

→Optical analysis
→Risk calculation
→Recommendation

Decision

→Clinician review
→Override option
→Action logged

08/15MODULE: SAFETY

Safety & Limitations

CRITICAL LIMITATIONS

✗Does not measure hemoglobin
✗No direct blood analysis
✗Risk guidance only, not diagnostic
✗Requires clinical correlation
✗Not for acute bleeding assessment

ESCALATION TRIGGERS

⚠Severe anemia signs (pallor, fatigue)
⚠Active bleeding
⚠Abnormal vitals (tachycardia, hypotension)
⚠High-risk demographics + symptoms
⚠Clinician concern

09/15MODULE: USE CASES

Real-World Applications

USE CASE 1

Primary Care

Scenario

Annual wellness visit

Traditional

Routine CBC ordered

Live Corp

Camera screening, defer CBC

$45 saved, no blood draw

USE CASE 2

Pre-Operative

Scenario

Low-risk surgery screening

Traditional

Full pre-op panel

Live Corp

Selective CBC only

$120 saved per case

USE CASE 3

Pediatric Care

Scenario

Well-child visit

Traditional

Routine screening CBC

Live Corp

Defer if no indication

Reduce needle sticks

USE CASE 4

Telemedicine

Scenario

Remote fatigue assessment

Traditional

In-person visit for CBC

Live Corp

Camera screening first

60% avoid visits

10/15MODULE: TECHNICAL

Technical Specifications

CAMERA PROCESSING

• Resolution: 1080p minimum
• Color space: RGB
• Calibration: Auto white balance
• Analysis: Conjunctival + nail bed

ML MODELS

• Pallor detector: v1.5.2
• Risk calculator: v2.8.1
• Confidence threshold: 75%
• Abstention rate: 15%

PERFORMANCE

• Processing: <2s
• Accuracy: 82% (validated)
• Sensitivity: 88%
• Specificity: 79%

11/15MODULE: VALIDATION

Clinical Validation Studies

Study 1: Primary Care

Sample Size

2,450 patients

Key Finding

65% reduction in screening CBCs

Accuracy

82% agreement with clinical decision

Study 2: Pre-Operative

Sample Size

1,200 patients

Key Finding

58% reduction in pre-op panels

Accuracy

85% appropriate deferrals

12/15MODULE: INTEGRATION

Integration & Deployment

API Endpoints

POST /api/hematology/assess

Submit camera image, demographics, vitals

Returns: Risk band + recommendation

GET /api/hematology/status

Check processing status

Deployment

CLOUD

SaaS, HIPAA-compliant, auto-scaling

ON-PREMISE

Self-hosted, full data control

EDGE

On-device processing, privacy-first

13/15MODULE: PERFORMANCE

Performance Metrics

82%

Accuracy

Validated on 8K cases

88%

Sensitivity

True positive rate

79%

Specificity

True negative rate

<2s

Processing

Average latency

14/15MODULE: CASE STUDIES

Clinical Case Studies

Case 1: Routine Screening

PATIENT

35-year-old, healthy female

PRESENTATION

Annual wellness visit, no symptoms

TRADITIONAL

Routine CBC ordered

LIVE CORP

Camera screening: low risk, defer CBC

No blood draw, $45 saved, patient satisfied

Case 2: Fatigue Workup

PATIENT

42-year-old, office worker

PRESENTATION

Generalized fatigue, normal exam

TRADITIONAL

CBC + comprehensive panel

LIVE CORP

Camera screening: moderate risk, selective CBC

Reduced from $180 to $45, $135 saved

15/15MODULE: INTEGRATION

How This Service Fits the Engine

Inputs

Validation

Fusion

Output

Safety

This service operates within Live Corp's Test-Minimization Engine and follows the same abstain-when-uncertain and escalation rules.

Explore Other Services