ThrombUS+ - Wearable DVT Monitoring Device

Background and Motivation

What the Device Does

• Continuously monitors for DVT formation in real time
• Detects clots early before symptoms appear
• Assesses DVT risk on an ongoing basis
• Guides prevention exercises through serious gaming and extended reality (XR) interfaces
• Alerts clinicians via an AI-powered decision support unit

Technology Components

Target Patient Populations

• Postoperative patients
• Patients undergoing lengthy surgical procedures
• Cancer patients
• Bedridden patients (at home or in care units)
• Pregnant women and those in the postpartum period
• Passengers on long-haul flights (DVT risk management)

Project Details

Clinical Validation Plan

1. Study A (>60 participants) - Early feasibility: collecting baseline sensor signals from healthy volunteers
2. Study C1 (25-50 participants) - Early feasibility of the integrated prototype in postoperative ward patients
3. Study C2 (50-100 participants) - Prospective, double-blinded, pilot study comparing ThrombUS+ AI-driven DVT detection against standard ultrasound in patients suspected of having DVT

Relevance to Medical Professionals

How ThrombUS+ Works and Detects DVT

The Physical Setup

• Ultrasound transducer arrays
• Bioimpedance electrodes
• Light/optical sensors
• Inertial measurement units (IMUs/accelerometers)

Detection Modality 1: Wearable Ultrasound

• Ultrasound waves are emitted into the leg tissue from transducers in the cuff
• Sound waves reflect differently off blood, vessel walls, and clot material
• The system reconstructs vascular images showing whether a vein is compressible (normal) or obstructed by thrombus (clot present)
• This is essentially a miniaturized, continuous version of the duplex ultrasound used in hospitals - but worn on the patient 24/7

• VERMON bulk transducer arrays - traditional piezoelectric technology, miniaturized
• Fraunhofer IPMS CMUT arrays (Capacitive Micromachined Ultrasonic Transducers) - next-generation MEMS-based sensors that are smaller, cheaper, and offer higher sensitivity and bandwidth; mass-producible at low cost

Detection Modality 2: Impedance Plethysmography (IPG)

• A small, safe electrical current is passed through the leg tissue
• Blood is a good electrical conductor; clotted or stagnant blood and edema change the tissue's electrical impedance (resistance)
• When a DVT forms and blocks venous outflow, blood pools in the limb - this causes a measurable, characteristic change in impedance
• The system continuously tracks these changes and flags abnormal patterns

Detection Modality 3: Light Reflection Rheography (LRR)

• Infrared light is shone into the skin of the lower leg
• The amount of light reflected back changes depending on the volume of blood in the superficial and deep veins beneath the skin
• In a healthy leg, when the calf muscle pump works (during movement), it empties the veins - the LRR signal changes in a predictable pattern
• When a DVT is present, venous outflow is obstructed: blood does not drain normally, and the refilling time after muscle activity is abnormally fast - the LRR signal is abnormal
• This is a time-efficient, well-suited screening method for lower leg (calf) veins - complementing IPG's weakness

Detection Modality 4: Activity Sensors (IMUs)

• Accelerometers and inertial measurement units track whether the patient is moving their leg, walking, or completely still
• Immobility is a major DVT risk factor (Virchow's triad: stasis, endothelial injury, hypercoagulability)
• Prolonged inactivity triggers a higher risk score in the AI model
• The system also uses this data to drive serious gaming - the patient receives interactive exercises via extended reality, encouraging movement to prevent clot formation

The AI Layer: Putting it All Together

1. Fuses the signals - correlates ultrasound images, impedance changes, LRR waveforms, and activity levels together
2. Detects DVT in real time - trained on large clinical datasets from three clinical studies
3. Scores DVT risk continuously - even before a clot is visible, rising risk trends trigger early alerts
4. Alerts clinicians - sends notifications when DVT is detected or risk crosses a threshold
5. Guides the patient - extended reality (XR) interface instructs the patient on leg exercises

Summary Flow

Patient wears cuff on lower limb
         ↓
[Ultrasound] → Vein imaging → Clot visible?
[Impedance IPG] → Blood flow obstruction? → Proximal DVT?
[LRR optical] → Venous refill pattern? → Calf DVT?
[IMU activity] → Is the patient immobile? → Risk score ↑
         ↓
   AI Engine fuses all signals
         ↓
Real-time DVT detection + Risk alert → Clinician notified
         ↓
XR + Gaming → Patient exercises → Prevention

1. Research/prototype components (not sold as a consumer product with a fixed price)
2. Custom-built assemblies for projects like ThrombUS+

Cost Estimate: Miniature Ultrasound Transducer (UST) + Near Infrared (NIR) Sensor Combo

Important caveat

1. Miniature Ultrasound Transducer (UST)

2. Near Infrared (NIR) Sensor / Light Reflection Rheography (LRR) Module

3. If You Are Building a Combined System (DIY / Research)

4. Why There Is No Single Listed Price

• Used in research projects like ThrombUS+ (€9.5M EU-funded, not a product for sale yet)
• Available as separate research components from suppliers like Fraunhofer IPMS, VERMON, medis GmbH
• Custom-quoted by manufacturers depending on volume, integration level, and application

Bottom Line

• Fraunhofer IPMS (ipms.fraunhofer.de) for CMUT modules - contact for prototype pricing
• medis GmbH (medis.company) for LRR/NIR modules used in ThrombUS+
• VERMON for ultrasound transducer arrays

Single Prototype Cost Estimate

Component 1 - Miniature Ultrasound Transducer (UST)

Component 2 - Near Infrared (NIR) Sensor

Component 3 - Supporting Electronics

Total Prototype Cost Summary

Key Points for Your Product Design Cost

• Single prototype cost is always higher than per-unit cost at volume - components are priced for individual buyers, not bulk
• Once you move to a batch of 100+ units, component costs typically drop 40-60%
• You will also need to budget for development time, PCB design, firmware, and regulatory testing (CE/FDA) if selling commercially - this is typically $50,000-$500,000+ depending on your target market
• The biggest cost driver is the UST - choosing between a CMUT chip vs. a simple piezo element changes your BOM cost by 10x