Automating Compression Ultrasonography of Human Thigh Tissue and Vessels via Strain Estimation.

Rytis Jurkonis from Kaunas University of Technology (Lithuania) presented their recent work entitled “Automating Compression Ultrasonography of Human Thigh Tissue and Vessels via Strain Estimation." Rytis presented on the methodology along the novel wearable hardware developed to automate compression ultrasonography for DVT detection in the lower limbs. In addition, preliminary results were shared, highlighting the feasibility of an operator-independent method to perform compression ultrasonography.

Presented at BIOSTEC 2025 in Porto, Portugal.

About ThrombUS+: Our interdisciplinary approach centers around creating a novel wearable diagnostic device utilizing autonomous, AI-driven DVT detection. This groundbreaking device incorporates wearable ultrasound hardware, impedance plethysmography, and light reflection rheography for early clot detection. ThrombUS+ is designed for postoperative patients, those undergoing lengthy surgical procedures, cancer patients, bedridden individuals at home or in care units, and women during pregnancy and postpartum.

1. Automating Compression Ultrasonography of
Human Thigh Tissue and Vessels via Strain
Rytis Jurkonis
1
,Rimvydas Eitminavičius
1
,Vaidotas Marozas
1
,
Andrius Sakalauskas
2
1
Biomedical Engineering Institute, Kaunas University of Technology,
K. Baršausko 59, Kaunas, Lithuania
2
TELEMED, Ultrasound Medical Systems, Savanorių 178A, Vilnius,
Lithuania

2. Introduction
< ... very little pressure is needed to totally obliterate the
lumen of the normal vein...
Sonographic Criteria for Diagnosis of DVT
The following sonographic findings were used for the
diagnosis of DVT: (a) demonstration of an echogenic soft-
tissue mass within the lumen of the vein, and/or (b)
inability to totally obliterate the lumen of the vein by probe
compression... >
[ Raghavendraet al., 1986 ]
Isneed for:
•quantitative criteria about the limits of compression
•objective criteria, especially from sonography images.
2
https://pressbooks.palni.org/ultrasoundphysicsan
ditsapplicationinmedicine/chapter/vascular-
sonography/

3. Aim of study
The study explores the feasibility of automated
compression ultrasonography of thigh tissues by
proposing a hardware actuator and compression
control solution.
3

4. Materials and Methods (1/4)
Prototype of mountingonthigh:
•two rigid C-shaped frames;
•locking slot for fixing the imaging transducer;
•fixed with two nonelastic adjustable straps;
•two bladders on back of thigh;
4

5. Materials and Methods (2/4)
Two bladders:
•rectangular, 60 x 100 mm size;
Silent piezoelectric air pump:
•model UXPB5400000A (Lee Company,
Westbrook, CT, USA);
•free flow output 1.35 L/min;
•pressure up to 140 mmHg.
5

6. Materials and Methods (3/4)
Ultrasonic imaging:
•„ArtUs1H“ beamformer (Telemed, Lithuania);
•Low-profilelinear array (LF11-5H60-A3).
Imaging options:
•depth5 cm;
•focaldistance5 cm;
•frequency 5 to7 MHz;
•Dynamicrange 36 dB.
Pneumatic actuator(airpump).
6

7. Materials and Methods (4/4)
7
Input real-time streams:
•B-modepictures512 x 512 pixels;
•RF signal sampled 40MHz 16 bits.
MATLAB GUIfunctions:
•imagingsettings;
•initialize imagingandcompressionsimultineously;
•fromRF signals calculating displacements and strain
oftissues [Zahiri and Salcudean, 2006];
•images andstraintracerepresentation;
•new sampleinstraintrace> Threshold: NoorYes?
•data saving into file;
•finalize compressionsonographysession.

8. Data collection (1/5)
Body positioning:
•testbed with back-support elevated at a 30;
•leg freely extended.
Tissue compressions variants:
1.manually, by manoeuvring the imaging
transducer;
2.by using an automated tissue compression
actuatorintegratedwithimagingtransducer.
8
30
o

9. Data collection: by-hand (2/5)
9
Initial(strainlevel3%) Final(strainlevel11%)
By-hand compression:
•front side of thigh – routine clinical practice;
•reaction of tissues to compressionexpressed with strain trace in time.
•compression dossing with 2, 5, 8, and 11% strain thresholds;

10. Data collection: automated (3/5)
Automated compression:
1.with piezo-pump inflated
bladders at back of thigh;
2.compression dossing with 1.5,
2.5, 3.5, 4.5 and 5.5% strain
thresholds.
10
Compressions recorded three times each.
Image sequences saved for off-line analysis.

11. Data collection: automatedmonitoring (4/5)
Time instances, when reactions of lumen to dosed compressions were
recorded. Doses of compressions were interchanged with strain threshold 4.5%
- 2.5% - 4.5% and so on in more than 60 compressions(150 minutes).
11

12. Data analysis (1/5)
Figure 9: Cross-sectional images of femoral vessels during compression session of the thigh tissues. The artery is the
most left circular pattern in all images and does not collapse at the end of the compression session.
12

13. Data analysis (2/5)
Examplesof primarydataof imagingduringmotorizedPrototypeNo1
compressions: toprow–indexedimagesof structuresin the sequence; bottom
row–changesof veinduringcompressionand releaseof tissues. The thighsize
was56 cmin circumference.
13

14. Data analysis (3/5)
14
Image analysis:
•positionsofsurfacesfromthresholdingin
imageintensityprofiles;
•topandbottomwalls deformations over time;
•Lumen(verticaldimension) decreasefrom
differenceofwallsdeformations.

15. Data analysis (4/5)
Correlated:
•Diameter decrease;
•Strain trace;
Real-timecalculatedstraintracecanquantifydosewhilecompressing.
15

16. Data collection (5/5)
Thresholdofstrainfor:
•by-hand compression2, 5, 8, and 11%
•automated compression:1.5, 2.5, 3.5,
4.5 and 5.5%.
16
Compressions recorded three times per thresholdforeachperson.
Image sequences andstraintrace: saved for off-line verification.

17. Data analysis (5/5)
Verticalsize(diameter) decrease[%] from off-lineanalysis.
17

18. Results (1/6)
Strain dosed compressionsonthigh:
•by-hand from front side
•automated with actuator from back side
18

19. Results (2/6)
Figure 11: Cross-sectional images of blood vessels during manual compression (up to strain 11%) of thigh
tissues: intermediate reactions of one of the subjects characterized in Figures 6.
19

20. Results (3/6)
Strain dosed compressionsonthigh:
•by-hand from front side
•automated with actuator from back side
20

21. Results (4/6)
Eachpointofresultsisfromsingle
compression;
Singleattemptofbodyandleg
positioning;
Singlemountingofactuatorintegrated
withimagingtransducer.
21

22. Results (5/6)
In-vivo monitoring with
compression up to 2.5% of
strain.
Lumendecrease median
12% with median pressure
of 53 mmHg.
22

23. Results (6/6)
In-vivo monitoring with
compression up to 4.5%
of strain.
Lumendecrease
median 45% with
median pressure of 79
mmHg.
23

24. Conclusions
In this study, an ultrasonography tissue compression actuator for automated long-term
monitoring of venous vessels in the lower extremities is proposed.
The actuator is controlled by a tissue strain parameter that regulates compression.
The results show a negative correlation between venous lumen closure and the tissue
strain parameter, suggesting that using tissue strain as a feedback mechanism is
feasible.
However, the variability at a higher strain threshold is relatively big, leaving room for
improvement.
Inter-subject variability could be addressed through initialcalibration, while intra-
subject variability for monitoring applications could be managed by defining a more
targetedanalysisregion insonographyimage.
24

25. Acknowledgements
This work is funded under the Horizon Europe Innovation Action ThrombUS+
(Grant Agreement No.101137227), co-funded by the European Union.
"Wearable Continuous Point-of-Care Monitoring, Risk Estimation and Prevention
for Deep Vein Thrombosis"
25

26. Openfordiscussion
[email protected]
https://biomedicine.ktu.edu
Biomedical Engineering Institute, Kaunas University of Technology, Lithuania
26
Cite as: Jurkonis R, Eitminavičius R, Marozas V and Sakalauskas
A. (2025). Automating Compression Ultrasonography of
Human Thigh Tissue and Vessels via Strain Estimation. In
Proceedings of the 18th International Joint Conference on
Biomedical Engineering Systems and Technologies - Volume 1,
ISBN 978-989-758-731-3, ISSN 2184-4305, pages 239-245.
https:/doi.org/10.5220/0013264300003911