scholarly journals Non-invasive diagnosis of deep vein thrombosis from ultrasound imaging with machine learning

2021 ◽  
Vol 4 (1) ◽  
Author(s):  
Bernhard Kainz ◽  
Mattias P. Heinrich ◽  
Antonios Makropoulos ◽  
Jonas Oppenheimer ◽  
Ramin Mandegaran ◽  
...  
Keyword(s):  
Deep Vein Thrombosis ◽  
Deep Learning ◽  
Predictive Value ◽  
Gold Standard ◽  
Learning Algorithm ◽  
Blood Clot ◽  
Decision Algorithm ◽  
Vein Thrombosis ◽  
Compression Ultrasound ◽  
Deep Vein

AbstractDeep vein thrombosis (DVT) is a blood clot most commonly found in the leg, which can lead to fatal pulmonary embolism (PE). Compression ultrasound of the legs is the diagnostic gold standard, leading to a definitive diagnosis. However, many patients with possible symptoms are not found to have a DVT, resulting in long referral waiting times for patients and a large clinical burden for specialists. Thus, diagnosis at the point of care by non-specialists is desired. We collect images in a pre-clinical study and investigate a deep learning approach for the automatic interpretation of compression ultrasound images. Our method provides guidance for free-hand ultrasound and aids non-specialists in detecting DVT. We train a deep learning algorithm on ultrasound videos from 255 volunteers and evaluate on a sample size of 53 prospectively enrolled patients from an NHS DVT diagnostic clinic and 30 prospectively enrolled patients from a German DVT clinic. Algorithmic DVT diagnosis performance results in a sensitivity within a 95% CI range of (0.82, 0.94), specificity of (0.70, 0.82), a positive predictive value of (0.65, 0.89), and a negative predictive value of (0.99, 1.00) when compared to the clinical gold standard. To assess the potential benefits of this technology in healthcare we evaluate the entire clinical DVT decision algorithm and provide cost analysis when integrating our approach into diagnostic pathways for DVT. Our approach is estimated to generate a positive net monetary benefit at costs up to £72 to £175 per software-supported examination, assuming a willingness to pay of £20,000/QALY.

scholarly journals Non-invasive Diagnosis of Deep Vein Thrombosis from Ultrasound with Machine Learning

2021 ◽  
Author(s):  
Bernhard Kainz ◽  
Antonios Makropoulos ◽  
Jonas Oppenheimer ◽  
Christopher Deane ◽  
Sven Mischkewitz ◽  
...  
Keyword(s):  
Deep Vein Thrombosis ◽  
Deep Learning ◽  
Predictive Value ◽  
Gold Standard ◽  
Learning Algorithm ◽  
Blood Clot ◽  
Decision Algorithm ◽  
Vein Thrombosis ◽  
Compression Ultrasound ◽  
Deep Vein

AbstractDeep Vein Thrombosis (DVT) is a blood clot most found in the leg, which can lead to fatal pulmonary embolism (PE). Compression ultrasound of the legs is the diagnostic gold standard, leading to a definitive diagnosis. However, many patients with possible symptoms are not found to have a DVT, resulting in long referral waiting times for patients and a large clinical burden for specialists. Thus, diagnosis at the point of care by non-specialists is desired.We collect images in a pre-clinical study and investigate a deep learning approach for the automatic interpretation of compression ultrasound images. Our method provides guidance for free-hand ultrasound and aids non-specialists in detecting DVT.We train a deep learning algorithm on ultrasound videos from 246 healthy volunteers and evaluate on a sample size of 51 prospectively enrolled patients from an NHS DVT diagnostic clinic. 32 DVT-positive patients and 19 DVT-negative patients were included. Algorithmic DVT diagnosis results in a sensitivity of 93.8% and a specificity of 84.2%, a positive predictive value of 90.9%, and a negative predictive value of 88.9% compared to the clinical gold standard.To assess the potential benefits of this technology in healthcare we evaluate the entire clinical DVT decision algorithm and provide cost analysis when integrating our approach into a diagnostic pathway for DVT. Our approach is estimated to be cost effective at up to $150 per software examination, assuming a willingness to pay $26 000/QALY.

Evaluation of a New Rapid Quantitative D-dimer Assay in Patients with Clinically Suspected Deep Vein Thrombosis

1996 ◽  
Vol 75 (03) ◽  
pp. 412-416 ◽  
Author(s):  
Armando D’Angelo ◽  
Gabriella D’Alessandro ◽  
Loredana Tomassini ◽  
Jean Louis Pittet ◽  
G Dupuy ◽  
...  
Keyword(s):  
Deep Vein Thrombosis ◽  
Predictive Value ◽  
Cost Savings ◽  
Pretest Probability ◽  
First Episode ◽  
Consecutive Series ◽  
Vein Thrombosis ◽  
Baseline Testing ◽  
Deep Vein ◽  
D Dimer

SummaryThe sensitivity and specificity for deep vein thrombosis (DVT) of a new rapid, quantitative and precise (total imprecision < 10%) D-dimer assay suitable for individual measurements (VIDAS D-DIMER, bio-Merieux, France) were evaluated in a consecutive series of 103 in- and out-patients submitted to serial compression ultrasonography (C-US) for the clinical suspicion of DVT (n = 66) or of DVT recurrence (n = 37) and symptoms lasting from 1 to 15 days. DVT was found in 22 patients at baseline testing and no patient with an initially negative C-US developed vein incompressibility at follow up. The time elapsed from the onset of symptoms was negatively associated with D-dimer levels both in patients with and in those without DVT. In the entire series of patients, the sensitivity of a positive D-dimer test (≥1.0 Μg/ml) for the presence of DVT was 96% (21/22 patients, 95% confidence interval 75-100%) with a specificity of 75% (64-84%), a negative predictive value of 98% (90-100%), a positive predictive value of 51% (35-67%), and an overall accuracy of 80% (70-87%). A normal D-dimer value (0.22 Μg/ml) was observed in one patient with DVT and symptoms lasting from 15 days. The approach of withholding C-US testing in patients with symptoms lasting from less than 11 days and D-dimer levels below the cut-off value was compared to serial C-US testing alone in a cost-effectiveness analysis subdividing the 66 patients with a first episode according to their clinical pretest probability of DVT. Thrombosis was detected in 6.7% of the patients in the low probability group (n = 15), 16.7% of the patients in the moderate probability group (n = 24), 51.9% of the patients in the high probability group (n = 27) and 8.1% of patients with suspected DVT recurrence. Calculated cost-savings for each DVT diagnosed ranged from 5% in the high pretest probability group to 55% in the low pretest probability group and to 77% in patients with suspected DVT recurrence.The safety of avoiding C-US testing in symptomatic patients with a negative D-dimer test should be evaluated in clinical management studies.

Is Colour Doppler Ultrasound a Sensitive Screening Method in Diagnosing Deep Vein Thrombosis after Hip Surgery ?

1996 ◽  
Vol 75 (02) ◽  
pp. 242-245 ◽  
Author(s):  
Marie Magnusson ◽  
Bengt I Eriksson ◽  
Peter Kãlebo ◽  
Ramon Sivertsson
Keyword(s):  
Deep Vein Thrombosis ◽  
High Risk ◽  
Doppler Ultrasound ◽  
Predictive Value ◽  
Screening Method ◽  
Colour Doppler ◽  
Vein Thrombosis ◽  
Colour Doppler Ultrasound ◽  
Asymptomatic Patients ◽  
Deep Vein

SummaryPatients undergoing orthopedic surgery are at high risk of developing deep vein thrombosis. One hundred and thirty-eight consecutive patients undergoing total hip replacement or hip fracture surgery were included in this study. They were surveilled with colour Doppler ultrasound (CDU) and bilateral ascending contrast phlebography. The prevalence of proximal and distal DVT in this study was 5.8% and 20.3% respectively.CDU has a satisfactory sensitivity in patients with symptomatic deep vein thrombosis, especially in the proximal region. These results could not be confirmed in the present study of asymptomatic patients. The sensitivity was 62.5% (95% confidence interval: C.I. 24-91%) and the specificity 99.6% (C.I. 98-100%) for proximal DVT; 53.6% (C.I. 34-73%) and 98% (C.I. 96-99%) respectively for distal thrombi. The overall sensitivity was 58.1% (C.I. 39-75%) and the specificity 98% (C.I. 96-99%). The positive predictive value was 83.3% (C.I. 36-99%) and 75% (C.I. 51-91%) for proximal and distal DVT respectively. The negative predictive value was 98.9% (C.I. 98-100%) and 94.9% (C.I. 92-98%) for proximal and distal DVT respectively. The results of this study showed that even with a highly specialised and experienced investigator the sensitivity of CDU was too low to make it suitable for screening purposes in a high risk surgical population.

Predictive Value of Decreasing Antithrombin III in Deep-Vein Thrombosis

1980 ◽  
Vol 44 (03) ◽  
pp. 135-137 ◽  
Author(s):  
Thorkild Lund Andreasen
Keyword(s):  
Deep Vein Thrombosis ◽  
Predictive Value ◽  
Coronary Care Unit ◽  
Antithrombin Iii ◽  
Predictive Values ◽  
Vein Thrombosis ◽  
At Iii ◽  
Coronary Care ◽  
Time Of Admission ◽  
Deep Vein

SummaryAntithrombin III (At-III) was measured at the time of admission and two days later in 131 patients laid up in a coronary care unit. The patients were examined for deep-vein thrombosis (DVT) clinically and by means of 125I-fibrinogen scanning. 19 patients developed DVT. In 11 subjects with and 25 without DVT At-III decreased more than 10%. And in 7 with and 17 without DVT At-III decreased more than 15%. One person with DVT had subnormal At-III. By using decrease of At-III or subnormal initial At-III to predict DVT the following predictive value (PV) were found. Decrease ≤ 10%, PV pos.= 0.32 and PV neg. = 0.93. Decrease ≤ 15%, PV pos. = 0.32 and PV neg. = 0.90. The positive predictive values obtained were too low to let decreasing At-III give occasion for prophylactic anticoagulant treatment.

D-Dimer Test and Diagnosis of Deep Vein Thrombosis: A Comparative Study of 7 Assays

1996 ◽  
Vol 76 (04) ◽  
pp. 518-522 ◽  
Author(s):  
A Elias ◽  
I Aptel ◽  
B Huc ◽  
J J Chale ◽  
F Nguyen ◽  
...  
Keyword(s):  
Deep Vein Thrombosis ◽  
Negative Predictive Value ◽  
Predictive Value ◽  
Latex Agglutination ◽  
Lower Limbs ◽  
First Episode ◽  
Lower Sensitivity ◽  
Vein Thrombosis ◽  
Deep Vein ◽  
D Dimer

SummaryThe current D-Dimer ELISA methods provide high sensitivity and negative predictive value for the diagnosis of deep vein thrombosis but these methods are not suitable for emergency or for individual determination. We have evaluated the performance of 3 newly available fast D-Dimer assays (Vidas D-Di, BioMerieux; Instant IA D-Di, Stago; Nycocard D-Dimer, Nycomed) in comparison with 3 classic ELISA methods (Stago, Organon, Behring) and a Latex agglutination technique (Stago). One-hundred-and-seventy-one patients suspected of presenting a first episode of deep vein thrombosis were investigated. A deep vein thrombosis was detected in 75 patients (43.8%) by ultrasonic duplex scanning of the lower limbs; in 11 of them the thrombi were distal and very limited in size (<2 cm). We compared the performance of the tests by calculating their sensitivity, specificity, positive and negative predictive value for different cut-off levels and by calculating the area under ROC curves. The concordance of the different methods was evaluated by calculating the kappa coefficient. The performances of the 3 classic ELISA and of the Vidas D-Di were comparable and kappa coefficients indicated a good concordance between the results provided by these assays. Their sensitivity slightly declined for detection of the very small thrombi. Instant IA D-Di had a non-significantly lower sensitivity and negative predictive value than the 4 previous assays; however its performance was excellent for out-patients. As expected, the Latex assay had too low a sensitivity and negative predictive value to be recommended. In our hands, Nycocard D-Dimer also exhibited low sensitivity and negative predictive value, which were significantly improved when the plasma samples were tested by the manufacturer. Thus significant progress has been made, allowing clinical studies to be planned to compare the safety and cost-effectiveness of D-Dimer strategy to those of the conventional methods for the diagnosis of venous thrombosis.

Diagnosis of Deep Vein Thrombosis by Combination of Doppler Ultrasound Flow Examination and Strain Gauge Plethysmography

1982 ◽  
Vol 47 (02) ◽  
pp. 141-144 ◽  
Author(s):  
H Bounameaux ◽  
B Krähenbühl ◽  
S Vukanovic
Keyword(s):  
Deep Vein Thrombosis ◽  
Negative Predictive Value ◽  
Doppler Ultrasound ◽  
Strain Gauge ◽  
Predictive Value ◽  
Lower Limbs ◽  
Vein Thrombosis ◽  
Non Invasive ◽  
Strain Gauge Plethysmography ◽  
Deep Vein

SummaryDoppler ultrasound flow examination, strain gauge plethysmography and contrast venography were performed in 160 lower limbs of 80 in-patients. Deep vein thrombosis (DVT) was suspected in 87 limbs. Using measurement of venous stop-flow pressure, the Doppler method had an overall sensitivity of 83%. By combined use of Doppler and Plethysmography, sensitivity was increased to 96%. Specificity was 62% and 51%, respectively. With a positive and a negative predictive value of 80% and 73%, respectively, the combination of both non-invasive methods cannot reliably replace venography in the diagnosis of DTV, although all (40/40) thromboses proximal to or involving the popliteal segment were detected by either Doppler and Plethysmography or both.After exclusion of 14 patients (18%) suffering from conditions known to alter the results of these non-invasive methods, the positive predictive value of abnormal findings in both Doppler and Plethysmography was increased to 94% for suspected limbs, whilst negative predictive value of both negative Doppler and Plethysmography was 90%, allowing the avoidance of venography in these patients.

scholarly journals The Predictive Value of the aCL and Anti-β2GPI at the Time of Acute Deep Vein Thrombosis—A Two-Year Prospective Study

Biomedicines ◽  
2021 ◽  
Vol 9 (8) ◽  
pp. 901
Author(s):  
Katja Perdan-Pirkmajer ◽  
Polona Žigon ◽  
Anja Boc ◽  
Eva Podovšovnik ◽  
Saša Čučnik ◽  
...  
Keyword(s):  
Deep Vein Thrombosis ◽  
Prospective Study ◽  
Anticoagulant Therapy ◽  
Predictive Value ◽  
Anticoagulation Therapy ◽  
Time Frame ◽  
Vein Thrombosis ◽  
Therapeutic Decisions ◽  
Acute Deep Vein Thrombosis ◽  
Deep Vein

Antiphospholipid syndrome (APS) is an important cause of deep vein thrombosis (DVT). According to current APS classification criteria, APS cannot be confirmed until 24 weeks after DVT. This time frame results in frequent discontinuation of anticoagulant treatment before APS is diagnosed. Therefore, the aim of our study was to evaluate the potential predictive value of anticardiolipin (aCL) and anti-β2glycoprotein I (anti-β2GPI) before discontinuation of anticoagulation therapy. Patients with newly diagnosed DVT were included into a 24-month prospective study. All patients received anticoagulant therapy. aCL and anti-β2GPI were determined at inclusion and every four weeks for the first 24 weeks and then one and two years after inclusion. APS was confirmed in 24/221 (10.9%) patients. At the time of acute DVT 20/24 (83.3%), APS patients had positive aCL and/or anti-β2GPI. Two patients had low aCL levels and two were negative at the time of acute DVT but later met APS criteria due to lupus anticoagulant (LA). Our data indicate that negative aCL and/or anti-β2GPI at the time of acute DVT make further aPL testing unnecessary; however, LA should be determined after discontinuation of anticoagulant therapy. Positive aCL and/or anti-β2GPI at the time of acute DVT have a strong positive predictive value for APS and may support therapeutic decisions.

Mortality rates and risk factors for asymptomatic deep vein thrombosis in medical patients

2005 ◽  
Vol 93 (01) ◽  
pp. 76-79 ◽  
Author(s):  
Alain Leizorovicz ◽  
Alexander Cohen ◽  
Alexander Turpie ◽  
Carl-Gustav Olsson ◽  
Samuel Goldhaber ◽  
...  
Keyword(s):  
Risk Factors ◽  
Deep Vein Thrombosis ◽  
Mortality Rates ◽  
Hazard Ratio ◽  
Group Iii ◽  
Vein Thrombosis ◽  
Group I ◽  
Compression Ultrasound ◽  
The Difference ◽  
Deep Vein

SummaryThe clinical importance of asymptomatic proximal and distal deep vein thrombosis (DVT) remains uncertain and controversial. The aim of this retrospective,post-hoc analysis was to examine mortality and risk factors for development of proximal DVT in hospitalized patients with acute medical illness who were recruited into a randomized, prospective clinical trial of thromboprophylaxis with dalteparin (PREVENT).We analyzed 1738 patients who had not sustained a symptomatic venous thromboembolic event by Day 21 and who had a complete compression ultrasound of the proximal and distal leg veins on Day 21. We examined the 90-day mortality rates in patients with asymptomatic proximal DVT (Group I, N = 80), asymptomatic distal DVT (Group II, N = 118) or no DVT (Group III, N = 1540).The 90-day mortality rates were 13.75%, 3.39%, and 1.92% for Groups I–III, respectively. The difference in mortality between Group I and Group III was significant (hazard ratio 7.63, 95% CI = 3.8–15.3;p < 0.0001),whereas the difference between Groups II and III did not reach significance (hazard ratio 1.36, 95% CI = 0.41–4.45).The association of asymptomatic proximal DVT with increased mortality remained highly significant after adjusting for differences in baseline demographics and clinical variables. Risk factors significantly associated with the development of proximal DVT included advanced age (p = 0.0005), prior DVT (p = 0.001), and varicose veins (p = 0.04). In conclusion, the high mortality rate in patients with asymptomatic proximal DVT underscores its clinical relevance and supports targeting of asymptomatic proximal DVT as an appropriate endpoint in clinical trials of thromboprophylaxis.

Sign in / Sign up

Export Citation Format

Share Document