Development of an IT tool to support post-market surveillance and expert panels in detecting sentinel signals relevant to serious incidents in high-risk medical devices: pilot on italian data

Introduction

The EU Medical Device Regulation 2017/745, in force since May 26th 2021, defines new rules for medical device (MD) certification and post-market surveillance (PMS). For high-risk implantable MD, in the certification process notified bodies are obliged to consult clinical Expert Panels (EP) that could decide to proceed for an extensive review of the supporting clinical data provided by the manufacturer, also in view of sentinel signals, such as significantly increased rate of reported serious incidents for a specific MD groups.

Purpose

1) To develop an ICT tool to automatically collect and display in an aggregated way the accessible curated regulatory information on MD alerts and recalls to capture possible trends in reported incidents that could be used both for scientific analysis and as information source to EP. 2) To conduct a pilot feasibility study on the Italian data, characterized already by the same European Medical Device Nomenclature (EMDN), organized in a multi-level hierarchical tree code to define a MD, as it will be used in Europe.

Methods

Web scraping was used to retrieve data of 7622 safety notices (SN) from 2009 to 2021 from the Italian Ministry of Health website. The EMDN code was missing in 68% of cases: to retrieve it, the MD best match was searched within a separate public list of about 1.5M MD on the Italian market containing the EMDN code, using Natural Language Processing techniques and pairwise entity resolution with Cosine similarity to identify similar manufacturers and MD. The performance of this approach was tested on the 2440 SN for which the EMDN code was available as gold standard. A mash up was then performed to integrate data, and to present it to the final user through a graphical interface.

Results

The implemented entity resolution method was able to correctly assign the correct manufacturer to the MD in each SN in 99% of the cases. Moreover, the correct EMDN code at level 1 (22 categories available) was assigned in 2382 SN (97.62%), at level 2 (146 anatomical or functional groups available) in 2366 SN (96.97%), at level 3 (multiple types available) in 2304 SN (94.45%). The developed interface (Figure 1) allows querying the database by manufactures, devices, type of SN (1a), and selecting the EMDN nomenclature up to the fourth level (1b). As a result, the relevant information is shown, including trends over the selected period and the link to the SN on the original website (2).

Conclusions

The proposed approach was able to cope with the uncompleteness of the publicly available data in the SN, thus allowing proper matching of MD with its EMDN code up to level 3 with very good performance. In this way, grouping of SN relevant to a specific MD category/group/type could be used as possible sentinel for increased rates in reported serious incidents in high-risk MD. Extension of this approach to aggregate SN from other EU nations could result in an effective support tool in PMS.

Funding Acknowledgement

Type of funding sources: Public grant(s) – EU funding. Main funding source(s): EU Horizon 2020 - Project CORE-MD

A deep-learning approach for scar detection in early contrast-enhanced cardiac computed tomography images

Background/Introduction

Diagnosis of myocardial fibrosis has a relevant prognostic and therapeutic role. Scar-tissue analysis is commonly performed with late gadolinium contrast-enhanced (CE) cardiac magnetic resonance (CMR). However, CMR might be contraindicated or unavailable. Coronary computed tomography (CCT), which is the technique of choice for many routine assessments, is emerging as an alternative to CMR.

Purpose

The objective of this study is to evaluate whether artificial intelligence (AI) could allow identification of myocardial fibrosis from routine early CE-CCT images.

Methods

Fifty consecutive patients with left ventricular dysfunction (LVD), who underwent both CE-CMR and (early and late) CE-CCT were retrospectively selected. According to the late enhancement CMR patterns, patients were classified as with ischemic (n=15, 30%) or non-ischemic (n=25, 70%) LVD.

Scar regions were manually traced on late CE-CCT using CE-CMR as ground-truth.

On early CE-CCT images, the myocardial sectors were extracted according to AHA 16-segment model and labeled as with scar or not based on the late CE-CCT manual tracing.

A deep-learning model was developed to classify each sector (Picture 1).

Results

Of the initial 44187 sectors computed out of the 8285 slices available from the early CE-CCT images, 4594 sectors (10%) presented scar.

The CNN approach on the early CE-CCT images yielded a classification accuracy for all sectors of 71% (95% confidence interval (CI): 63%-79%) obtained through 5-fold cross validation.

The mean sensitivity, positive predictive value (PPV) and negative predictive value (NPV) for the testing fold resulted in 73% (95% CI: 66%-79%), 56% (95% CI: 48%-65%) and 85% (95% CI: 82%- 88%), respectively. The mean AUC across the five folds was 76% (95% CI: 72%-81%).

In a per-segment analysis of the 16-segment AHA model the bull's eye segmental comparison of CE-CMR and respective early CE-CCT findings an 91% agreement was achieved (Picture 2).

Conclusions

AI can detect both ischemic and non-ischemic myocardial fibrosis from routine noninvasive coronary scans, without additional contrast-agent administration or radiational dose, thus assisting diagnosis and management of patients with LV dysfunction and coronary artery disease.

Funding Acknowledgement

Type of funding sources: None.

Digital therapeutics: a systematic review of clinical trials characteristics

Introduction

Digital therapeutics (DTx) are a subset of digital health tools delivering evidence-based therapeutic interventions that are driven by high quality software programs to prevent, manage, or treat a medical disorder or disease. They are studied using randomized clinical trial methodology and reviewed, cleared or certified by regulatory bodies as required to support product claims regarding risk, efficacy, and intended use.

Purpose

To perform a systematic review of clinical research/studies conducted in the field of DTx with the aim to describe studies where DTx were used, classifying them by digital intervention and condition, and analysing and reporting the characteristics of clinical trials.

Methods

The U.S. National Library of Medicine ClinicalTrials.gov was searched using the terms “digital therapeutics”, “digital therapeutic”, “digital therapy”, and “digital therapies” within the fields “Intervention/treatment” and “Title/Acronym”, and the resulting trial characteristics were extracted and analysed.

Results

In total, 560 clinical trials were retrieved on January 10, 2021. Most of them (n=424, 75.7%) were excluded because they were observational studies (n=82), non-randomized/single arm assignment studies (n=123), not involving any digital health tool (n=181), or involving digital health tools not classified as DTx (n=38). Of the remaining 136 trials, the DTx intervention was delivered through apps (n=57, 41.9%), web-based systems (n=35, 25.7%), videogames (n=12, 8.8%), virtual reality (n=6, 4.4%), text messages (n=5, 3.7%), social media platform (n=4, 2.9%), computer-based systems (n=3, 2.2%), or other (n=14, 10.3%), and applied to the following clinical scenarios: mental health (n=47, 34.6%), chronic pain and chronic diseases (n=26, 19.1%), smoking and other substances abuse or addiction (n=17, 12.5%), insomnia and sleeping disorders (n=12, 8.8%), obesity and physical activity (n=11, 8.1%), cardiovascular diseases (n=10, 7.3%), and other conditions (n=13, 9.6%). Apps were used more frequently for chronic pain (54%) and sleeping disorders (67%), while videogames and web-based systems were adopted for mental health in 21.3% and 38% of the trials, respectively, and text messages were preferred in 18.2% of obesity and 17.6% of addiction trials, respectively Sixty-eight trials (50%) started in the last three years (2019, 2020, and 2021), while 54.4% were ongoing, 33.8% completed, 2.9% stopped early, and 8.8% with status unknown.

Conclusions

The term “digital therapeutics” was very often incorrectly used by researchers when they register their trials in ClinicalTrials.gov, improperly including studies involving the use of digital health tools to support a drug intake or monitor a condition. Mobile apps, web-based systems and videogames were the most adopted technologies to deliver DTx, while mental health diseases, chronic pain, and addiction were the conditions in which they were most frequently studied.

Funding Acknowledgement

Type of funding sources: None.

A framework for risk stratification of burnout due to Covid-19 in nurses using personas" creation

Funding Acknowledgements

Type of funding sources: None.

Background

Covid-19 pandemic has generated a huge amount of stress upon all healthcare workers, with an increased risk of developing burnout syndrome and Post Traumatic Stress Disorder (PTSD). Mobile health could provide tools for preventive interventions, whose efficacy could be improved by applying User-Centered Design (UCD) techniques to create personas to stratify the associated risk and tailor those interventions.

Purpose

We propose a framework for the definition of personas applied to an Italian nurse population to highlight different levels of risk relevant to burnout, anxiety, depression, and PTSD induced by Covid-19.

Methods

Data was gathered by web-questionnaire delivered to nurses in Italy between April and May 2020. Quantitative information about respondents’ lifestyle, professional experience, impact of Covid-19 pandemic on both work conditions and personal life were collected, together with psychological indexes as Maslach Burnout Inventory (MBI), Impact of Event Scale (IES) and Patient Health Questionnaire (PHQ), used to define the risk level of burnout, PTSD, and anxiety and depression, respectively. Collected variables (55) were first reduced using Principal Component Analysis; then, k-medoids clustering with Partitioning Around Medoids algorithm was applied, where the best value of k was determined by silhouette and total within sum of square differences heuristics, together with evaluation from domain experts. Statistical analysis was applied to define which variables were able to differentiate among the k clusters, constituting the basis for the definition of a Persona card (i.e., a template with textual and graphical information) for each of the obtained clusters.

Results

From the 175 respondents nurses, a total of three clusters (personas) were identified. They presented very strong dissimilarities both in lifestyle and in how nurses reacted to the pandemic, showing distinct profiles associated to three different levels of risk of developing burnout and PSTD: the highest (representing 38.3% of all respondents), associated to a married nurse with children living with family, also using Personal Protective Equipment (PPE) at home, with workload highly impacted by COVID-19; the middle one, associated to an engaged nurse living alone with workload highly impacted by COVID-19; the lowest, associated to a married nurse living only with the husband/wife and wearing PPE at home, with workload moderately impacted by the pandemic. All personas showed a good amount of knowledge about burnout.

Conclusions

The application of the proposed framework to nurses showed three clusters that evidenced three risk profiles associated to different personas, each requiring distinctive treatment. This approach constitutes the first step for the development of tailored m-health interventions aimed at monitoring and preventing burnout and PTSD among healthcare workers in critical situations, such as the one provoked by Covid-19 pandemic.

P586Left ventricular volume and mass adaptation after 58-days head-down bed-rest assessed by cine-MRI, and effectiveness of high-intensity jump training countermeasure

Aims

Prolonged immobilization generates cardiac deconditioning, a risk factor for cardiovascular disease, and efficient countermeasures (CM) are needed to prevent it. Our aim was to assess by Cine-MRI the effects of long-term strict head-down (−6 degrees) bed-rest (BR) deconditioning, and the effectiveness of high-intensity jump training CM, on left ventricular (LV) function and mass.

Methods

23 male participants (29±6 years, 181±6 cm, 77±7 kg) were enrolled. The experiment was conducted at: envihab (Koln, DLR, Germany) as part of the European Space Agency BR studies. Volunteers were randomly allocated to the jump training group (JUMP, n=12) or the control group (CTRL, n=11). A typical training session consisted of 4x10 countermovement jumps and 2x10 hops in a sledge jump system, with 5–6 sessions per week.

Steady-state free precession cine-MRI images were obtained (25 frames/cardiac cycle, 8mm thickness, no gap, no overlap) as stack of short-axis images covering the whole LV from base to apex, before (PRE) and after 58-days (HDT58) of BR. Endocardial and epicardial semi-automated contouring was performed using custom software.

Results

In CTRL group, at HDT58 a reduction in LV mass (9%), end-diastolic (21%), end-systolic (8%) and stroke volume (23%) were observed, while ejection fraction did not change. In JUMP group, the reduction in LV end-diastolic volume was only by 9%, followed by a decrease in end-systolic (10%) and stroke volume (10%), with a preservation of LV mass. In both groups, ejection fraction did not change.

Left ventricular changes after 58d BR EDV (ml) ESV (ml) SV (ml) EF (%) Mass (g) CTRL PRE 171 (163; 191) 60 (58; 65) 113 (102; 124) 64 (63; 67) 133 (121; 160) HDT58 143 (131; 149)* 56 (50; 58)* 86 (83; 92)* 62 (60; 63)* 127 (106; 138)* JUMP PRE 156 (140; 204) 70 (61; 84) 91 (78; 117) 58 (56; 60) 118 (104; 134) HDT58 141 (135; 175)* 63 (54; 69)* 90 (78; 92)* 58 (55; 60) 113 (105; 128) Results expressed as median (25th; 75th percentiles). *p<0.05 Wilcoxon non parametric paired test (PRE vs HDT58). CTRL: control group; JUMP: countermeasure group.

Conclusions

Cardiac adaptation to deconditioning due to immobilization induced by BR resulted in a reduction of cardiac volumes and function, together with a decrease in LV mass. Interestingly, the applied JUMP countermeasure appeared able to partially reverse these effects, in particular by reducing the decrease in end-diastolic volume and preserving LV mass. This information could be useful for better understanding physiologic changes in patients undergoing long periods of immobilization, as well as to apply the studied countermeasure during space flight to reduce cardiac deconditioning.

Acknowledgement/Funding

Italian Space Agency (contract 2018-7-U.0),CNES/DAR 48ehz747.01950965,BELSPO, via the European Space Agency PRODEX program (PEA 4000110826)

Sensitivity analysis of ventricular activation and electrocardiogram in tailored models of heart-failure patients

Cardiac resynchronization therapy is not effective in a variable proportion of heart failure patients. An accurate knowledge of each patient's electroanatomical features could be helpful to determine the most appropriate treatment. The goal of this study was to analyze and quantify the sensitivity of left ventricular (LV) activation and the electrocardiogram (ECG) to changes in 39 parameters used to tune realistic anatomical-electrophysiological models of the heart. Electrical activity in the ventricles was simulated using a reaction-diffusion equation. To simulate cellular electrophysiology, the Ten Tusscher-Panfilov 2006 model was used. Intracardiac electrograms and 12-lead ECGs were computed by solving the bidomain equation. Parameters showing the highest sensitivity values were similar in the six patients studied. QRS complex and LV activation times were modulated by the sodium current, the cell surface-to-volume ratio in the LV, and tissue conductivities. The T-wave was modulated by the calcium and rectifier-potassium currents, and the cell surface-to-volume ratio in both ventricles. We conclude that homogeneous changes in ionic currents entail similar effects in all ECG leads, whereas the effects of changes in tissue properties show larger inter-lead variability. The effects of parameter variations are highly consistent between patients and most of the model tuning could be performed with only ~10 parameters.

Beat-to-beat heart rate detection by smartphone's accelerometers: validation with ECG

Mobile phones offer the possibility to monitor and track health parameters. Our aim was to test the feasibility and accuracy of measuring beat-to-beat heart rate using smartphone accelerometers by recording the vibrations generated by the heart during its function and transmitted to the chest wall, i.e. the so-called seismocardiographic signal (SCG).

METHODS

9 healthy male volunteers were studied in supine (SUP) and in standing (ST) posture. A smartphone (iPhone6, Apple) was positioned on the thorax (POS1) to acquire SCG signal. While supine, a second smartphone was positioned on the navel (POS2). The SCG signal was recorded for 3 minutes during spontaneous respiration, synchronous with 3-leads ECG. Using a fully automated algorithm based on amplitude thresholding after rectification, the characteristic peak of the SCG signal (IVC) was detected and used to compute beat-to-beat heart duration, to be compared with the corresponding RR intervals extracted from the ECG.

RESULTS

A 100% feasibility of the approach resulted for POS1 in SUP, while 89% in POS2, and 78% for POS1 in ST. In supine, for each smartphones' position, the automated algorithm correctly identified the cardiac beats with >98% accuracy. Linear correlation (r2) with RR was very high (>0.98) in each posture and position, with no bias and narrow limits of agreement.

CONCLUSIONS

The obtained results proved the feasibility of the proposed approach and the robustness of the applied algorithm in measuring the beat-to-beat heart rate from smartphone-derived SCG, with high accuracy compared to conventional ECG-derived measure.

Evaluation of respiratory- and postural-induced changes on the ballistocardiogram signal by time warping averaging

OBJECTIVE

The aim of this work was to evaluate the potential changes in the ballistocardiogram (BCG) signal induced by different respiratory patterns and posture, by using the dynamic time warping (DTW) technique.

APPROACH

BCG signals were recorded in a group of 20 healthy volunteers, simultaneously with an electrocardiogram (ECG). Two recordings, one in a supine (SUP) and one in a standing (ST) position, including spontaneous breathing, two 1 min apneas (at full and empty-lungs, respectively) and 30 s of Valsalva, were analyzed. A warped averaged waveform was computed for each phase, from which amplitude and temporal parameters were extracted to characterize each condition.

MAIN RESULTS

Variations were found in both amplitude and duration of BCG-derived parameters among manoeuvres, especially when compared to spontaneous breathing, suggesting a complex interaction between intra-thoracic pressure changes acting on venous return, together with the autonomic nervous system modulation on heart rate. The effect of a hydrostatic pressure gradient elicited by postural conditions was also evident.

SIGNIFICANCE

Posture and respiratory manoeuvres affect the BCG signal in different ways, probably as a result of changes induced in preload and afterload. This supports the need to define separate normality ranges for each posture and/or breathing conditions, as well as the importance of applying specific manoeuvres to highlight any pathological response in the computed BCG parameters.

Evaluation of ensemble averaging methods in 3D ballistocardiography

Ballistocardiography (BCG) is a non-invasive technique which measures the acceleration of a body induced by cardiovascular activity, namely the force exerted by the beating heart. Measuring a BCG in a gravity-free environment provides ideal conditions where the subject is completely decoupled from its environment. Furthermore, because gravity constrains the motion in two dimensions, the non-negligible accelerations taking place in the third dimension are lost. In every experimental situation, the measured BCG signal contains artifacts pertaining to different causes. One of them is the undesirable involuntary movements of the subject. Ensemble averaging (EA) tackles the issue of constructing a typical one cardiac cycle BCG signal which best represents a longer recording. The present work compares state-of-the-art EA methods and proposes two novel techniques, one taking into account the ECG sub-intervals and the other one based on Dynamic Time Warping. The effects of lung volume are also assessed.

Effects of 5 days of head-down bed rest, with and without short-arm centrifugation as countermeasure, on cardiac function in males (BR-AG1 study)

This study examined cardiac remodeling and functional changes induced by 5 days of head-down (−6°) bed rest (HDBR) and the effectiveness of short-arm centrifugation (SAC) in preventing them in males. Twelve healthy men (mean age: 33 ± 7) were enrolled in a crossover design study (BR-AG1, European Space Agency), including one sedentary (CTRL) and two daily SAC countermeasures (SAC1, 30 min continuously; SAC2, 30 min intermittently) groups. Measurements included plasma and blood volume and left ventricular (LV) and atrial (LA) dimensions by transthoracic echocardiography (2- and 3-dimensional) and Doppler inflows. Results showed that 5 days of HDBR had a major impact on both the geometry and cardiac function in males. LV mass and volume decreased by 16 and 14%, respectively; LA volume was reduced by 36%; Doppler flow and tissue Doppler velocities were reduced during early filling by 18 and 12%, respectively; and aortic flow velocity time integral was decreased by 18% with a 3% shortening of LV ejection time. These modifications were presumably due to decreased physiological loading and dehydration, resulting in reduced plasma and blood volume. All these changes were fully reversed 3 days after termination of HDBR. Moreover, SAC was not able to counteract these changes, either when applied continuously or intermittently.

Left ventricular modelling: a quantitative functional assessment tool based on cardiac magnetic resonance imaging

We present the development and testing of a semi-automated tool to support the diagnosis of left ventricle (LV) dysfunctions from cardiac magnetic resonance (CMR). CMR short-axis images of the LVs were obtained in 15 patients and processed to detect endocardial and epicardial contours and compute volume, mass and regional wall motion (WM). Results were compared with those obtained from manual tracing by an expert cardiologist. Nearest neighbour tracking and finite-element theory were merged to calculate local myocardial strains and torsion. The method was tested on a virtual phantom, on a healthy LV and on two ischaemic LVs with different severity of the pathology. Automated analysis of CMR data was feasible in 13/15 patients: computed LV volumes and wall mass correlated well with manually extracted data. The detection of regional WM abnormalities showed good sensitivity (77.8%), specificity (85.1%) and accuracy (82%). On the virtual phantom, computed local strains differed by less than 14 per cent from the results of commercial finite-element solver. Strain calculation on the healthy LV showed uniform and synchronized circumferential strains, with peak shortening of about 20 per cent at end systole, progressively higher systolic wall thickening going from base to apex, and a 10° torsion. In the two pathological LVs, synchronicity and homogeneity were partially lost, anomalies being more evident for the more severely injured LV. Moreover, LV torsion was dramatically reduced. Preliminary testing confirmed the validity of our approach, which allowed for the fast analysis of LV function, even though future improvements are possible.

Assessment of the cardiovascular regulation during robotic assisted locomotion in normal subjects: autoregressive spectral analysis vs empirical mode decomposition

Robotic assisted locomotion systems are recently gaining appreciation as methods to rehabilitate individuals with lost sensory motor function. In the present study we compare autoregressive power spectral analysis and empirical mode decomposition (EMD) applied to the analysis of short-term heart period variability regarding their ability to typify autonomic response during a robotic assisted locomotion session consisting in the following phases: 1) sitting position; 2) standing position; 3) suspension during subject instrumentation; 4) robotic assisted treadmill locomotion with partial body weight support; 5) standing recovery after exercise. Results showed a significant tachycardia during the suspension phase, but no significant changes of spectral indexes. On the contrary, when spectral indexes were derived according to EMD, changes were evidenced during the suspension and walking phases. The EMD method is more powerful than autoregressive spectral analysis in detecting variations of parasympathetic and sympathetic modulations elicited by a robotic-assisted locomotion protocol.

Empirical mode decomposition to assess baroreflex gain from spontaneous variability during exercise in humans

Estimation of the baroreflex gain has become an important tool in clinical practice in order to assess cardiac autonomic system control. Spectral analysis and sequence analysis techniques based on the spontaneous variability of systolic arterial pressure and heart period have been proposed to evaluate the baroreflex gain. These analyses can be significantly altered by the presence of nonstationarities. Recently, the empirical mode decomposition (EMD), a signal processing technique particularly suitable for nonstationary series, has been proposed as a new tool for data analysis. The aim of this study is to propose EMD-based approaches to the evaluation of the baroreflex gain to account for the possible presence of nonstationarities of systolic arterial pressure and heart period series.

Semi-automated analysis of coronary flow Doppler images: validation with manual tracings

Coronary flow velocity reserve (CFVR) is conventionally obtained by manual tracings of Doppler profiles, as ratio of stress vs baseline diastolic peak velocity. When <1.9, this parameter evidences reduced coronary flow and possible microcirculatory disease. Our goals were: 1) to develop a novel technique for semi-automated detection of Doppler flow velocity profile, allowing the automated computation of CFVR and other parameters; 2) to validate this technique in comparison with conventional measurements obtained by manual tracing; 3) to test for differences between normal (N) subjects and patients with rheumatoid arthritis (RA). Linear correlation and Bland-Altman analyses showed that the proposed method was highly accurate and repeatable compared to the manual measurements. Comparison between N and RA groups evidenced significant differences in some of the automated parameters.

Evaluation of alterations on mitral annulus velocities, strain, and strain rates due to abrupt changes in preload elicited by parabolic flight

We tested the hypothesis that in normal subjects, cardiac tissue velocities, strain, and strain rates (SR), measured by Doppler tissue echocardiography (DTE), are preload dependent. To accomplish it, immediately preceding image acquisition, reversible, repeatable, acute nonpharmacological changes in preload were induced by parabolic flight. DTE has been proposed as a new approach to assess left ventricular regional myocardial function by computing tissue velocities, strain, and SR. However, preload dependence of these parameters in normal subjects still remains controversial. DTE images (Philips) were obtained in 10 normal subjects in standing upright position at normogravity (1 G z), hypergravity (1.8 G z), and microgravity (0 G z) with and without −50 mmHg lower body negative pressure (LBNP). Myocardial velocity curves in the basal interventricular septum were reconstituted offline from DTE images, from which peak systolic (S′), early (E′) and late (A′) diastolic velocities, SR, and peak systolic strain (PSε) were measured and averaged over four beats. At 1.8 G z (reduced venous return), S′, E′, and A′ decreased by 21%, 21%, and 26%, respectively, compared with 1-G z values, while at 0 G z (augmented venous return), E′, A′, and PSε increased by 57%, 53%, and 49%, respectively. LBNP reduced E′ and PSε. In conclusion, our results were in agreement with those obtained in animal models, in which preload was changed in a controlled, acute, and reversible manner, and image acquisition was performed immediately following preload modifications. The hypothesis of preload dependence was confirmed for S′, E′, A′, and PSε, while SR appeared to be preload independent, probably reflecting intrinsic myocardial properties.

Quantification of coronary flow velocity reserve by means of semiautomated analysis of coronary flow Doppler images

Coronary flow reserve (CFVR) is conventionally obtained by manual tracings of Doppler profiles, as ratio of control vs stress diastolic peak velocity. This parameter could help in discriminating between normal (N) and microcirculatory pathologic (P) subjects, even the clinical meaning of 1.9<CFVR</=2.5 it's still not clear. Our goal was to develop a technique for automated tracings of Doppler flow velocity profile to reduce subjectivity and to allow the extraction of other parameters, which could be of clinical interest. This technique was applied to 15 N (CFVR>3) and 15 P (CFVR<1.8) subjects, to assess whose of the new parameters could be able to discriminate between these groups. Results indicated that many of the new parameters were able to evidence significant differences between N and P, thus representing new clinical indices useful for the diagnosis.

Nearly automated analysis of coronary Doppler flow velocity from transthoracic ultrasound images: validation with manual tracings

Coronary flow velocity reserve is obtained by manual tracings of transthoracic coronary Doppler flow velocity profiles as the ratio of stress versus baseline diastolic peak velocities. This approach introduces subjectivity in the measurements and limits the information which could be exploited from the Doppler velocity profile. Accordingly, our goals were to develop a technique for nearly automated detection of Doppler coronary flow velocity profile, and automatically compute both conventional and additional amplitude, derivative and temporal parameters, and validate it with manual tracings. A total of 100 patients (17 normals, 15 patients with severe coronary stenosis, 41 with connective tissue disease and 27 with diabetes mellitus) were studied. Linear correlation and Bland-Altman analyses showed that the proposed method was highly accurate and repeatable compared to the manual measurements. Comparison between groups evidenced significant differences in some of the automated parameters, thus representing potentially additional indices useful for the noninvasive diagnosis of microcirculatory or coronary artery disease.

Cardiac response to robotic assisted locomotion in normal subjects: a preliminary study

Robotic assisted locomotion systems are recently gaining appreciation and diffusion as useful methods to rehabilitate individuals with lost sensorimotor function. Our aim was to evaluate potential changes in the autonomic nervous system activity (by ECG and spectral analysis), due to the experimental protocol, which include suspension of the subject to be instrumented on the system. A group of 10 normal subjects was studied during the rehabilitation protocol. Results showed a significant tachycardia and a reduced variance, during orthostatic stress induced by the suspension phase in comparison with sitting baseline condition but no significant increase of LF normalized power as it would be expected during a sympathetic activation.

Objective evaluation of changes in left ventricular and atrial volumes during parabolic flight using real-time three-dimensional echocardiography

We tested the feasibility of real-time three-dimensional (3D) echocardiographic (RT3DE) imaging to measure left heart volumes at different gravity during parabolic flight and studied the effects of lower body negative pressure (LBNP) as a countermeasure. Weightlessness-related changes in cardiac function have been previously studied during spaceflights using both 2D and 3D echocardiography. Several technical factors, such as inability to provide real-time analysis and the need for laborious endocardial definition, have limited its usefulness. RT3DE imaging overcomes these limitations by acquiring real-time pyramidal data sets encompassing the entire ventricle. RT3DE data sets were obtained (Philips 7500, X3) during breath hold in 16 unmedicated normal subjects in upright standing position at different gravity phases during parabolic flight (normogravity, 1 Gz; hypergravity, 1.8 Gz; microgravity, 0 Gz), with LBNP applied (−50 mmHg) at 0 Gz in selected parabolas. RT3DE imaging during parabolic flight was feasible in 14 of 16 subjects. Data were analyzed (Tomtec) to quantify left ventricular (LV) and atrial (LA) volumes at end diastole and end systole, which significantly decreased at 1.8 Gz and increased at 0 Gz. While ejection fraction did not change with gravity, stroke volume was reduced by 16% at 1.8 Gz and increased by 20% at 0 Gz, but it was not significantly different from 1 Gz values with LBNP. RT3DE during parabolic flight is feasible and provides the basis for accurate quantification of LV and LA volume changes with gravity. As LBNP counteracted the increase of LV and LA volumes caused by changes in venous return, it may be effectively used for preventing cardiac dilatation during 0 Gz.

Improved quantification of left ventricular mass based on endocardial and epicardial surface detection with real time three dimensional echocardiography

OBJECTIVE

To develop a technique for volumetric analysis of real time three dimensional echocardiography (RT3DE) data aimed at quantifying left ventricular (LV) mass and to validate the technique against magnetic resonance (MR) assumed as the reference standard.

DESIGN

RT3DE, which has recently become widely available, provides dynamic pyramidal data structures that encompass the entire heart and allows four dimensional assessment of cardiac anatomy and function. However, analysis techniques for the quantification of LV mass from RT3DE data are fundamentally two dimensional, rely on geometric modelling, and do not fully exploit the volumetric information contained in RT3DE datasets. Twenty one patients underwent two dimensional echocardiography (2DE), RT3DE, and cardiac MR. LV mass was measured from 2DE and MR images by conventional techniques. RT3DE data were analysed to semiautomatically detect endocardial and epicardial LV surfaces by the level set approach. From the detected surfaces, LV mass was computed directly in the three dimensional space as voxel counts.

RESULTS

RT3DE measurement was feasible in 19 of 21 patients and resulted in higher correlation with MR (r = 0.96) than did 2DE (r = 0.79). RT3DE measurements also had a significantly smaller bias (-2.1 g) and tighter limits of agreement (2SD = +/-23 g) with MR than did the 2DE values (bias (2SD) -34.9 (50) g). Additionally, interobserver variability of RT3DE (12.5%) was significantly lower than that of 2DE (24.1%).

CONCLUSIONS

Direct three dimensional model independent LV mass measurement from RT3DE images is feasible in the clinical setting and provides fast and accurate assessment of LV mass, superior to the two dimensional analysis techniques.

Tracking of Left Ventricular Long Axis From Real-Time Three-Dimensional Echocardiography Using Optical Flow Techniques

Two-dimensional echocardiography (2DE) is routinely used in clinical practice to measure left ventricular (LV) mass, dimensions, and function. The reliability of these measurements is highly dependent on the ability to obtain nonforeshortened long axis (LA) images of the left ventricle from transthoracic apical acoustic windows. Real time three-dimensional echocardiography (RT3DE) is a novel imaging technique that allows the acquisition of dynamic pyramidal data structures encompassing the entire ventricle and could potentially overcome the effects of LA foreshortening. Accordingly, the aim of this paper was to develop a nearly automated method based on optical flow techniques for the measurement of the left ventricular (LV) LA throughout the cardiac cycle from RT3DE data. The LV LA measurements obtained with the automated technique has been compared with LA measurements derived from manual selection of the LA from a volumetric display of RT3DE data. High correlation (r = .99, SEE = 1.8%, y = .94x + 5.3), no significant bias (-0.18 mm), and narrow limits of agreement (SD: 1.91 mm) were found. The comparison between the LA length derived from 2DE and RT3DE data showed significant underestimation of the 2DE based measurements. In conclusion, this study proves that RT3DE data overcome the effects of foreshortening and indicates that the method we propose allows fast and accurate quantification of LA length throughout the cardiac cycle.

Quantification of left ventricular modification in weightlessness conditions from the spatio-temporal analysis of 2D echocardiographic images

Two-dimensional echocardiography (2DE) performed during flights with a parabolic trajectory to simulate weightlessness provides a unique means to study left ventricular (LV) modifications to prevent post-flight orthostatic intolerance in astronauts. However, conventional analysis of 2DE is based on manual tracings and depends on experience. Accordingly, the aim was objectively to quantify, from 2DE images, the LV modifications related to different gravity levels, by applying a semi-automated level-set border detection technique. The algorithm validation was performed by the comparison of manual tracing results, obtained by two independent observers with 20 images, with the semi-automated measurements. To quantify LV modifications, three consecutive cardiac cycles were analysed for each gravity phase (1 Gz, 1.8 Gz, 0 Gz). The level-set procedure was applied frame-by-frame to detect the LV endocardial contours and obtain LV area against time curves, from which end-diastolic (EDA) and end-systolic (ESA) areas were computed and averaged to compensate for respiratory variations. Linear regression (y = 0.91x + 1.47, r = 0.99, SEE:0.80cm2) and Bland-Altman analysis (bias = -0.58 cm2, 95% limits of agreement= +/- 2.14cm2) showed excellent correlation between the semi-automatic and manually traced values. Inter-observer variability was 5.4%, and the inter-technique variability was 4.1%. Modifications in LV dimensions during the parabola were found: compared with 1 Gz values, EDA and ESA were significantly reduced at 1.8 Gz by 8.8 +/- 5.5% and 12.1 +/- 10.1%, respectively, whereas, during 0 Gz, EDA and ESA increased by 13.3 +/- 7.3% and 11.6 +/- 5.1%, respectively, owing to abrupt changes in venous return. The proposed method resulted in fast and reliable estimations of LV dimensions, whose changes caused by different gravity conditions were objectively quantified.

Feasibility of real-time 3D echocardiography in weightlessness during parabolic flight

Aim of the study was to test the feasibility of transthoracic real-time 3D (Philips) echocardiography (RT3D) during parabolic flight, to allow direct measurement of heart chambers volumes modifications during the parabola. One RT3D dataset corresponding to one cardiac cycle was acquired at each gravity phase (1 Gz, 1.8 Gz, 0 Gz, 1.8 Gz) during breath-hold in 8 unmedicated normal subjects (41 +/- 8 years old) in standing upright position. Preliminary results, obtained by semi-automatically tracing left ventricular (LV) and left atrial (LA) endocardial contours in multiple views (Tomtec), showed a significant (p<0.05) reduction, compared to 1 Gz, of LV and LA volumes with 1.8 Gz, and a significant increase with 0 Gz. Further analysis will focus on the right heart.

Changes in Doppler mitral inflow patterns during parabolic flight

Aim of the study was to evaluate by transthoracic Doppler the alterations in mitral inflow velocity pattern caused by acute changes in loading conditions occurring during parabolic flights. Each parabola included normogravity (1 Gz, 1 min), mild hypergravity (1.8 Gz, 20 sec), microgravity (0 Gz, 24 sec) and mild hypergravity (1.8 Gz, 20 sec) phases. Pulsed-Doppler images were digitally acquired in 11 unmedicated subjects (46 +/- 5 years), in standing upright position and supine resting. Doppler profiles were semi-automatically traced and inflow parameters extracted and averaged onto three consecutive beats. Only in standing position, significant alterations during microgravity (p<0.05) were noted in several parameters.

Time-variant spectral analysis of heart rate variability during parabolic flight with and without LBNP

Modifications of autonomic activity during parabolic flight were studied by a time-variant model able to estimate low (LF, 0.04-0.14 Hz) and high (HF, 0.14-0.35 Hz) frequency spectral components on a beat-to-beat basis. Ten subjects were studied with and without lower body negative pressure (LBNP). ECG and Gz load were digitized (500 Hz) and RR interval variability series extracted. Beat-to-beat mean RR, variance, LF and HF power were obtained. One-way ANOVA (p<0.01) was used to compare values obtained during starting 1Gz (I), first 1.8Gz (II), 0Gz (III), second 1.8Gz (IV), ending 1Gz (V). Without LBNP, total and LF power increased during 0Gz to 1.69 +/- 1.41 and 2.87 +/- 4.66 respectively (mean +/- SD, normalized by phase I value). With LBNP, their change during 0Gz (1.38 +/- 1.37 and 1.54 +/- l.04 respectively) reached significance only with phase II and phase V. Phase I HF power was higher than in the other phases, both without and with LBNP.

Analysis of cardiac left-ventricular volume based on time warping averaging

The cardiac left-ventricular (LV) volume signal, obtained by acoustic quantification, is affected by noise and respiratory modulation, resulting in a large beat-to-beat variability that affects the computation of LV function indices. A new method is proposed to improve the evaluation of LV indices by applying a signal averaging technique based on dynamic time warping to consecutive LV volume waveforms. Volume signals obtained from ten normal young (NY) subjects (mean age +/- SD: 25+/-5 years) were used to evaluate the performance of this algorithm. To evaluate its clinical utility, the effects of ageing and pharmacologically induced changes on LV function were assessed by studying, respectively, ten normal (N) adult subjects (age 64+/-8 years) and ten patients with dilated cardiomyopathy during a control and low-dose dobutamine (10 microg kg(-1) min(-1)) study. Indices of LV function were highly consistent, with a variability of less than 8%, even when only 16 beats were averaged, independently of their selection inside the whole recording. When compared with beat-to-beat measures, the averaging of 16 beats significantly reduced (by more than 50%) the interbeat variability of all indexes. Expected alterations in both diastolic and systolic function were evidenced both with ageing (peak filling atrial contraction and ejection rates: from 275+/-77 mls(-1), 76+/-30 ml s(-1) 230+/-70 mls(-1), respectively, in NY, to 160+/-33 mls(-1), 125+/-39 mls(-1), 163+/-54 mls(-1) in N) and with dobutamine (peak filling and ejection rates from 160+/-72 mls(-1) and 183+/-86 mls(-1) respectively, in control, to 253+/-75 mls(-1) and 251+/-105 mls(-1) with dobutamine). Signal averaging with time warping allows fast and improved assessment of LV function.

Changes in left ventricular size during parabolic flights by two-dimensional echocardiography and level set method

This study aims to evaluate changes on cardiac chambers size, induced by gravitational stresses. During parabolic flight, seven subjects underwent 2-D transthoracic echocardiography at three different gravity phases (1 Gz, 1.8 Gz, and 0 Gz). LV endocardial borders were detected applying a semi-automatic segmentation procedure based on level set methods. LV cavity area was computed frame-by-frame for a whole cardiac cycle during each gravity phase. Expected modifications in LV area with different gravity were found: at 1.8 Gz, end-diastolic (ED) and end-systolic (ES) areas were significantly (p<0.05) reduced of 10.7 +/- 5.4% and 21.6 +/- 11.1% respectively, compared to 1 Gz values, while they were increased of 11.2 +/- 5.4% and 11.1 +/- 6% during 0 Gz. Fractional area change was augmented of 20.9 +/- 29.1% at 1.8 Gz, while it remained unchanged at 0 Gz, compared with 1 Gz values. Furthermore, LV filling due to atrial contraction was increased at 0 Gz of 39 +/- 35.6%.

RR-arterial pressure variability relationships

Methodological aspects of a causal black-box model of heart period/arterial pressure interaction, arterial pressure closed-loop regulation and respiration effects on both heart period and arterial pressure are revisited in the "time" (more exactly heart beat count) domain. Parameters are estimated from experimental data (model identification) by means of multiple linear regressions of actual samples over the past ones. The elements composing either heart period or systolic arterial pressure variability are visualised as beat-by-beat series. Indexes describing the signal interactions, the loop properties and the spectral components of the variability series are consequently summarised. In 17 normal young volunteers, the analysis was carried out during active standing, rest, mild clinostatic pedalling exercise at 10%, 20%, and 30% of the maximum effort, and recovery. A negative effect of heart period changes on systolic arterial pressure of - 13.3 mm Hg/s was found at rest. This effect, though augmented by exercise, appeared insignificant in explaining arterial pressure variability. Arterial baroreflex was assessed by alphaart index which had a value of 5.18 mm Hg/ms at rest, 3.78 mm Hg/ms during active standing, and decreased progressively with exercise down to 0.55 mmHg/ms. The pressure regulation loop displayed a tendency to amplify disturbances at low frequency (around 0.1 Hz) 5.94 times at rest, augmented to 8.88 times during standing, 7.55 at 30% exercise. The first parameter of the pressure auto-regression was slightly higher than 1 at rest and even more augmented during standing, thus, indicating a tendency of arterial pressure perturbations to persist from one beat to the next. These mechanisms appear important in the genesis of low-frequency pressure waves. Nonetheless, the trace of different sources was evident in the regression residuals. Noticeably, during exercise it explained 10.16% of total heart period variability compared to 12.49% related to the low-frequency oscillations of closed-loops. The origin of high-frequency waves synchronous with respiration appeared miscellaneous as well. Arterial pressure appeared negligibly affected by heart period changes. Conversely, a limited effect of arterial pressure waves was found on heart period superimposed to a large effect of cardiopulmonary reflexes directly modulating the sinus node. In conclusion, both high-frequency and low-frequency waves are composite phenomena and a multi-channel analysis comparing heart period and arterial pressure variability yields a variety of figures assessing cardiovascular regulation and cardiorespiratory coupling.

Combined assessment of myocardial perfusion and regional left ventricular function by analysis of contrast-enhanced power modulation images

BACKGROUND

Echocardiographic contrast media have been used to assess myocardial perfusion and to enhance endocardial definition for improved assessment of left ventricular (LV) function. These methodologies, however, have been qualitative or have required extensive offline image analysis. Power modulation is a recently developed imaging technique that provides selective enhancement of microbubble-generated reflections. Our goal was to test the feasibility of using power modulation for combined quantitative assessment of myocardial perfusion and regional LV function in an animal model of acute ischemia.

METHODS AND RESULTS

Coronary balloon occlusions were performed in 18 anesthetized pigs. Transthoracic power modulation images (Agilent 5500) were obtained during continuous intravenous infusion of the contrast agent Definity (DuPont) at baseline and during brief coronary occlusion and reperfusion and were analyzed with custom software. At each phase, myocardial perfusion was assessed by calculation, in 6 myocardial regions of interest, of mean pixel intensity and the rate of contrast replenishment after high-power ultrasound impulses. LV function was assessed by calculation of regional fractional area change from semiautomatically detected endocardial borders. All ischemic episodes caused detectable and reversible changes in perfusion and function. Perfusion defects, validated with fluorescent microspheres, were visualized in real time and confirmed by a significant decrease in pixel intensity in the left anterior descending coronary artery territory after balloon inflation and reduced rate of contrast replenishment. Fractional area change decreased significantly in ischemic segments and was restored with reperfusion.

CONCLUSIONS

Power modulation allows simultaneous online assessment of myocardial perfusion and regional LV wall motion, which may improve the echocardiographic diagnosis of myocardial ischemia.

Evaluation of respiratory influences on left ventricular function parameters extracted from echocardiographic acoustic quantification

This study was designed to assess, using the echocardiographic acoustic quantification technique, the influence of respiration on left ventricular (LV) function and its modifications connected with the ageing process, quantifying in a non-invasive way the respiratory contribution to the LV volume variability. An automated algorithm is applied to extract the beat-to-beat measurements of LV function parameters from the LV volume signal, obtained from recordings lasting a few minutes. Mean values, amount of variability and spectral content were studied in a population of 17 normal young (mean age 25 +/- 1 years) and 12 normal old (mean age 64 +/- 2 years) subjects. Mean values of the beat-to-beat measurements of LV function parameters were able to point out alterations connected with the ageing process in peak filling rate, peak atrial filling rate and peak ejection rate. Spectral analysis, applied to the extracted variability series, displayed a predominance of the high-frequency (HF) component corresponding to respiration in all LV function parameters; moreover, age related changes of HF variability were observed in peak ejection rate. The HF power spectrum component of beat to beat series extracted from the LV signal can provide a non-invasive assessment of the fluctuations in ventricular parameters associated with respiration.