56
Federico Lombardi, MD, FESC UOC Cardiologia, IRCCS Ospedale Maggiore Policlinico Università degli Studi di Milano. Email: [email protected] No conflict of interest in relation to this presentation

No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Federico Lombardi, MD, FESCUOC Cardiologia, IRCCS Ospedale Maggiore PoliclinicoUniversità degli Studi di Milano.Email: [email protected]

What is next? How to organize the best follow up of HF patients No conflict of interest in relation to this presentation

Page 2: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

•How new technologies and apps will affect follow up of HF patients:

•Physical activity•Quality of sleep•Recording of cardiac rhythm and detection of atrial fibrillation

Page 3: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

•New issues:regulation, validation, control, patients’ engagement, medical

certification vs over the counter sale, legal aspects…….

Page 4: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

4

From hospital to home to everywhere.Vital Parameters everywhere, everytime for long-term recordings → huge amount of data

• Challenges for physicians and BME:i) Innovative sensorsii) More sophisticated algorithms for

correct detection of biomarkers in an «uncontrolled» environment

iii) Compromise to be reached: on-board vs on-server processing

By courtesy of S. Cerruti

Page 5: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Is six min walking test over ?

Page 6: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Examples of wearable devices for the monitoring of physiological

parameters and activity status of the subject

Zheng, et al., 2014

Page 7: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Sarmento et al, Int J Cardiol 2018

Page 8: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Central apnea

Obstructive apnea

Page 9: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Wearable / textile sensors for vital signals acquisition

Smartex

CSEMResp (25 Hz, 1/beat)

BA

BR

ECG (250 Hz)

HR, HRV

By courtesy of S. Cerruti

Page 10: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Sleep Signal Acquisition

•Bed Foil (VTT)

Bed Sensor with 8 channel piezo foils

By courtesy of S. Cerruti

Page 11: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Devices for the acquisition of vital parameters at home

Sensorized

bracelets

Mobile EEG

device

Bed sensors

By courtesy of S. Cerruti

Page 12: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Devices which can be directly put in contact with the human body for the

monitoring of vital parameters: a) multifunction device for temp & stress; b)

capacity pressure sensor for radial pulse; c) e-skin (press & temp); d) respiration;

e) tattoo electrochemical biosensor for lactate measurements (sweat)

By courtesy of S. Cerruti

Page 13: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

From ILR to smartphone to clinician portal

Page 14: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

A new and emerging question: How to manage all these data?

Page 15: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Cadmus-Bertand et al Ann Int Med 2017

Over the counter: non medical device

Page 16: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 17: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 18: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Face Detection

Motion flow tracking

Contactless measurament of physiological parameter

Maximum peak (PPG)

Post Processing

Video BVP

cardiac interval

Dicrotic notch

Minimum peak (PPG)

Beat to beat variation of skin color

Corino, Iozzia & Lombardi submitted

Page 19: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 20: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 21: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Self-applied ECG patch improves the rate of AF diagnosisA Digital End-to-End, Nationwide, Pragmatic Trial of Screening for Undiagnosed Atrial

Fibrillation Within a Health Insurance System Using a Self-Applied ECG Patch: Primary Results of the mHealth Screening to Prevent Strokes (mSToPS) Trial

Page 22: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Methods: 2655 eligible individuals without a diagnosis of AF were enrolled via a web-based platform to undergo active monitoring at home with an iRhythm Zio patch for an average of 12 days (cases) and for each case, 2 age-, sex- and CHA2DS2-VASc-matched observational controls were chosen. The primary endpoint was the number of participants with first diagnosis of AF at 1 year.

Main results

1- year new diagnosis of AF was 6.3% in 1738 participants who were actively monitored compared to 2.3% in 3476 matched controls (unadj OR 2.8; 95%CI:2.1-3.7, P<0.0001 and adjOR 3.0; 95%CI:2.2-4.0, P<0.0001).

Median total AF burden during monitoring was 0.9% and median duration of longest AF period was 185.5 min (92.8% >5 minutes, 37.7% >6 hours).

Active monitoring was associated with increased initiation of anticoagulation therapy compared to routine care (5.4% vs. 3.4%, P=0.0004)

Steinhubl et al, ACC 2018

Page 23: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Wineinger et al, Heart Rhythm 2019

Page 24: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Wineinger et al, Heart Rhythm 2019

Page 25: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 26: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Cardiio Rhythm developed by CardiioCambridge MA.

AliveCor Kardia Application.

Page 27: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Kardia Mobile ecg recording

The present and……

Page 28: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

In 204 patients (48 in AF), compared to expert

12-lead ECG diagnosis, AliveCor lead I analyzed

by an embedded automated algorithm (iECG)

detected AF with 98% sensitivity, 97% specificity

and 97% accuracy [Lau JK et al, Int J Cardiol 2013].

Potential benefits of community AF screening in aging population

at risk with substantial impact on reducing stroke burden.

Smartphone as ECG monitor and AF detector

In 1000 pharmacy customers aged ≥65 years,

newly identified AF was found by iECG in 1.5%,

all with CHA2DS2-VASc score≥2, with 98.5%

sensitivity and 91.4% specificity.

[Lowres N et al, Thrombosis and Haemostasis 2014]

Use of smartphones as ECG monitor could allow

patients for rapid, real-world cardiac monitoring from

virtually anywhere, to confirm AF when symptomatic,

allowing for more timely treatment and management.

Page 29: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Curtis et al, PACE 2018 in press

Page 30: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Curtis et al, PACE 2018 in press

Page 31: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

The future: pilotless driving, parking support…….. and electrocardiographic monitoring

Page 32: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Conclusions:

• Will all these new devices and apps improve quality of life and clinical management of our patients?

Page 33: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 34: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

SLEEP STUDIES

Page 35: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Photograph of a complete electrocardiograph showing the manner in which the electrodes were attached to the patient; the hands and one foot were immersed in jars of salt solution.

Page 36: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

Evolution of the electrocardiogram from the electrometer. The upper record was made using the capillary electrometer, the middle record is a "corrected curve," and the lower record was made using Einthoven's string galvanometer. The exact source for the lower portion of this figure is unknown because it was not shown in the original figure published in 1903. It did appear in Fishman AP, Richards DW (eds): Circulation of the Blood: Men and Ideas. New York, Oxford University Press, 1964, p 295.

Page 37: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm

The present time: remote monitoring

Heart Rhythm Society 2015

Page 38: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 39: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 40: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 41: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 42: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 43: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 44: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 45: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 46: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 47: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 48: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 49: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 50: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 51: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 52: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 53: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 54: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 55: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm
Page 56: No conflict of interest in relation to this presentationIn 204 patients (48 in AF), compared to expert 12-lead ECG diagnosis, AliveCor lead I analyzed by an embedded automated algorithm