WO2005006209A1 - Technique et appareil de detection de fibrillation auriculaire paroxystique - Google Patents

Technique et appareil de detection de fibrillation auriculaire paroxystique Download PDF

Info

Publication number
WO2005006209A1
WO2005006209A1 PCT/IE2004/000096 IE2004000096W WO2005006209A1 WO 2005006209 A1 WO2005006209 A1 WO 2005006209A1 IE 2004000096 W IE2004000096 W IE 2004000096W WO 2005006209 A1 WO2005006209 A1 WO 2005006209A1
Authority
WO
WIPO (PCT)
Prior art keywords
segment
paf
subject
ecg
cardiac
Prior art date
Application number
PCT/IE2004/000096
Other languages
English (en)
Inventor
Brian Hickey
Conor Heneghan
Original Assignee
University College Dublin National University Of Ireland, Dublin
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by University College Dublin National University Of Ireland, Dublin filed Critical University College Dublin National University Of Ireland, Dublin
Publication of WO2005006209A1 publication Critical patent/WO2005006209A1/fr

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
    • A61B5/346—Analysis of electrocardiograms
    • A61B5/349—Detecting specific parameters of the electrocardiograph cycle
    • A61B5/361—Detecting fibrillation
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
    • A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems

Definitions

  • This invention relates to a method and apparatus for detecting probable paroxysmal atrial fibrillation.
  • Atrial fibrillation is the most common heart arrhythmia in clinical practice, and has serious morbidity and mortality. AF is a significant risk factor for stroke; about 15% of strokes occur in people with AF. The prevalence of AF increases with age, with an estimated 10 % of the 80-89 age group suffering from AF. AF can either be chronic or intermittent; intermittent AF is referred to as paroxysmal atrial fibrillation (PAF).
  • PAF paroxysmal atrial fibrillation
  • Atrial fibrillation is currently diagnosed by searching for episodes of atrial fibrillation detected using a 12 lead electrocardiogram, an ambulatory Holter monitor (which records a single or two-channel ECG over 24-72 hours) or an event recorder (which records one or two channels of ECG when triggered by patient).
  • An episode of AF can be recognised by the presence of a low-amplitude variable signal prior to ventricular depolarisation and the absence of normal amplitude P waves. If an episode of AF occurs during the recording time the patient is diagnosed as suffering from atrial fibrillation, otherwise not. No attempt is currently made to diagnose atrial fibrillation in cases where no actual episode of AF is recorded.
  • US2002143266 discloses a technique for detecting episodes of atrial fibrillation based on analysis of R- intervals and beat types.
  • the technique uses a hidden Markov model as a classification technique, and also carries out detection of P waves to assist in classification.
  • WO0224068 describes a technique for using the R-R difference series to differentiate between sections of electrocardiogram where a subject is in normal rhythm and in atrial fibrillation.
  • US6178347 discusses another generic episode-based classification system which subtracts out the QRS complex, and then carries out spectral analysis of the residual signal in order to identify episodes of atrial fibrillation and atrial flutter. This idea is also known in the technical literature (for example, see Stridh M. and L. Sornmo.
  • the present invention differs by using spectral features and detection of atrial premature contractions (rather than approximate entropy) to derive information from the non-AF-episode ECG.
  • Previous work has also considered assessment of autonomic tone using heart rate variability to assess the likelihood of an AF episode being initiated (Huang J.L., Z.C. Wen, .L. Lee, M.S. Chang, S.A. Chen. Changes in autonomic tone before the onset of paroxysmal atrial fibrillation. International J of Cardiology, 66:275-283, 1998., Hickey B., C. Heneghan. Screening for paroxysmal atrial fibrillation using atrial premature contractions and spectral measures.
  • the present invention differs in that it combines information from atrial premature contractions together with spectral R-R measures, and also does not solely focus on predicting AF onset. It is an object of the invention to provide a method and apparatus for detecting probable paroxysmal atrial fibrillation (PAF) in a subject from a prolonged recording of an electrocardiogram (ECG) of that subject not necessarily containing an episode of AF.
  • PAF paroxysmal atrial fibrillation
  • the invention provides a method of detecting probable paroxysmal atrial fibrillation (PAF) in a subject from a prolonged recording of an electrocardiogram (ECG) of that subject not necessarily containing an episode of atrial fibrillation, the method comprising the following steps: (A) for each of a plurality of segments of the ECG: (i) determining cardiac intervals, (ii) calculating the correlation between the high and low frequency spectral content of the cardiac interval sequence at predetermined intervals during the ECG segment, and (iii) classifying the subject into a first or second group according to whether the degree of correlation is above or below a predetermined threshold, if the subject is in the first group: (iv) determining the power spectral density (PSD) of a cardiac interval sequence at a plurality of frequencies for at least a part of the segment, and (v) classifying the magnitudes of the power at each frequency to provide a probability of PAF, if t s s ⁇ bjed is in 1be second group: (
  • cardiac interval means a cardiac interbeat interval, such as the RR, PP, TT interval or other time interval between electrically equivalent points on the cardiac cycle, or a cardiac intrabeat interval such as the PR or QT interval.
  • a cardiac interbeat interval such as the RR, PP, TT interval or other time interval between electrically equivalent points on the cardiac cycle, or a cardiac intrabeat interval such as the PR or QT interval.
  • This invention ignores the conventional approach of diagnosing paroxysmal atrial fibrillation. Conventional methods depend heavily on whether episodes of AF occur during the ECG recording time, whereas the current invention does not. If no episode occurs during the recording time, the test fails using conventional methods. However, with this invention, the test can still succeed. Furthermore, this invention can be based on a single lead ECG, although not limited thereto, unlike most current techniques that use multiple leads.
  • Step 1 ECG Acquisition
  • ECG electrocardiogram
  • the raw ECG signal acquired as above was processed using a linear phase band pass FIR filter windowed using a Kaiser-Bessel window to reduce baseline wander and EMG artefact.
  • the band stop frequencies were set at 0.25 Hz and 20 Hz.
  • QRS detection was implemented using a Hubert Transform method. From this set of R peaks, the RR intervals were determined by known techniques.
  • the ECG was now divided up into twenty contiguous 30-minute segments and the following two-stage classification system performed for each segment.
  • Steps 3 - 5 Classification Stage 1
  • the first stage of classification provides a measure of such changes.
  • the correlation between the high frequency (HF) and low frequency (LF) spectral content of the RR interval sequence was calculated at 2-minute intervals during the 30 minute segment.
  • Step 3 The HF and LF components were derived from the interval based power spectral density (PSD) estimates of non-overlapping two-minute RR interval sequences.
  • PSD power spectral density
  • the RR interval sequence contains both normal and ectopic beats, i.e., no attempt was made to restrict the analysis to NN intervals.
  • To obtain a zero-mean sequence the mean RR interval was subtracted from each segment. The sequence was zero padded to the nearest integer multiple of 98 exceeding the length of the sequence, and the Fast Fourier Transform (FFT) was taken of the entire sequence. The absolute values of the FFT coefficients were then squared to yield a periodogram estimate of the PSD.
  • FFT Fast Fourier Transform
  • Adjacent frequency bins were then combined to result in a 98 point PSD estimate (of which only bins 0-49 are relevant since 50-97 provide identical information as 1-48).
  • the magnitudes of these PSD bins from 0.04 to 0J5 cycles/interval were added to yield the LF component, and similarly the bins from 0J5 to 0.4 cycles/interval yielded the HF component (this is in line with the conventional frequency ranges used in defining LF and HF components).
  • Step 4 For the 30-minute segment, this process yielded two sets of 15 values, and the correlation coefficient between these sets of numbers was taken, where the correlation coefficient was calculated as follows:
  • x ( and y t are the LF and HF powers respectively, x and y are their mean values, and ⁇ x and ⁇ y are the standard deviations.
  • Step 5 From this analysis, two subgroups of records were developed - those records with a correlation coefficient above 0.75 (group A), and those records with a correlation coefficient below 0.75 (group B).
  • group A A physiological interpretation is that the subjects in Group A experience parallel changes in sympathetic and parasympathetic activation, whereas those in Group B experience some decoupling. This decoupling could be due to something as simple as a postural change, or may reflect an origin of unknown cause.
  • Steps 6 - 10 Classification stage 2
  • the PSD of the RR interval sequence was once again taken and calculated as above. Adjacent frequency bins were then combined to result in a 64 point PSD estimate, Step 7, of which only bins 0-32 are relevant since 33-63 provide identical information as 1-31. This analysis was carried out on the final 10 minutes of the 30 minute segment. The magnitudes of these PSD bins were used as features in a linear discriminant classifier (LDC), Step 7, to be described below.
  • LDC linear discriminant classifier
  • Atrial Premature Contractions APCs
  • PAF Atrial Premature Contractions
  • Step 9 the same analysis was carried out in Step 9 as was carried out in Step 6 for Group A subjects, but this time over the final 20 minutes of the 30-minute segment.
  • Step 10 the magnitudes of the PSD bins were used as features in a linear discriminant classifier.
  • a linear discriminant classifier (LDC), based on Fisher's rule, was used.
  • LDC linear discriminant classifier
  • the classifier outputs a set of numbers representing the probability estimate of each class, in response to a set of input features.
  • Linear discriminants partition the feature space into different classes using a set of hyper- planes. Optimisation of the model is achieved through direct calculation and is extremely fast relative to other models such as neural networks.
  • the common covariance matrix is defined as:
  • This formulation provides a mapping from discriminant value to posterior probabilities wliich tends to bias the probabilities towards the extremes (since linear discriminant classifiers generally tend to underestimate the class probabilities).
  • the final class assigned to x is the class with the highest posterior probability.
  • the above analysis outputs a probability PPAF for each 30-minute segment.
  • a 30- minute recording was classified as PAF in origin if P P A F was greater than or equal to 60%.
  • the ECG was classified as coming from a AF patient if:
  • the Physionet Normal Sinus Rhythm Database (NSRD).
  • the database consists of 17 records, each approximately 24 hours long, recorded using lead II. These patients have no significant arrhythmias, and include 5 men, aged 26 to 45, and 13 women aged 20 to 50.
  • the database consists often overnight records of approximately 8-hours duration from subjects with no known cardiac pathology, aged between 23 and 40 (modified N5). The signals were sampled at 128 Hz, with 12-bit resolution.
  • the MIT-BIH AF database (AFDB); This database includes 23 long-term ECG recordings of human subjects with atrial fibrillation (mostly paroxysmal). The individual recordings are each approximately 10 hours in duration, and contain two ECG signals each sampled at 250 Hz with 12-bit resolution. The number of episodes of AF varies from 0 to 39.
  • the classification system achieved 100% accuracy in separating 26 Non-PAF and 23 PAF patients.
  • Step 1 Apart from the recording of the initial ECG, Step 1, it will be recognised by the person skilled in the art that the entire method can be implemented in software on a programmable general computer, such as a PC. However, this does not rule out the method being implemented in special purpose devices in, for example, microcode. Further, while the ECG used in the above embodiment is taken externally of the subject, it is quite possible to obtain the ECG from implanted devices.
  • the individual segments subject to Steps 2 to 10 need not be 30 minutes ling, but they should preferably be at least 15 minutes long.
  • the segments need not be contiguous — they could overlap (slower processing but greater accuracy) or be non-contiguous (faster processing but reduced accuracy).
  • the RR intervals be 2 minutes in Step 3, but they are preferably at least 1 minute intervals.
  • the portion of each 30-minute segment used in Steps 6 and 9 could be the whole segment, but preferably in each case it is a significant portion of the segment at or near the end of the segment.
  • the number of frequencies used in Steps 6 and 9 need not be 64 but is preferably at least 16.
  • the RR intervals were determined for use in Step 3
  • a different cardiac interval could be used, for example, the RR, PP, TT interval or other time interval between electrically equivalent points on the cardiac cycle.
  • a cardiac intrabeat interval such as the PR or QT interval could be used.

Abstract

La présente invention concerne une technique permettant d'évaluer la probabilité qu'une personne souffre de fibrillation auriculaire paroxystique (PAF). Cette technique est fondée sur l'analyse de tronçons de 30 minutes d'électrocardiogramme (ECG) d'un patient souffrant de fibrillation auriculaire paroxystique (PAF), sans qu'il soit nécessaire que ces segments contiennent un épisode réel de fibrillation auriculaire (AF). Pour chaque segment, l'évaluation est fondée sur la sortie d'un dispositif de classification discriminant linéaire supervisé (LCD). Un des deux LCD est utilisé selon qu'il existe une forte corrélation entre le la puissance de spectre de basse fréquence et de haute fréquence dans la densité spectrale de puissance RR (ou dans un autre intervalle cardiaque) sur ce segment de 30 minutes. S'il existe une forte corrélation, le LCD utilise des caractéristiques spectrales calculées sur une fenêtre de dix minutes, dans le cas d'une faible corrélation, des caractéristiques spectrales et des contractions prématurées auriculaires sont utilisées comme caractéristiques. La classification globale du sujet est ainsi obtenue par combinaison de ces classifications sur des tronçons individuels.
PCT/IE2004/000096 2003-07-11 2004-07-12 Technique et appareil de detection de fibrillation auriculaire paroxystique WO2005006209A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
IE20030518 2003-07-11
IES2003/0518 2003-07-11

Publications (1)

Publication Number Publication Date
WO2005006209A1 true WO2005006209A1 (fr) 2005-01-20

Family

ID=34044141

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IE2004/000096 WO2005006209A1 (fr) 2003-07-11 2004-07-12 Technique et appareil de detection de fibrillation auriculaire paroxystique

Country Status (1)

Country Link
WO (1) WO2005006209A1 (fr)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2007043903A1 (fr) * 2005-10-14 2007-04-19 Medicalgorithmics Sp. Z O.O. Procede, dispositif et systeme pour l'analyse de signal d'electrocardiographie (ecg) a derivations limitees
US7308308B1 (en) 2004-09-16 2007-12-11 Pacesetter, Inc. Method to monitor progression of atrial fibrillation and to detect its susceptibility for termination
WO2008007236A2 (fr) * 2006-06-07 2008-01-17 Koninklijke Philips Electronics N.V. Détection de fibrillation auriculaire
US7813791B1 (en) 2007-08-20 2010-10-12 Pacesetter, Inc. Systems and methods for employing an FFT to distinguish R-waves from T-waves using an implantable medical device
US8818496B2 (en) 2005-10-14 2014-08-26 Medicalgorithmics Ltd. Systems for safe and remote outpatient ECG monitoring
CN108601540A (zh) * 2015-12-07 2018-09-28 智能解决方案技术公司 心房颤动检测系统及使用方法
CN110037687A (zh) * 2019-04-09 2019-07-23 上海数创医疗科技有限公司 基于改进卷积神经网络的室性早搏心跳定位方法和装置

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
HICKEY B ET AL INSTITUTE OF ELECTRICAL AND ELECTRONICS ENGINEERS: "Screening for paroxysmal atrial fibrillation using atrial premature contractions and spectral measures", COMPUTERS IN CARDIOLOGY 2002. MEMPHIS, TN, SEPT. 22 - 25, 2002, NEW YORK, NY : IEEE, US, vol. VOL. 29, 22 September 2002 (2002-09-22), pages 217 - 220, XP010624126, ISBN: 0-7803-7735-4 *
HICKEY B; HENEGHAN C; DE CHAZAL P: "Non-episode-dependent assessment of paroxysmal atrial fibrillation through measurement of RR interval dynamics and atrial premature contractions", ANNALS OF BIOMEDICAL ENGINEERING, vol. 32, no. 5, May 2004 (2004-05-01), pages 677 - 687, XP008035521 *
HUANG J L ET AL: "Changes of autonomic tone before the onset of paroxysmal atrial fibrillation", INTERNATIONAL JOURNAL OF CARDIOLOGY. 30 OCT 1998, vol. 66, no. 3, 30 October 1998 (1998-10-30), pages 275 - 283, XP002296812, ISSN: 0167-5273 *
ZONG W ET AL INSTITUTE OF ELECTRICAL AND ELECTRONICS ENGINEERS: "A methodology for predicting paroxysmal atrial fibrillation based on ECG arrhythmia feature analysis", COMPUTERS IN CARDIOLOGY 2001. ROTTERDAM, THE NETHERLANDS, SEPT. 23 - 26, 2001, NEW YORK, NY : IEEE, US, 23 September 2001 (2001-09-23), pages 125 - 128, XP010573232, ISBN: 0-7803-7266-2 *

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7308308B1 (en) 2004-09-16 2007-12-11 Pacesetter, Inc. Method to monitor progression of atrial fibrillation and to detect its susceptibility for termination
US9846764B2 (en) 2005-10-14 2017-12-19 Medicalgorithmics S.A. Systems for safe and remote outpatient ECG monitoring
US11183305B2 (en) 2005-10-14 2021-11-23 Medicalgorithmics S.A. Systems for safe and remote outpatient ECG monitoring
WO2007043903A1 (fr) * 2005-10-14 2007-04-19 Medicalgorithmics Sp. Z O.O. Procede, dispositif et systeme pour l'analyse de signal d'electrocardiographie (ecg) a derivations limitees
US7753856B2 (en) 2005-10-14 2010-07-13 Medicalgorithmics Ltd. Method, device and system for cardio-acoustic signal analysis
US10262111B2 (en) 2005-10-14 2019-04-16 Medicalgorithmics S.A. Systems for safe and remote outpatient ECG monitoring
US8818496B2 (en) 2005-10-14 2014-08-26 Medicalgorithmics Ltd. Systems for safe and remote outpatient ECG monitoring
US9351652B2 (en) 2005-10-14 2016-05-31 Medicalgorithmics S.A. Systems for safe and remote outpatient ECG monitoring
WO2008007236A3 (fr) * 2006-06-07 2008-04-24 Koninkl Philips Electronics Nv Détection de fibrillation auriculaire
WO2008007236A2 (fr) * 2006-06-07 2008-01-17 Koninklijke Philips Electronics N.V. Détection de fibrillation auriculaire
US7813791B1 (en) 2007-08-20 2010-10-12 Pacesetter, Inc. Systems and methods for employing an FFT to distinguish R-waves from T-waves using an implantable medical device
CN108601540A (zh) * 2015-12-07 2018-09-28 智能解决方案技术公司 心房颤动检测系统及使用方法
CN110037687A (zh) * 2019-04-09 2019-07-23 上海数创医疗科技有限公司 基于改进卷积神经网络的室性早搏心跳定位方法和装置

Similar Documents

Publication Publication Date Title
Asgari et al. Automatic detection of atrial fibrillation using stationary wavelet transform and support vector machine
US8755876B2 (en) Methods and systems for atrial fibrillation detection
Lin et al. Heartbeat classification using normalized RR intervals and wavelet features
De Chazal et al. Using wavelet coefficients for the classification of the electrocardiogram
Costin et al. Atrial fibrillation onset prediction using variability of ECG signals
De Chazal Detection of supraventricular and ventricular ectopic beats using a single lead ECG
KR20130085632A (ko) 적응적 국부 임계화를 이용한 심전도 신호의 ｒ 피크 검출방법
Wang A new method for evaluating ECG signal quality for multi-lead arrhythmia analysis
Riasi et al. Prediction of ventricular tachycardia using morphological features of ECG signal
KR102559333B1 (ko) 수면호흡장애 환자의 인공지능 기반 심혈관질환 예측 모델을 이용한 심혈관질환 예측시스템
WO2005006209A1 (fr) Technique et appareil de detection de fibrillation auriculaire paroxystique
Alcaraz et al. Wavelet bidomain sample entropy analysis to predict spontaneous termination of atrial fibrillation
Hickey et al. Non-episode-dependent assessment of paroxysmal atrial fibrillation through measurement of RR interval dynamics and atrial premature contractions
Hickey et al. Screening for paroxysmal atrial fibrillation using atrial premature contractions and spectral measures
Weng et al. Atrial fibrillation detection using stationary wavelet transform analysis
Don et al. Analysis of electrocardiogram signals of arrhythmia and ischemia using fractal and statistical features
Wrobel et al. Optimizing the automated detection of atrial fibrillation episodes in long-term recording instrumentation
Khan et al. Heart rate variability analysis using approximate entropy and detrended fluctuation for monitoring heart condition
Ittatirut et al. Detection of Premature Ventricular Contraction for real-time applications
IE20040470U1 (en) Method and apparatus for detecting paroxysmal atrial fibrillation
Shouldice et al. Automated detection of paroxysmal atrial fibrillation from inter-heartbeat intervals
Hadia et al. Morphology-based detection of premature ventricular contractions
Johura et al. ECG signal for artrial fibrillation detection
Elgendi et al. Premature atrial complexes detection using the Fisher Linear Discriminant
Lin et al. Bispectrum and histogram features for the identification of atrial fibrillation based on electrocardiogram

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NA NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): BW GH GM KE LS MW MZ NA SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
122 Ep: pct application non-entry in european phase