Geri Dön

Yapay sinir ağları kullanılarak EKG verilerinin sıkıştırılması

Başlık çevirisi mevcut değil.

  1. Tez No: 55985
  2. Yazar: YÜCEL KOÇYİĞİT
  3. Danışmanlar: DOÇ.DR. MEHMET KORÜREK
  4. Tez Türü: Yüksek Lisans
  5. Konular: Biyomühendislik, Bioengineering
  6. Anahtar Kelimeler: Elektrokardiyografi, Yapay sinir ağları, Electrocardiography, Artificial neural networks
  7. Yıl: 1996
  8. Dil: Türkçe
  9. Üniversite: İstanbul Teknik Üniversitesi
  10. Enstitü: Fen Bilimleri Enstitüsü
  11. Ana Bilim Dalı: Belirtilmemiş.
  12. Bilim Dalı: Belirtilmemiş.
  13. Sayfa Sayısı: Belirtilmemiş.

Özet

Tanlan EKG per%|»riu=3M Prâ=X2.77S39 Wı -ll 1\^f^^^^ M il Hj i u^n--/"1^^ aaafl | *3 ll«i-#«Uwo4l^^ IBürünen uari sauisi=l«WI Fiğ. 2. Graphics of the orginal and reconstructed ECG from MIT/BIH tapes 100 for 10 hidden nodes, and graphics of difference (5 traıes) between them process. As above mentioned, learning coefficient (s) and momentum (a) must be selected to minimize the error in training process. Reconstructed processing for compressed ECG is provided with 10 input nodes (because of 10 hidden nodes in compression) ör 5 input nodes (because of 5 hidden nodes in compression) and 200 output nodes. Number of hidden nodes are selected to minimize the error along with s and a. For normal ECG, orginal and reconstructed ECG is shown in Fiğ 2. An advantage of mis technique is to compress large numbers of data, i. e. 10 cycles (samples) from ECG's are used for training and 100 cycles (samples) can be compressed. x

Özet (Çeviri)

Lry»^-s\8Efl&»^(5) A,0, =-*(«,/<»,)(6) where s is a small positive constant. By calculating the right hand side of (5) and (6), it follows that *,wfl=sS*°*W A,0y=*yrf(8) where /'(netjYtj -o^) (whenj belongs to theoutputlayer.) **=].(9) f'(net^dpkw^ (othenvise) k Note that k in the above summation represents every unit k whose output follows into unit j. in order to accelerate the computation, the momenttim teraıs are added on (7), (8), A rwfl (n +1) = s6pj ofi + <zA, wy, (n)(10) A,0, (» + !) = eSa + aAp9J (n)(l 1) where n represents the number of learning cycles, and a is a small positive value. in this thesis, using artificial neural network (ANN) compression and reconstructed of the some ECG's are done. These ECG's are from MTT/BIH database tapes 100, 107 and 109. Tapes 100, 107 and 109 are respectively normal, paced beat and left bundle branch block ECG's. Orginal data sampled at 360 Hz and this sampling rate is decreased to 200 Hz in this thesis. Compression processing which is run on computer with Pentium-75 microprocessor in Borland C is provided, as above mentioned, with 200 input nodes, 10 and 5 hidden nodes and 200 output nodes. At the same time, these values give compression ratio, i.e. 20:1 and 40:1. Therefore the utilized ANN structures are as 200:10:200 and 200:5:200. Generally, 20 cycles from ECG's are used in training ixTanlan EKG per%|»riu=3M Prâ=X2.77S39 Wı -ll 1\^f^^^^ M il Hj i u^n--/“1^^ aaafl | *3 ll«i-#«Uwo4l^^ IBürünen uari sauisi=l«WI Fiğ. 2. Graphics of the orginal and reconstructed ECG from MIT/BIH tapes 100 for 10 hidden nodes, and graphics of difference (5 traıes) between them process. As above mentioned, learning coefficient (s) and momentum (a) must be selected to minimize the error in training process. Reconstructed processing for compressed ECG is provided with 10 input nodes (because of 10 hidden nodes in compression) ör 5 input nodes (because of 5 hidden nodes in compression) and 200 output nodes. Number of hidden nodes are selected to minimize the error along with s and a. For normal ECG, orginal and reconstructed ECG is shown in Fiğ 2. An advantage of mis technique is to compress large numbers of data, i. e. 10 cycles (samples) from ECG's are used for training and 100 cycles (samples) can be compressed. xLry»^-s\8Efl&»^(5) A,0, =-*(«,/<»,)(6) where s is a small positive constant. By calculating the right hand side of (5) and (6), it follows that *,wfl=sS*°*W A,0y=*yrf(8) where /'(netjYtj -o^) (whenj belongs to theoutputlayer.) **=].(9) f'(net^dpkw^ (othenvise) k Note that k in the above summation represents every unit k whose output follows into unit j. in order to accelerate the computation, the momenttim teraıs are added on (7), (8), A rwfl (n +1) = s6pj ofi + <zA, wy, (n)(10) A,0, (» + !) = eSa + aAp9J (n)(l 1) where n represents the number of learning cycles, and a is a small positive value. in this thesis, using artificial neural network (ANN) compression and reconstructed of the some ECG's are done. These ECG's are from MTT/BIH database tapes 100, 107 and 109. Tapes 100, 107 and 109 are respectively normal, paced beat and left bundle branch block ECG's. Orginal data sampled at 360 Hz and this sampling rate is decreased to 200 Hz in this thesis. Compression processing which is run on computer with Pentium-75 microprocessor in Borland C is provided, as above mentioned, with 200 input nodes, 10 and 5 hidden nodes and 200 output nodes. At the same time, these values give compression ratio, i.e. 20:1 and 40:1. Therefore the utilized ANN structures are as 200:10:200 and 200:5:200. Generally, 20 cycles from ECG's are used in training ixTanlan EKG per%|»riu=3M Prâ=X2.77S39 Wı -ll 1\^f^^^^ M il Hj i u^n--/”1^^ aaafl | *3 ll«i-#«Uwo4l^^ IBürünen uari sauisi=l«WI Fiğ. 2. Graphics of the orginal and reconstructed ECG from MIT/BIH tapes 100 for 10 hidden nodes, and graphics of difference (5 traıes) between them process. As above mentioned, learning coefficient (s) and momentum (a) must be selected to minimize the error in training process. Reconstructed processing for compressed ECG is provided with 10 input nodes (because of 10 hidden nodes in compression) ör 5 input nodes (because of 5 hidden nodes in compression) and 200 output nodes. Number of hidden nodes are selected to minimize the error along with s and a. For normal ECG, orginal and reconstructed ECG is shown in Fiğ 2. An advantage of mis technique is to compress large numbers of data, i. e. 10 cycles (samples) from ECG's are used for training and 100 cycles (samples) can be compressed. xLry»^-s\8Efl&»^(5) A,0, =-*(«,/<»,)(6) where s is a small positive constant. By calculating the right hand side of (5) and (6), it follows that *,wfl=sS*°*W A,0y=*yrf(8) where /'(netjYtj -o^) (whenj belongs to theoutputlayer.) **=].(9) f'(net^dpkw^ (othenvise) k Note that k in the above summation represents every unit k whose output follows into unit j. in order to accelerate the computation, the momenttim teraıs are added on (7), (8), A rwfl (n +1) = s6pj ofi + <zA, wy, (n)(10) A,0, (» + !) = eSa + aAp9J (n)(l 1) where n represents the number of learning cycles, and a is a small positive value. in this thesis, using artificial neural network (ANN) compression and reconstructed of the some ECG's are done. These ECG's are from MTT/BIH database tapes 100, 107 and 109. Tapes 100, 107 and 109 are respectively normal, paced beat and left bundle branch block ECG's. Orginal data sampled at 360 Hz and this sampling rate is decreased to 200 Hz in this thesis. Compression processing which is run on computer with Pentium-75 microprocessor in Borland C is provided, as above mentioned, with 200 input nodes, 10 and 5 hidden nodes and 200 output nodes. At the same time, these values give compression ratio, i.e. 20:1 and 40:1. Therefore the utilized ANN structures are as 200:10:200 and 200:5:200. Generally, 20 cycles from ECG's are used in training ixTanlan EKG per%|»riu=3M Prâ=X2.77S39 Wı -ll 1\^f^^^^ M il Hj i u^n--/"1^^ aaafl | *3 ll«i-#«Uwo4l^^ IBürünen uari sauisi=l«WI Fiğ. 2. Graphics of the orginal and reconstructed ECG from MIT/BIH tapes 100 for 10 hidden nodes, and graphics of difference (5 traıes) between them process. As above mentioned, learning coefficient (s) and momentum (a) must be selected to minimize the error in training process. Reconstructed processing for compressed ECG is provided with 10 input nodes (because of 10 hidden nodes in compression) ör 5 input nodes (because of 5 hidden nodes in compression) and 200 output nodes. Number of hidden nodes are selected to minimize the error along with s and a. For normal ECG, orginal and reconstructed ECG is shown in Fiğ 2. An advantage of mis technique is to compress large numbers of data, i. e. 10 cycles (samples) from ECG's are used for training and 100 cycles (samples) can be compressed. x

Benzer Tezler

  1. Tez No

    748209

    Heart failure detection with deep networks based on electrocardiography and impedance cardiography signals

    Kalp yetersizliğinin elektrokardiyografi ve empedans kardiyografi işaretleri kullanılarak derin ağlara dayalı tespiti

    ŞEVKET GÜNDÜZ

    Yüksek Lisans

    İngilizce

    İngilizce

    2022

    KardiyolojiMarmara Üniversitesi

    Elektrik-Elektronik Mühendisliği Ana Bilim Dalı

    PROF. DR. CABİR VURAL

    PROF. DR. ALİ SERDAR FAK

  2. Tez No

    100763

    Yapay sinir ağları ve genetik algoritmalar kullanılarak EKG vurularının sınıflandırılması

    Classification of ECG beats by using artifical neural networks and genetic algorithms

    ZÜMRAY DOKUR

    Doktora

    Türkçe

    Türkçe

    2000

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    DOÇ.DR. TAMER ÖLMEZ

  3. Tez No

    848408

    Developing a novel artificial intelligence based method for diagnosing chronic obstructive pulmonary disease

    Kronik obstrüktif akciğer hastalığı teşhisi için yapay zeka tabanlı yeni bir yöntem geliştirilmesi

    İNANÇ MORAN

    Doktora

    İngilizce

    İngilizce

    2023

    Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrolİstanbul Teknik Üniversitesi

    Bilgisayar Mühendisliği Ana Bilim Dalı

    PROF. DR. DENİZ TURGAY ALTILAR

  4. Tez No

    266434

    Uyku apnesi türlerinin sınıflandırılması

    Classification of sleep apnea types

    MEHMET FEYZİ AKŞAHİN

    Doktora

    Türkçe

    Türkçe

    2010

    Elektrik ve Elektronik MühendisliğiBaşkent Üniversitesi

    Elektrik-Elektronik Mühendisliği Ana Bilim Dalı

    PROF. DR. OSMAN EROĞUL

  5. Tez No

    276763

    Erken ventriküler kasılmalarda ysa tabanlı bir sınıflandırıcının FPGA ile gerçekleştirilmesi

    An FPGA implementation of ann-based premature ventricular contraction classifier

    AHMET TURAN ÖZDEMİR

    Doktora

    Türkçe

    Türkçe

    2010

    Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve KontrolErciyes Üniversitesi

    Elektronik Mühendisliği Ana Bilim Dalı

    PROF. DR. KENAN DANIŞMAN

Geri Dön