Andere Kunden interessierten sich auch für
![Sensor de Fibra Óptica na Avaliação Cardiovascular:]( "Sensor de Fibra Óptica na Avaliação Cardiovascular:")
Sensor de Fibra Óptica na Avaliação Cardiovascular:26,99 €![Comparative Evaluation of Optical Coherence Tomography (OCT)]( "Comparative Evaluation of Optical Coherence Tomography (OCT)")
Comparative Evaluation of Optical Coherence Tomography (OCT)26,99 €![Decussatio pyramid and optical chiasm evolutionary process]( "Decussatio pyramid and optical chiasm evolutionary process")
Decussatio pyramid and optical chiasm evolutionary process40,99 €![Non-invasive assessment of central arterial pulses with optical fibres]( "Non-invasive assessment of central arterial pulses with optical fibres")
Non-invasive assessment of central arterial pulses with optical fibres42,99 €![CAD System for Breast Cancer Masses Detection]( "CAD System for Breast Cancer Masses Detection")
CAD System for Breast Cancer Masses Detection36,99 €![Retinal Nerve Fiber Layer (RNFL)Thickness In Black Glaucoma Suspects]( "Retinal Nerve Fiber Layer (RNFL)Thickness In Black Glaucoma Suspects")
Retinal Nerve Fiber Layer (RNFL)Thickness In Black Glaucoma Suspects32,99 €![Electrospun Fiber Membranes as Wound Dressing Materials]( "Electrospun Fiber Membranes as Wound Dressing Materials")
Electrospun Fiber Membranes as Wound Dressing Materials44,99 €
According to the World Health Organization (WHO) cardiovascular diseases (CVDs) are the prime cause of death in developed countries. Arterial pulse waveform (APW) analysis provides a non-invasive way for early diagnosis of CVDs. The present work reports the elaboration of an automatic algorithm to extract and classify the acquired data of APW signals based on the risk of CVD related problems. To extract the data several signal pre-processing methods were applied for noise reduction and pulse segmentation. Then, a K-Means clustering algorithm was applied to select the higher quality APWs. The discrepancy between a parallel manual extraction of the waves and the result of K-Means algorithm was practically null. A pool of 32 waveform characterization parameters were extracted including time domain, wavelet transform, root mean square error and frequency domain features. Some parameters were used in the K-Means clustering and the rest were used to train a multi-layer perceptron neural network for APW classification of low risk/high risk of CVDs in each patient. The results obtained are very promising. A portion of this work won the Nascimento Leitão prize.