TY - GEN
T1 - A novel method for the conversion of scanned electrocardiogram (ECG) image to digital signal
AU - Lewis, Macline Crecsilla
AU - Maiya, Manjunatha
AU - Sampathila, Niranjana
PY - 2018/1/1
Y1 - 2018/1/1
N2 - Electrocardiogram (ECG) is the record of origin and propagation of electrical potential through cardiac muscles. It provides information about heart functioning. Generally, ECG is printed on thermal paper. The person having heart abnormalities will have to maintain all the records for the diagnosis purpose, which requires large storage space and is minimized by storing in the computer using scanner. The stored data is processed manually, which is time consuming. So an automatic algorithm that is developed does the conversion of the ECG image to digital signal. In order to convert the image, image processing methods like binarization, morphological techniques have been used. Usage of morphological skeletonization helps in converting the image to digital signal form by finding the skeleton of the ECG signal. The performance of the conversion algorithm is analyzed using root-mean-square error (RMSE), and it was found good. The average error found between the binarized image and the skeletonized image is nearly 7.5%.
AB - Electrocardiogram (ECG) is the record of origin and propagation of electrical potential through cardiac muscles. It provides information about heart functioning. Generally, ECG is printed on thermal paper. The person having heart abnormalities will have to maintain all the records for the diagnosis purpose, which requires large storage space and is minimized by storing in the computer using scanner. The stored data is processed manually, which is time consuming. So an automatic algorithm that is developed does the conversion of the ECG image to digital signal. In order to convert the image, image processing methods like binarization, morphological techniques have been used. Usage of morphological skeletonization helps in converting the image to digital signal form by finding the skeleton of the ECG signal. The performance of the conversion algorithm is analyzed using root-mean-square error (RMSE), and it was found good. The average error found between the binarized image and the skeletonized image is nearly 7.5%.
UR - http://www.scopus.com/inward/record.url?scp=85040231866&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85040231866&partnerID=8YFLogxK
U2 - 10.1007/978-981-10-5520-1_34
DO - 10.1007/978-981-10-5520-1_34
M3 - Conference contribution
AN - SCOPUS:85040231866
SN - 9789811055195
T3 - Advances in Intelligent Systems and Computing
SP - 363
EP - 373
BT - International Conference on Intelligent Computing and Applications - ICICA 2016
PB - Springer Verlag
T2 - 3rd International Conference on Intelligent Computing and Applications, ICICA 2016
Y2 - 21 December 2016 through 22 December 2016
ER -