ﺗﺸﺨﯿﺺ آرﯾﺘﻤﯽ ﻫﺎي ﻗﻠﺒﯽ ﺑﺮاﺳﺎس ﺗﺒﺪﯾﻞ ﺑﺴﺘﻪ ﻣﻮﺟﮏ و اﻟﮕﻮرﯾﺘﻢ ﻓﺎﮐﺘﻮرﮔﯿﺮي ﻣﺎﺗﺮﯾﺲ ﻏﯿﺮ ﻣﻨﻔﯽ ﺗُﻨُﮏ

ECG Arrhythmia Classification Based on Wavelet Packet Transform and Sparse Non-Negative Matrix Factorization

ﭼﮑﯿﺪه: ﺑﯿﻤﺎري ﻫﺎي ﻗﻠﺒﯽ ﯾﮑﯽ از ﺷﺎﯾﻊ ﺗﺮﯾﻦ ﻋﻮاﻣﻞ ﻣﺮگ و ﻣﯿﺮ در ﻣﺤﺪوده ﻫﺎي ﺳﻨﯽ ﻣﺨﺘﻠﻒ ﻣﯽ ﺑﺎﺷﺪ و ﺗﻌﯿﯿﻦ دﻗﯿﻖ ﻧﻮع آرﯾﺘﻤﯽ ﺑﺮاﺳﺎس ﭘﺮدازش ﺳﯿﮕﻨﺎل ﻫﺎي ﻗﻠﺒﯽ ﻣﯽ ﺗﻮاﻧﺪ در ﮐﻨﺎر داﻧﺶ ﭘﺰﺷﮑﯽ ﺑﻪ ﺗﺼﻤﯿﻢ ﮔﯿﺮي درﺳﺖ در ﻣﻮرد وﺿﻌﯿﺖ ﺑﯿﻤﺎر ﻣﻨﺘﻬﯽ ﮔﺮدد. در اﯾﻦ زﻣﯿﻨﻪ ﺗﺸﺨﯿﺺ ﻧﻮع آرﯾﺘﻤﯽ و اﻧﺘﺨﺎب ﺷﯿﻮه درﻣﺎﻧﯽ ﻣﻨﺎﺳﺐ ﺑﺮ اﺳﺎس آن ﻣﯽ ﺗﻮاﻧﺪ ﺑﻪ ﯾﮏ ﻣﺴﺌﻠﻪ ﭼﺎﻟﺶ ﺑﺮاﻧﮕﯿﺰ ﺗﺒﺪﯾﻞ ﮔﺮدد زﯾﺮا اﻣﮑﺎن ﺑﺮوز ﺧﻄﺎ در اﯾﻦ ﺗﺼﻤﯿﻢ ﮔﯿﺮي ﺗﻮﺳﻂ ﭘﺰﺷﮏ وﺟﻮد دارد. ﺑﻪ ﻣﻨﻈﻮر ﺑﺮرﺳﯽ دﻗﯿﻖ ﺟﺰﺋﯿﺎت ﺳﯿﮕﻨﺎل ﻗﻠﺒﯽ ﺛﺒﺖ ﺷﺪه از ﺑﯿﻤﺎر، ﺑﮑﺎرﮔﯿﺮي ﺗﮑﻨﯿﮏ ﻫﺎي ﭘﺮدازش و ﺗﺤﻠﯿﻞ ﺳﯿﮕﻨﺎل ﻣﯽ ﺗﻮاﻧﺪ اﻫﻤﯿﺖ ﺑﺴﯿﺎري داﺷﺘﻪ ﺑﺎﺷﺪ. در اﯾﻦ ﻣﻘﺎﻟﻪ، ﺗﺸﺨﯿﺺ ﻧﻮع آرﯾﺘﻤﯽ ﺑﻪ ﮐﻤﮏ ﺗﺮﮐﯿﺐ وﯾﮋﮔﯽ ﻫﺎي ﻣﻮرﻓﻮﻟﻮژﯾﮑﯽ و ﺿﺮاﯾﺐ ﺗﺒﺪﯾﻞ ﺑﺴﺘﻪ ﻣﻮﺟﮏ ﺻﻮرت ﻣﯽ ﮔﯿﺮد. ﺑﻪ ﻣﻨﻈﻮر ﮐﺎﻫﺶ ﺑﻌﺪ اﯾﻦ دﺳﺘﻪ وﯾﮋﮔﯽ ﻫﺎ از اﻟﮕﻮرﯾﺘﻢ ﺗﺤﻠﯿﻞ ﻣﻮﻟﻔﻪ ﻫﺎي اﺳﺎﺳﯽ ﺗُﻨُﮏ ﺳﺎﺧﺘﺎر ﯾﺎﻓﺘﻪ اﺳﺘﻔﺎده ﻣﯽ ﺷﻮد. ﺳﭙﺲ از اﯾﻦ ﺑﺮدار وﯾﮋﮔﯽ ﺑﻪ ﻣﻨﻈﻮر ﯾﺎدﮔﯿﺮي ﻣﺪل ﻫﺎي ﺑﺎزﻧﻤﺎﯾﯽ ﮐﻨﻨﺪه ﺳﺎﺧﺘﺎر داده ﻣﺮﺑﻮط ﺑﻪ ﻫﺮ ﻧﻮع آرﯾﺘﻤﯽ ﻗﻠﺒﯽ ﺑﻪ ﮐﻤﮏ اﻟﮕﻮرﯾﺘﻢ ﻓﺎﮐﺘﻮرﮔﯿﺮي ﻣﺎﺗﺮﯾﺲ ﻏﯿﺮﻣﻨﻔﯽ ﺗُﻨُﮏ اﺳﺘﻔﺎده ﻣﯽ ﮔﺮدد. دﺳﺘﻪ ﺑﻨﺪي داده ﻫﺎ در اﯾﻦ روش ﺑﺮاﺳﺎس ﻣﻘﺪار اﻧﺮژي ﺿﺮاﯾﺐ ﺗُﻨُﮏ ﺣﺎﺻﻞ از ﺑﺎزﻧﻤﺎﯾﯽ داده ﺻﻮرت ﻣﯽ ﮔﯿﺮد. ﻧﺘﺎﯾﺞ روش ﭘﯿﺸﻨﻬﺎدي ﺑﺎ ﻧﺘﺎﯾﺞ ﺣﺎﺻﻞ از ﺳﺎﯾﺮ روش ﻫﺎي ﻣﻄﺮح در اﯾﻦ ﺣﻮزه و ﻧﯿﺰ ﺳﺎﯾﺮ ﻃﺒﻘﻪ ﺑﻨﺪﻫﺎي ﻣﺒﺘﻨﯽ ﺑﺮ ﺷﺒﮑﻪ ﻋﺼﺒﯽ و ﻣﺎﺷﯿﻦ ﺑﺮدار ﭘﺸﺘﯿﺒﺎن ﻣﻘﺎﯾﺴﻪ ﺷﺪه اﺳﺖ. ﻧﺘﺎﯾﺞ ﺣﺎﺻﻞ از اﯾﻦ ﺷﺒﯿﻪ ﺳﺎزي ﻫﺎ ﻧﺸﺎن ﻣﯽ دﻫﺪ ﮐﻪ روش ﭘﯿﺸﻨﻬﺎدي ﻣﺒﺘﻨﯽ ﺑﺮ وﯾﮋﮔﯽ ﻫﺎي ﺗﺮﮐﯿﺒﯽ ﻣﻌﺮﻓﯽ ﺷﺪه و ﻣﺪل ﻫﺎي آﻣﻮزش دﯾﺪه ﻗﺎدر ﺑﻪ دﺳﺘﻪ ﺑﻨﺪي اﻧﻮاع آرﯾﺘﻤﯽ ﻗﻠﺒﯽ ﺑﺎ دﻗﺖ ﺑﺎﻻ ﺧﻮاﻫﺪ ﺑﻮد.

Classification of brain tumors using MRI images along with medical knowledge can lead to proper decision-making on the patient's condition. Also, classification of benign or malignant tumors is one of the challenging issues due to the need for detailed analysis of tumor tissue. Therefore, addressing this field using image processing techniques can be very important. In this paper, various types of texture-based and statistical-based features are used to determine the type of brain tumor and different types of features are applied in this classification procedure. Sparse coding and dictionary learning techniques are used to learn the over-complete models based on the characteristics of each data category. The classification process is carried out based on the calculated energy of the sparse coefficients. Also, the results of this categorization are compared with the results of the classification based on the neural network and support vector machine. The simulation results show that the proposed method based on the selected combinational features and learning the over-complete dictionaries can be able to classify the types of brain tumors precisely.