Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network
Abstract Leukemia is a cancer of white blood cells characterized by immature lymphocytes. Due to blood cancer, many people die every year. Hence, the early detection of these blast cells is necessary for avoiding blood cancer. A novel deep convolutional neural network (CNN) 3SNet that has depth-wise...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-10-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-44210-7 |
| _version_ | 1797559705487278080 |
|---|---|
| author | D. P. Yadav Deepak Kumar Anand Singh Jalal Ankit Kumar Kamred Udham Singh Mohd Asif Shah |
| author_facet | D. P. Yadav Deepak Kumar Anand Singh Jalal Ankit Kumar Kamred Udham Singh Mohd Asif Shah |
| author_sort | D. P. Yadav |
| collection | DOAJ |
| description | Abstract Leukemia is a cancer of white blood cells characterized by immature lymphocytes. Due to blood cancer, many people die every year. Hence, the early detection of these blast cells is necessary for avoiding blood cancer. A novel deep convolutional neural network (CNN) 3SNet that has depth-wise convolution blocks to reduce the computation costs has been developed to aid the diagnosis of leukemia cells. The proposed method includes three inputs to the deep CNN model. These inputs are grayscale and their corresponding histogram of gradient (HOG) and local binary pattern (LBP) images. The HOG image finds the local shape, and the LBP image describes the leukaemia cell's texture pattern. The suggested model was trained and tested with images from the AML-Cytomorphology_LMU dataset. The mean average precision (MAP) for the cell with less than 100 images in the dataset was 84%, whereas for cells with more than 100 images in the dataset was 93.83%. In addition, the ROC curve area for these cells is more than 98%. This confirmed proposed model could be an adjunct tool to provide a second opinion to a doctor. |
| first_indexed | 2024-03-10T17:49:06Z |
| format | Article |
| id | doaj.art-ad3bd23a1c954ebf9e2a2345fe994cbd |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-03-10T17:49:06Z |
| publishDate | 2023-10-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-ad3bd23a1c954ebf9e2a2345fe994cbd2023-11-20T09:26:00ZengNature PortfolioScientific Reports2045-23222023-10-0113111510.1038/s41598-023-44210-7Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural networkD. P. Yadav0Deepak Kumar1Anand Singh Jalal2Ankit Kumar3Kamred Udham Singh4Mohd Asif Shah5Department of Computer Engineering and Applications, G.L.A. UniversityDepartment of Computer Science, NIT MeghalayaDepartment of Computer Engineering and Applications, G.L.A. UniversityDepartment of Computer Engineering and Applications, G.L.A. UniversitySchool of Computing, Graphic Era Hill UniversityKebri Dehar UniversityAbstract Leukemia is a cancer of white blood cells characterized by immature lymphocytes. Due to blood cancer, many people die every year. Hence, the early detection of these blast cells is necessary for avoiding blood cancer. A novel deep convolutional neural network (CNN) 3SNet that has depth-wise convolution blocks to reduce the computation costs has been developed to aid the diagnosis of leukemia cells. The proposed method includes three inputs to the deep CNN model. These inputs are grayscale and their corresponding histogram of gradient (HOG) and local binary pattern (LBP) images. The HOG image finds the local shape, and the LBP image describes the leukaemia cell's texture pattern. The suggested model was trained and tested with images from the AML-Cytomorphology_LMU dataset. The mean average precision (MAP) for the cell with less than 100 images in the dataset was 84%, whereas for cells with more than 100 images in the dataset was 93.83%. In addition, the ROC curve area for these cells is more than 98%. This confirmed proposed model could be an adjunct tool to provide a second opinion to a doctor.https://doi.org/10.1038/s41598-023-44210-7 |
| spellingShingle | D. P. Yadav Deepak Kumar Anand Singh Jalal Ankit Kumar Kamred Udham Singh Mohd Asif Shah Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network Scientific Reports |
| title | Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network |
| title_full | Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network |
| title_fullStr | Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network |
| title_full_unstemmed | Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network |
| title_short | Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network |
| title_sort | morphological diagnosis of hematologic malignancy using feature fusion based deep convolutional neural network |
| url | https://doi.org/10.1038/s41598-023-44210-7 |
| work_keys_str_mv | AT dpyadav morphologicaldiagnosisofhematologicmalignancyusingfeaturefusionbaseddeepconvolutionalneuralnetwork AT deepakkumar morphologicaldiagnosisofhematologicmalignancyusingfeaturefusionbaseddeepconvolutionalneuralnetwork AT anandsinghjalal morphologicaldiagnosisofhematologicmalignancyusingfeaturefusionbaseddeepconvolutionalneuralnetwork AT ankitkumar morphologicaldiagnosisofhematologicmalignancyusingfeaturefusionbaseddeepconvolutionalneuralnetwork AT kamredudhamsingh morphologicaldiagnosisofhematologicmalignancyusingfeaturefusionbaseddeepconvolutionalneuralnetwork AT mohdasifshah morphologicaldiagnosisofhematologicmalignancyusingfeaturefusionbaseddeepconvolutionalneuralnetwork |