Classification of scanning electron microscope images of pharmaceutical excipients using deep convolutional neural networks with transfer learning
Convolutional Neural Networks (CNNs) are image analysis techniques that have been applied to image classification in various fields. In this study, we applied a CNN to classify scanning electron microscopy (SEM) images of pharmaceutical raw material powders to determine if a CNN can evaluate particl...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2022-12-01
|
| Series: | International Journal of Pharmaceutics: X |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S259015672200024X |
| _version_ | 1797989789275783168 |
|---|---|
| author | Hiroaki Iwata Yoshihiro Hayashi Aki Hasegawa Kei Terayama Yasushi Okuno |
| author_facet | Hiroaki Iwata Yoshihiro Hayashi Aki Hasegawa Kei Terayama Yasushi Okuno |
| author_sort | Hiroaki Iwata |
| collection | DOAJ |
| description | Convolutional Neural Networks (CNNs) are image analysis techniques that have been applied to image classification in various fields. In this study, we applied a CNN to classify scanning electron microscopy (SEM) images of pharmaceutical raw material powders to determine if a CNN can evaluate particle morphology. We tested 10 pharmaceutical excipients with widely different particle morphologies. SEM images for each excipient were acquired and divided into training, validation, and test sets. Classification models were constructed by applying transfer learning to pretrained CNN models such as VGG16 and ResNet50. The results of a 5-fold cross-validation showed that the classification accuracy of the CNN model was sufficiently high using either pretrained model and that the type of excipient could be classified with high accuracy. The results suggest that the CNN model can detect differences in particle morphology, such as particle size, shape, and surface condition. By applying Grad-CAM to the constructed CNN model, we succeeded in finding particularly important regions in the particle image of the excipients. CNNs have been found to have the potential to be applied to the identification and characterization of raw material powders for pharmaceutical development. |
| first_indexed | 2024-04-11T08:25:00Z |
| format | Article |
| id | doaj.art-ba50632cc49e46f0beba35e478f9fa2b |
| institution | Directory Open Access Journal |
| issn | 2590-1567 |
| language | English |
| last_indexed | 2024-04-11T08:25:00Z |
| publishDate | 2022-12-01 |
| publisher | Elsevier |
| record_format | Article |
| series | International Journal of Pharmaceutics: X |
| spelling | doaj.art-ba50632cc49e46f0beba35e478f9fa2b2022-12-22T04:34:48ZengElsevierInternational Journal of Pharmaceutics: X2590-15672022-12-014100135Classification of scanning electron microscope images of pharmaceutical excipients using deep convolutional neural networks with transfer learningHiroaki Iwata0Yoshihiro Hayashi1Aki Hasegawa2Kei Terayama3Yasushi Okuno4Graduate School of Medicine, Kyoto University, 53 Shogoin-kawaharacho, Sakyo-ku, Kyoto 606-8507, JapanGraduate School of Medicine, Kyoto University, 53 Shogoin-kawaharacho, Sakyo-ku, Kyoto 606-8507, Japan; Pharmaceutical Technology Division, Nichi-Iko Pharmaceutical Co., Ltd., 205-1, Shimoumezawa Namerikawa-shi, Toyama 936-0857, Japan; Correspondence to: Y. Hayashi, Pharmaceutical Technology Division, Nichi-Iko Pharmaceutical Co., Ltd.; 205-1, Shimoumezawa Namerikawa-shi, Toyama 936-0857, Japan.Graduate School of Medicine, Kyoto University, 53 Shogoin-kawaharacho, Sakyo-ku, Kyoto 606-8507, JapanGraduate School of Medical Life Science, Yokohama City University, Tsurumi-ku, Yokohama 230-0045, JapanGraduate School of Medicine, Kyoto University, 53 Shogoin-kawaharacho, Sakyo-ku, Kyoto 606-8507, Japan; RIKEN Center for Computational Science, Kobe 650-0047, Japan; Correspondence to: Y. Okuno, Graduate School of Medicine, Kyoto University, 53 Shogoin-kawaharacho, Sakyo-ku, Kyoto 606-8507, Japan.Convolutional Neural Networks (CNNs) are image analysis techniques that have been applied to image classification in various fields. In this study, we applied a CNN to classify scanning electron microscopy (SEM) images of pharmaceutical raw material powders to determine if a CNN can evaluate particle morphology. We tested 10 pharmaceutical excipients with widely different particle morphologies. SEM images for each excipient were acquired and divided into training, validation, and test sets. Classification models were constructed by applying transfer learning to pretrained CNN models such as VGG16 and ResNet50. The results of a 5-fold cross-validation showed that the classification accuracy of the CNN model was sufficiently high using either pretrained model and that the type of excipient could be classified with high accuracy. The results suggest that the CNN model can detect differences in particle morphology, such as particle size, shape, and surface condition. By applying Grad-CAM to the constructed CNN model, we succeeded in finding particularly important regions in the particle image of the excipients. CNNs have been found to have the potential to be applied to the identification and characterization of raw material powders for pharmaceutical development.http://www.sciencedirect.com/science/article/pii/S259015672200024XConvolutional neural networksMachine learningScanning electron microscopeExcipientsPowderArtificial intelligence |
| spellingShingle | Hiroaki Iwata Yoshihiro Hayashi Aki Hasegawa Kei Terayama Yasushi Okuno Classification of scanning electron microscope images of pharmaceutical excipients using deep convolutional neural networks with transfer learning International Journal of Pharmaceutics: X Convolutional neural networks Machine learning Scanning electron microscope Excipients Powder Artificial intelligence |
| title | Classification of scanning electron microscope images of pharmaceutical excipients using deep convolutional neural networks with transfer learning |
| title_full | Classification of scanning electron microscope images of pharmaceutical excipients using deep convolutional neural networks with transfer learning |
| title_fullStr | Classification of scanning electron microscope images of pharmaceutical excipients using deep convolutional neural networks with transfer learning |
| title_full_unstemmed | Classification of scanning electron microscope images of pharmaceutical excipients using deep convolutional neural networks with transfer learning |
| title_short | Classification of scanning electron microscope images of pharmaceutical excipients using deep convolutional neural networks with transfer learning |
| title_sort | classification of scanning electron microscope images of pharmaceutical excipients using deep convolutional neural networks with transfer learning |
| topic | Convolutional neural networks Machine learning Scanning electron microscope Excipients Powder Artificial intelligence |
| url | http://www.sciencedirect.com/science/article/pii/S259015672200024X |
| work_keys_str_mv | AT hiroakiiwata classificationofscanningelectronmicroscopeimagesofpharmaceuticalexcipientsusingdeepconvolutionalneuralnetworkswithtransferlearning AT yoshihirohayashi classificationofscanningelectronmicroscopeimagesofpharmaceuticalexcipientsusingdeepconvolutionalneuralnetworkswithtransferlearning AT akihasegawa classificationofscanningelectronmicroscopeimagesofpharmaceuticalexcipientsusingdeepconvolutionalneuralnetworkswithtransferlearning AT keiterayama classificationofscanningelectronmicroscopeimagesofpharmaceuticalexcipientsusingdeepconvolutionalneuralnetworkswithtransferlearning AT yasushiokuno classificationofscanningelectronmicroscopeimagesofpharmaceuticalexcipientsusingdeepconvolutionalneuralnetworkswithtransferlearning |