DLLabelsCT: Annotation tool using deep transfer learning to assist in creating new datasets from abdominal computed tomography scans, case study: Pancreas

• Published in: PloS one Vol. 19; no. 12; p. e0313126
• Main Authors: Mustonen, Henrik, Isosalo, Antti, Nortunen, Minna, Nevalainen, Mika, Nieminen, Miika T., Huhta, Heikki
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 03.12.2024; Public Library of Science (PLoS)
• Subjects: Abdomen; Algorithms; Analysis; Annotations; Archives & records; Artificial intelligence; Artificial neural networks; Automation; Biology and Life Sciences; Computed tomography; Computer and Information Sciences; CT imaging; Data mining; Datasets; Deep Learning; Evaluation; Hospitals; Humans; Image analysis; Image processing; Image Processing, Computer-Assisted - methods; Machine learning; Medical imaging; Medical research; Medicine and Health Sciences; Neural networks; Neural Networks, Computer; Pancreas; Pancreas - diagnostic imaging; Pancreatic cancer; Research and Analysis Methods; Scanners; Tomography; Tomography, X-Ray Computed - methods; Transfer learning; Tumors; Finland; Netherlands; Chicago Illinois; United States > US; Amsterdam Netherlands; Japan; Germany
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0313126

The utilization of artificial intelligence (AI) is expanding significantly within medical research and, to some extent, in clinical practice. Deep learning (DL) applications, which use large convolutional neural networks (CNN), hold considerable potential, especially in optimizing radiological evaluations. However, training DL algorithms to clinical standards requires extensive datasets, and their processing is labor-intensive. In this study, we developed an annotation tool named DLLabelsCT that utilizes CNN models to accelerate the image analysis process. To validate DLLabelsCT, we trained a CNN model with a ResNet34 encoder and a UNet decoder to segment the pancreas on an open-access dataset and used the DL model to assist in annotating a local dataset, which was further used to refine the model. DLLabelsCT was also tested on two external testing datasets. The tool accelerates annotation by 3.4 times compared to a completely manual annotation method. Out of 3,715 CT scan slices in the testing datasets, 50% did not require editing when reviewing the segmentations made by the ResNet34-UNet model, and the mean and standard deviation of the Dice similarity coefficient was 0.82±0.24. DLLabelsCT is highly accurate and significantly saves time and resources. Furthermore, it can be easily modified to support other deep learning models for other organs, making it an efficient tool for future research involving larger datasets.