A convolutional neural network-based system to detect malignant findings in FDG PET/CT examinations
Abstract Background As the number of PET/CT scanners increases and FDG PET/CT becomes a common imaging modality for oncology, the demands for automated detection systems on artificial intelligence (AI) to prevent human oversight and misdiagnosis are rapidly growing. We aimed to develop a convolutional neural network (CNN)-based system that can classify whole-body FDG PET as 1) benign, 2) malignant, or 3) equivocal.Methods This retrospective study investigated 3,485 sequential patients with malignant or suspected malignant disease, who underwent whole-body FDG PET/CT at our institute. All the cases were classified into the 3 categories by a nuclear medicine physician. A residual network (ResNet)-based CNN architecture was built for classifying patients into the 3 categories. This network was trained with PET images. Five-fold cross-validations were carried out to estimate the classification performance. In addition, we examined whether the CNN could determine the location of the malignant uptake, be it in the head-and-neck region, chest, abdomen, or pelvic region.Results There were 1,280 (37%), 1,450 (42%) and 755 (22%) patients classified as benign, malignant and equivocal, respectively. In patient-based analysis, the CNN predicted benign and malignant images with 99.4% and 99.4% accuracy, respectively. Furthermore, in region-based analysis, the prediction was correct with the probability of 97.3% (head-and-neck), 96.6% (chest), 92.8% (abdomen) and 99.6% (pelvic region), respectively.Conclusion The CNN-based system reliably classified FDG PET images into 3 categories, indicating that it would be helpful for physicians as a double-checking system to prevent oversight and misdiagnosis.