ABSTRACT
The objective of the present study was to investigate whether deep learning could be applied successfully to the classification of images from colposcopy. For this purpose, a total of 158 patients who underwent conization were enrolled, and medical records and data from the gynecological oncology database were retrospectively reviewed. Deep learning was performed with the Keras neural network and TensorFlow libraries. Using preoperative images from colposcopy as the input data and deep learning technology, the patients were classified into three groups [severe dysplasia, carcinoma in situ (CIS) and invasive cancer (IC)]. A total of 485 images were obtained for the analysis, of which 142 images were of severe dysplasia (2.9 images/patient), 257 were of CIS (3.3 images/patient), and 86 were of IC (4.1 images/patient). Of these, 233 images were captured with a green filter, and the remaining 252 were captured without a green filter. Following the application of L2 regularization, L1 regularization, dropout and data augmentation, the accuracy of the validation dataset was ~50%. Although the present study is preliminary, the results indicated that deep learning may be applied to classify colposcopy images.
ABSTRACT
PURPOSE: This study aimed at evaluating the applicability of the concept of platinum sensitivity to recurrent cervical cancer. METHODS: The clinical information of patients with recurrent cervical cancer, who were initially treated with platinum-based chemotherapy and received second-line platinum-based chemotherapy at the time of recurrence between January 2008 and December 2012, was retrospectively reviewed. RESULTS: A total of 677 patients from 71 medical centers were analyzed. The median overall survival (OS) for patients with platinum-free interval (PFI) of <6, 6-11, 12-17, and ≥18 months was 12.1 (95% CI 11.0-14.1) months, 17.4 (15.5-20.4) months, 20.2 (17.9-27.6) months, and 29.9 (26.7-36.0) months, respectively (P < 0.0001, log-rank). The best cut-off value of PFI that affected OS was 7 months, analyzed by the minimum P value method. The median progression-free survival (PFS) for patients with less than and more than PFI of 7 months was 6.2 months (95% CI 4.8-9.3) and 21.0 months (18.9-24.8) (P < 0.0001, log-rank), respectively, and the median OS for patients with less than and more than PFI of 7 months was 12.3 months (11.2-14.1) and 24.2 months (20.8-25.8) (P < 0.0001, log-rank). Multivariate analysis revealed that PFI (P < 0.0001, HR 0.449, 95% CI 0.369-0.548) alone had a statistically significant association with OS. CONCLUSIONS: This study showed that the concept of platinum sensitivity could be applied to recurrent cervical cancer and PFI could be one of the independent prognostic factors for patients with recurrent cervical cancer who have previously been treated with platinum-based chemotherapy.