Computed tomography-guided microwave ablation of perivascular liver metastases from colorectal cancer: a study of the ablation zone, feasibility, and safety.

OBJECTIVES: To compare the ablation margins and safety of microwave ablation (MWA) of perivascular versus non-perivascular liver metastases from colorectal cancer (CRC) and to determine the risk factors for local tumor progression (LTP) after perivascular MWA. METHODS: Between June 2017 and June 2019, 84 metastases were treated: 39 perivascular (<5 mm from a vessel >3 mm), and 46 non-perivascular. Perivascular metastases were treated with either conventional or optimized protocols (maximum power and/or several heating cycles after repositioning the needle regardless of the initial tumor dimensions). The mean diameter of metastases was 15.4 mm (SD: 7.56). RESULTS: Vascular proximity did not result in a significant difference in ablation margins. The technical success rate, primary efficacy, and secondary efficacy were 90%, 66%, and 83%, respectively. Perivascular location was not a risk factor for time to LTP (p = 0.49), RFS (p = 0.52), or OS (p = 0.54). LTP was statistically related to the presence of a colonic obstruction (p < 0.05), number of metastases at the time of diagnosis (p < 0.05), type of protocol (p < 0.05), ablation margins (p < 0.001) and LTP was proportional to the number of liver resections before MWA (p < 0.05). There was no LTP in tumors ablated with margins over 10 mm. Two grade 4 complications occurred. CONCLUSION: MWA is an effective and safe treatment for perivascular liver metastases from CRC, provided that satisfactory margins are achieved. A maximalist attitude could be related to better local control.

Automatic cervical lymphadenopathy segmentation from CT data using deep learning.

PURPOSE: The purpose of this study was to develop a fast and automatic algorithm to detect and segment lymphadenopathy from head and neck computed tomography (CT) examination. MATERIALS AND METHODS: An ensemble of three convolutional neural networks (CNNs) based on a U-Net architecture were trained to segment the lymphadenopathies in a fully supervised framework. The resulting predictions were assessed using the Dice similarity coefficient (DSC) on examinations presenting one or more adenopathies. On examinations without adenopathies, the score was given by the formula M/(M+A) where M was the mean adenopathy volume per patient and A the volume segmented by the algorithm. The networks were trained on 117 annotated CT acquisitions. RESULTS: The test set included 150 additional CT acquisitions unseen during the training. The performance on the test set yielded a mean score of 0.63. CONCLUSION: Despite limited available data and partial annotations, our CNN based approach achieved promising results in the task of cervical lymphadenopathy segmentation. It has the potential to bring precise quantification to the clinical workflow and to assist the clinician in the detection task.