Transfer learning from RF to B-mode temporal enhanced ultrasound features for prostate cancer detection

Abstract

Our result suggests that the proposed adaptation technique is successful in reducing the divergence between TeUS RF and B-mode data.

Going Deep in Medical Image Analysis: Concepts, Methods, Challenges, and Future Directions

“…Considering the obvious lack of large-scale annotated datasets, Medical Imaging community has already started exploiting 'transfer learning' [287], [288], [289]. In transfer learning, one can learn a complex model using data from a source domain where large-scale annotated images are available (e.g.…”

Toward a real-time system for temporal enhanced ultrasound-guided prostate biopsy

“…We give a brief overview of these methods. For a detailed description of the models, the reader may refer to [4,8]. To generate the TeUS-based cancer detection models, we obtained TeUS data during fusion prostate biopsy.…”

“…We also augment the training data by creating ROIs using a sliding window of size 0.5 mm × 0.5 mm over the target region, which results in 1536 ROIs per target. Given the data augmentation strategy, we obtain a total number of 129,024 training samples including both TeUS radio frequency (RF) and TeUS B-mode data (see [8] for more details.). We further use the test data consisting of 171 cores to evaluate the trained model during the guidance system implementation, where 130 cores are labeled as benign and 31 cores are labeled as cancerous with GS ≥ 3 + 3.…”

“…We aim to learn a mapping from x i to y i in using recurrent neural networks (RNNs) to model the temporal information in TeUS B-mode data. For this purpose, first, we used the unsupervised domain adaption method presented in [8] to find a common feature space between TeUS Radio Frequency (RF) and TeUS B-mode data. Our RNN architecture is built with long short-term memory (LSTM) cells [14], where each cell maintains a memory over time.…”

Deep learning in medical imaging and radiation therapy

“…This property, referred to as transfer learning, is not unique to DL, but the large training data requirements of DL make it particularly useful in cases where relevant data for a particular task are scarce. For instance, in medical imaging, a DL system can be trained on a large number of natural images or those in a different modality to learn proper feature representations that allow it to “see.” The pretrained system can subsequently use these representations to produce an encoding of a medical image that is used for classification . Systems using transfer learning often outperform the state‐of‐the‐art methods based on traditional handcrafted features that were developed over many years with a great deal of expertise.…”