Scalable deep text comprehension for Cancer surveillance on high-performance computing.

BACKGROUND: Deep Learning (DL) has advanced the state-of-the-art capabilities in bioinformatics applications which has resulted in trends of increasingly sophisticated and computationally demanding models trained by larger and larger data sets. This vastly increased computational demand challenges the feasibility of conducting cutting-edge research. One solution is to distribute the vast computational workload across multiple computing cluster nodes with data parallelism algorithms. In this study, we used a High-Performance Computing environment and implemented the Downpour Stochastic Gradient Descent algorithm for data parallelism to train a Convolutional Neural Network (CNN) for the natural language processing task of information extraction from a massive dataset of cancer pathology reports. We evaluated the scalability improvements using data parallelism training and the Titan supercomputer at Oak Ridge Leadership Computing Facility. To evaluate scalability, we used different numbers of worker nodes and performed a set of experiments comparing the effects of different training batch sizes and optimizer functions. RESULTS: We found that Adadelta would consistently converge at a lower validation loss, though requiring over twice as many training epochs as the fastest converging optimizer, RMSProp. The Adam optimizer consistently achieved a close 2nd place minimum validation loss significantly faster; using a batch size of 16 and 32 allowed the network to converge in only 4.5 training epochs. CONCLUSIONS: We demonstrated that the networked training process is scalable across multiple compute nodes communicating with message passing interface while achieving higher classification accuracy compared to a traditional machine learning algorithm.

Case report: intra-tendinous ganglion of the anterior cruciate ligament in a young footballer.

A 20-year-old male medical student and keen rugby player presented with a 12-month history of progressively worsening right knee pain and stiffness with no history of trauma. Clinical examination revealed effusion and posterior knee pain exacerbated by end range movement and an extension lag of 15 degrees. Physiotherapy to improve the range of motion proved unsuccessful. Magnetic resonance imaging showed that the ACL was grossly thickened and displaced by material reported as mucoid in nature. There were also areas of focally high signal in relation to its tibial attachment and intra osseous small cysts. Arthroscopic examination revealed a ganglion related to the tibial attachment of the ACL and gross thickening and discoloration of the ACL. Biopsies were taken showing foci of mucoid degeneration in the ACL. A large intra-ACL mass of brownish coloured tissue was excised arthroscopically. Already at 2 weeks follow up the patient had greatly improved range of movement and was pain free. However, upon returning to rugby, joint instability was noticed and a tear of the ACL was confirmed. This rare clinical condition can be diagnosed with MRI and arthroscopic debridement effectively relieves symptoms. This case report illustrates that augmentation or reconstruction may end up being the definitive treatment for athletes. It may also offer some support to the argument that mucoid degeneration and ganglion cyst formation share a similar pathogenesis to intra-osseous cyst formation.