Application of reinforcement learning for segmentation of transrectal ultrasound images

By Farhang Sahba, M. Bilgen, Hamid R. Tizhoosh, Magdy M. A. Salama. 2008

283 citations recorded by the source corpus.

BackgroundManganese-enhanced MRI (MEI) offers a novel neuroimaging modality to trace corticospinal tract (CST) in live animals. This paper expands this capability further and tests the utility of MEI to image axonal fiber connectivity in CST of injured spinal cord (SC).MethodsA rat was injured at the thoracic T4 level of the SC. The CST was labeled with manganese (Mn) injected intracortically at two weeks post injury. Next day, the injured SC was imaged using MEI and diffusion tensor imaging (DTI) modalities.ResultsIn vivo MEI data obtained from cervical SC confirmed that CST was successfully labeled with Mn. Ex vivo MEI data obtained from excised SC depicted Mn labeling of the CST in SC sections caudal to the lesion, which meant that Mn was transported through the injury, possibly mediated by viable CST fibers present at the injury site. Examining the ex vivo data from the injury epicenter closely revealed a thin strip of signal enhancement located ventrally between the dorsal horns. This enhancement was presumably associated with the Mn accumulation in these intact fibers projecting caudally as part of the CST. Additional measurements with DTI supported this view.ConclusionCombining these preliminary results collectively demonstrated the feasibility of imaging fiber connectivity in experimentally injured SC using MEI. This approach may play important role in future investigations aimed at understanding the neuroplasticity in experimental SCI research.

Subfields: Cognitive Neuroscience, Health, Toxicology and Mutagenesis, Radiology, Nuclear Medicine and Imaging

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

1. Shamim Nemati
2. Chao Yu
3. Jiming Liu

In Radiology, Nuclear Medicine and Imaging

1. Shamim Nemati
2. Chao Yu
3. Jiming Liu

Citations in the corpus

• Reinforcement Learning in Healthcare: A Survey (2019, background): “Sahba et al. [196–199] applied basic Q-learning to the segmentation of prostate in transrectal ultrasound images (UI) .”