Search Results

You are looking at 1 - 6 of 6 items for

  • Author or Editor: Jason Wright x
  • All content x
Clear All Modify Search
Free access

Jennifer L. Quon, Michelle Han, Lily H. Kim, Mary Ellen Koran, Leo C. Chen, Edward H. Lee, Jason Wright, Vijay Ramaswamy, Robert M. Lober, Michael D. Taylor, Gerald A. Grant, Samuel H. Cheshier, John R. W. Kestle, Michael S. B. Edwards, and Kristen W. Yeom

OBJECTIVE

Imaging evaluation of the cerebral ventricles is important for clinical decision-making in pediatric hydrocephalus. Although quantitative measurements of ventricular size, over time, can facilitate objective comparison, automated tools for calculating ventricular volume are not structured for clinical use. The authors aimed to develop a fully automated deep learning (DL) model for pediatric cerebral ventricle segmentation and volume calculation for widespread clinical implementation across multiple hospitals.

METHODS

The study cohort consisted of 200 children with obstructive hydrocephalus from four pediatric hospitals, along with 199 controls. Manual ventricle segmentation and volume calculation values served as “ground truth” data. An encoder-decoder convolutional neural network architecture, in which T2-weighted MR images were used as input, automatically delineated the ventricles and output volumetric measurements. On a held-out test set, segmentation accuracy was assessed using the Dice similarity coefficient (0 to 1) and volume calculation was assessed using linear regression. Model generalizability was evaluated on an external MRI data set from a fifth hospital. The DL model performance was compared against FreeSurfer research segmentation software.

RESULTS

Model segmentation performed with an overall Dice score of 0.901 (0.946 in hydrocephalus, 0.856 in controls). The model generalized to external MR images from a fifth pediatric hospital with a Dice score of 0.926. The model was more accurate than FreeSurfer, with faster operating times (1.48 seconds per scan).

CONCLUSIONS

The authors present a DL model for automatic ventricle segmentation and volume calculation that is more accurate and rapid than currently available methods. With near-immediate volumetric output and reliable performance across institutional scanner types, this model can be adapted to the real-time clinical evaluation of hydrocephalus and improve clinician workflow.

Free access

Anthony C. Wang, George M. Ibrahim, Andrew V. Poliakov, Page I. Wang, Aria Fallah, Gary W. Mathern, Robert T. Buckley, Kelly Collins, Alexander G. Weil, Hillary A. Shurtleff, Molly H. Warner, Francisco A. Perez, Dennis W. Shaw, Jason N. Wright, Russell P. Saneto, Edward J. Novotny, Amy Lee, Samuel R. Browd, and Jeffrey G. Ojemann

OBJECTIVE

The potential loss of motor function after cerebral hemispherectomy is a common cause of anguish for patients, their families, and their physicians. The deficits these patients face are individually unique, but as a whole they provide a framework to understand the mechanisms underlying cortical reorganization of motor function. This study investigated whether preoperative functional MRI (fMRI) and diffusion tensor imaging (DTI) could predict the postoperative preservation of hand motor function.

METHODS

Thirteen independent reviewers analyzed sensorimotor fMRI and colored fractional anisotropy (CoFA)–DTI maps in 25 patients undergoing functional hemispherectomy for treatment of intractable seizures. Pre- and postoperative gross hand motor function were categorized and correlated with fMRI and DTI findings, specifically, abnormally located motor activation on fMRI and corticospinal tract atrophy on DTI.

RESULTS

Normal sensorimotor cortical activation on preoperative fMRI was significantly associated with severe decline in postoperative motor function, demonstrating 92.9% sensitivity (95% CI 0.661–0.998) and 100% specificity (95% CI 0.715–1.00). Bilaterally robust, symmetric corticospinal tracts on CoFA-DTI maps were significantly associated with severe postoperative motor decline, demonstrating 85.7% sensitivity (95% CI 0.572–0.982) and 100% specificity (95% CI 0.715–1.00). Interpreting the fMR images, the reviewers achieved a Fleiss’ kappa coefficient (κ) for interrater agreement of κ = 0.69, indicating good agreement (p < 0.01). When interpreting the CoFA-DTI maps, the reviewers achieved κ = 0.64, again indicating good agreement (p < 0.01).

CONCLUSIONS

Functional hemispherectomy offers a high potential for seizure freedom without debilitating functional deficits in certain instances. Patients likely to retain preoperative motor function can be identified prior to hemispherectomy, where fMRI or DTI suggests that cortical reorganization of motor function has occurred prior to the operation.