Fang-Cheng Yeh, MD, PhD

  • Assistant Professor
  • Director, High-Definition Fiber Tractography Lab

Fang-Cheng (Frank) Yeh, MD, PhD, joined the Department of Neurological Surgery in 2016 as a tenure-track assistant professor.

Prior to joining the faculty at the University of Pittsburgh, Dr. Yeh received his MD degree from National Taiwan University and completed his PhD study in biomedical engineering at Carnegie Mellon University in 2014.

Dr. Yeh is currently working on diffusion MRI and its role as image biomarkers for neurological and psychiatric disorders. His research focuses on novel applications of computational methods to brain connectome research, a challenging field with a lot of known unknowns and unsolved questions that require extensive technological development. He has developed several diffusion MRI methods and applied them to both clinical and translational studies.

Dr. Yeh is known for his development of DSI Studio, an integrated platform for diffusion MRI analysis, fiber tracking, and 3D tractography visualization. In 2018 alone, DSI Studio facilitated more than 100 peer-reviewed publications. DSI Studio provides the core technique for “high accuracy fiber tracking,” which has been widely used by many research groups to investigate how major fiber pathways are affected by neurological and psychiatric diseases. In an open competition sponsored by the International Society for Magnetic Resonance in Medicine (ISMRM) in 2015, Dr. Yeh’s method achieved the highest valid connection score (92.49%, ID:03) among 96 different approaches submitted by a total of 20 groups from around the world.

Specialized Areas of Interest

Diffusion MRI, tractography, network analysis, medical image analysis, pathology informatics.

Professional Organization Membership

International Society for Magnetic Resonance in Medicine

Education & Training

  • MD, National Taiwan University, 2006
  • PhD, Biomedical Engineering, Carnegie Mellon University, 2014

Honors & Awards

  • Chancellor’s Commercialization Fund Award, Pitt Ventures First Gear Program, University of Pittsburgh, 2019

Selected Publications

Yeh FC, Vettel JM, Singh A, Poczos B, Grafton ST, Erickson KI, Tseng WI, Verstynen TD. Quantifying Differences and Similarities in Whole-Brain White Matter Architecture Using Local Connectome FingerprintsPLoS Comput Biol 12(11):e1005203, 2016.

Yeh FC, Badre D, Verstynen T. Connectometry: A statistical approach harnessing the analytical potential of the local connectome. Neuroimage 125:162-71, 2016.

Fernández-Miranda JC, Wang Y, Pathak S, Stefaneau L, Verstynen T, Yeh FC. Asymmetry, connectivity, and segmentation of the arcuate fascicle in the human brain. Brain Struct Funct 220(3):1665-80, 2014.

Wang Y, Fernández-Miranda JC, Verstynen T, Pathak S, Schneider W, Yeh FC. Rethinking the role of the middle longitudinal fascicle in language and auditory pathways. Cereb Cortex 23(10):2347-56, 2013.

Yeh FC, Verstynen TD, Wang Y, Fernández-Miranda JC, Tseng WY. Deterministic diffusion fiber tracking improved by quantitative anisotropy. PLoS One 8(11):e80713, 2013.

Yeh FC, Tseng WY. NTU-90: a high angular resolution brain atlas constructed by q-space diffeomorphic reconstruction. Neuroimage 58(1):91-9, 2011.

Yeh FC, Wedeen VJ, Tseng WY. (2010). Generalized q-sampling imaging. IEEE Trans Med Imaging 29(9):1626-35, 2010.

Yeh FC, Parwani AV, Pantanowitz L, Ho C. Automated grading of renal cell carcinoma using whole slide imaging. Journal of Pathology Informatics 5:23, 2014.

Yeh FC, Ye Q, Hitchens TK, Wu YL, Parwani AV, Ho C. Mapping stain distribution in pathology slides using whole slide imaging. Journal of Pathology Informatics 5:1, 2014.

A complete list of Dr. Yeh's publications can be reviewed through the National Library of Medicine's publication database.

Research Activities

• Differential tractography as a track-based biomarker for neuronal injury

Diffusion MRI tractography has been used to map the axonal structure of the human brain, but its ability to detect neuronal injury is yet to be explored. Dr. Yeh has reported differential tractography, a new type of tractography that utilizes repeat MRI scans and a novel tracking strategy to map the exact segment of fiber pathways with a neuronal injury. He has examined differential tractography on multiple sclerosis, Huntington’s disease, amyotrophic lateral sclerosis, and epileptic patients. The results showed that the affected pathways shown by differential tractography matched well with the unique clinical symptoms of the patients, and the false discovery rate of the findings could be estimated using a sham setting to provide a reliability measurement. This novel approach enables a quantitative and objective method to monitor neuronal injury in individuals, allowing for diagnostic and prognostic evaluation of brain diseases.

• Structural Analysis and Optimization of Convolutional Neural Networks with a Small Sample Size

Deep neural networks have gained immense popularity in the Big Data problem; however, the availability of training samples can be relatively limited in specific application domains, particularly medical imaging, and consequently leading to overfitting problems. This “Small Data” challenge may need a mindset that is entirely different from the existing Big Data paradigm. Here, under the small data scenarios, Dr. Yeh has examined whether the network structure has a substantial influence on the performance and whether the optimal structure is predominantly determined by sample size or data nature. To this end, he has listed all possible combinations of layers given an upper bound of the VC-dimension to study how structural hyperparameters affected the performance. Results showed that structural optimization improved accuracy by 27.99%, 16.44%, and 13.11% over random selection for a sample size of 100, 500, and 1,000 in the MNIST dataset, respectively, suggesting that the importance of the network structure increases as the sample size becomes smaller. Furthermore, the optimal network structure was mostly determined by the data nature (photographic, calligraphic, or medical images), and less affected by the sample size, suggesting that the optimal network structure is data-driven, not sample size driven. 

Media Appearances