Ulas Ciftcioglu Email & Phone Number

Ankara, Turkey

San Francisco Bay Area

- Consulted with C. Light Technologies, to build a prognostic logistic regression model predicting whether disease level of a Multiple Sclerosis patient will worsen using fixational eye movement features, resulting in 59.4% balanced accuracy (https://bit.ly/forecastMS_post).- Implemented a forward feature selection method to overcome the curse of dimensionality, resulting in 10.7% improved balanced accuracy compared to using all features or dimensionality reduction by PCA.- Engineered new eye movement features based on measures of distributions from time series data to improve disease progression prediction, resulting in features that improved balanced accuracy by 2.2%.

Berkeley, California, United States

Greater Los Angeles Area

- Engineered a statistical testing method for time series data to quantify temporal stages of neuralprocessing, resulting in a twofold increase in the amount of usable data, vs previous standard methods.- Built a hybrid CNN and LSTM model using raw data and spectrograms of electrophysiological signals to extract running speed in time, resulting in predictions with a coefficient of determination (r^2) of 75%.- Filtered and preprocessed a public dataset (~100,000 brain cells), trained gradient boosting and DNN models to predict the brain region of a cell from its electrophysiological features, resulting in 94% accuracy; allowing researchers to determine the brain region they are recording to optimize experiments (please see Project section for more info, such as Github repo).

Greater Los Angeles Area

- Created a new metric from Fourier transform, built and selected analytical models to test how the brain processes visual information, discovering spatiotemporal features of vision coded in neural activity.- Built and trained a ridge regression model using cross validation to predict oscillatory activity in the brain, revealing that visual features encoded in neural oscillations are local rather than global.- Developed experimental setups and software to conduct experiments with state of the art high-density neural probes and acquisition systems, resulting in a fivefold increase in data collection.- Devised an index for a morphological feature of cells by analyzing 3D reconstructions and adapting the Thomson problem from particle physics, to quantify the symmetry in dendritic trees of brain cells.- Designed/performed animal experiments/surgeries, recorded electrophysiological data from 4 brain structures using multi‐channel (silicon probes) or single‐channel electrodes, used optogenetic stimulation. - As the president of USC Vision Science Association; organized and presented seminars, bringing together researchers in the field of visual neuroscience.- Thesis Title: Exploring Sensory Responses in the Different Subdivisions of the Visual Thalamus - Mentoring: 3 doctoral students, 1 technician, 1 undergraduate student- Principal Investigators: Judith Hirsch and Fritz Sommer

Ankara, Turkey

- Courses: Neuroimaging; Medical Image Analysis; Fundamentals of Medical Imaging; Pattern Classification for Biomedical Applications; Mathematics for Information Systems- Taught hands-on computational tools, created educational materials, oversaw assignments and projects for five graduate level courses (~100 students).

Ankara, Turkey

- Developed an evolutionary and region-merging image segmentation algorithm to detect gray matter tissue in MRI brain images, resulting in a 7% improvement in Jaccard index over a previous standard tool.- Analyzed time‐series functional MRI and functional optical neuroimaging data to investigate the activity of brain regions during behavioral tasks.- Tested the image segmentation algorithm with real and simulated image datasets with varying degree of noise and artifact levels. - Thesis Title: Improving the Subcortical Gray Matter Segmentation Using Evolutionary Hierarchical Region Merging- Principal Investigator: Didem Gokcay

Greater Philadelphia Area

- Worked on optical functional neuroimaging (theory, signal processing and experimentation) as part of successful efforts to implement functional Near‐Infrared Spectroscopy (fNIRS) technique in Middle East Technical University.

Ankara, Turkey

Worked on Electrocardiography (ECG) Instrumentation.

Manisa, Turkey

Worked on Television Electronic Hardware and Display Systems.

Doctor Of Philosophy - Phd, Neuroscience

Thesis Title: Exploring Sensory Responses in the Different Subdivisions of the Visual Thalamus Specialization Areas: Systems, Sensory.

Master Of Science - Ms, Medical Informatics

Thesis Title: Improving The Subcortical Gray Matter Segmentation Using Evolutionary Hierarchical Region Merging Specialization Areas.

Bachelor'S Degree, Electrical And Electronics Engineering, High Honors

Specialization Areas: Biomedical Engineering, Telecommunications, Signal Processing, Microwaves Minor Degree in Biology

High School Diploma, Mathematics

Specialization: Mathematics and Science