Diane Whitmer (dwhitmer@biomail.ucsd.edu)

Graduate Program:  Computational Neuroscience
Lab PI: Scott Makeig, Swartz Center for Computational Neuroscience
Undergraduate Institution: Dartmouth College
Med-into-Grad clinical training area: Neurology: Epilepsy
Main clinical mentors: Dr. Iragui (UCSD), Dr. Tecoma (UCSD), Dr. Worrell (Mayo Clinic)

Quote:  “This training has made me cognizant of the large gap between clinical practice and basic science research. The technology used in many clinical settings lags behind what is used in research by ~50 years. It has been one of many factors that have inspired me to work in clinical research as a post-doc. My clinical interests have extended beyond epilepsy, to the development of brain-computer interfaces for paralyzed patients. The same technology used to monitor epilepsy patients and for epilepsy research, provides a unique opportunity for studying the output of motor cortex that could help the development of therapeutics for patients with paralysis and other motor disorders.”

Rationale for Med-into-Grad training:
Medical training and identification of medically-relevant research issues:
Training in diagnostics & therapeutics, and identification of unmet diagnostic & therapeutic needs:
Diagnostic and therapeutic collaborations :
Long term impact.:
Advice for new trainees--autumn preparatory quarter:
Advice for new trainees--winter clinical training quarter:
Take home perspective on Med-into-Grad at UCSD:

Rationale for Med-into-Grad training:  My dissertation research involves using multi-scale electrophysiological brain data from epilepsy patients for cognitive neuroscience. Patients diagnosed with epilepsy originating from a focal region, and who are not responsive to pharmaceutical therapeutics, are candidates for treatment via surgical resection. Since these patients have electrodes implanted, they provide the unique opportunity to study the electrical signals directly from the human brain. My dissertation research entails collaboration between the Swartz Center for Computational Neuroscience, and the Mayo Clinic in Rochester, Minnesota. I am using concurrent scalp electroencephalogram (EEG) and intracranial EEG (iEEG) recordings from epilepsy patients, to understand the underlying spatiotemporal dynamics of visually cued movements. We apply independent component analysis (ICA) to decompose the signal mixtures as recorded on electrodes into underlying independent sources. We also use spectral analysis to examine changes the amplitude and phase-locking of oscillations, timed to visual and motor events. At the time that I applied for the med-into-grad program, my thesis proposal also included a chapter on using these approaches to study interictal (between seizure) epileptiform activity. I wanted to participate in this program to 1) learn how clinical EEG and iEEG data is interpreted and gain practice reading EEG signals in the “time domain” (as opposed to the “frequency domain” approach taken in my research), and 2) become informed of additional clinical research issues in neurology in general and epilepsy in particular.

Medical training and identification of medically-relevant research issues:

UCSD. My clinical training at UCSD with Dr. Iragui and Dr. Tecoma involved daily participation in interpretation of clinical EEG data, and daily rounds within the in-patient epilepsy monitoring unit, attending case conferences, shadowing doctors as they saw epilepsy and general neurology out-patients, observing a neurosurgery, and helping acquire patient data for pharmaceutical clinical trials.

Mayo Clinic. At the Mayo Clinic with Dr. Worrell, I observed neurosurgeries and clinical mapping studies, observed the out-patient treatment of epilepsy patients with Neuropace (an implanted device that delivers electrical stimulation for treatment of seizures), acquired new cognitive data from patients for my thesis and epilepsy data for colleagues in my lab, and attended general neurology rounds.

Based on my experience in these rotations, two of the most important areas of epilepsy research are seizure detection and seizure prediction. Seizure detection includes identifying both when a seizure is occurring, and also where in the brain (localization). Seizure localization diagnostics could be improved with better source localization models.  Seizure prediction is also an active area of research. Dr. Worrell’s team is using new technology to record seizures and interictal activity from simultaneously implanted cortical surface macroelectrodes and microwires. They have characterized the relationship between signals recorded on electrodes that average over different spatial scales, and have found “mini seizures” on microwire electrodes that precede seizures seen on larger electrodes.

Research collaborations:  I already had a research collaboration with Dr. Greg Worrell from the Mayo Clinic established, and have continued that collaboration through my dissertation work. As a post-doc in neuroscience, I plan to continue collaborating with Dr. Worrell, moving in a new research direction aimed at helping develop brain-computer interfaces for paralyzed patients.

Long term impact:  This training has made me cognizant of the large gap between clinical practice and basic science research. The technology used in many clinical settings lags behind what is used in research by ~50 years. It has been one of many factors that have inspired me to work in clinical research as a post-doc. My clinical interests have extended beyond epilepsy, to the development of brain-computer interfaces for paralyzed patients. The same technology used to monitor epilepsy patients and for epilepsy research, provides a unique opportunity for studying the output of motor cortex that could help the development of therapeutics for patients with paralysis and other motor disorders.

Training in diagnostics & therapeutics, and identification of unmet diagnostic & therapeutic needs:  Diagnostic methods include conducting an extensive history and physical exam, outpatient electroencephalography (EEG) recordings, in-patient video-EEG monitoring, in-patient monitoring with surgically implanted electrodes (iEEG), anatomical brain imaging including MR and CT, and in some cases ictal SPECT.
For therapeutics, most epilepsy patients are treated with anti-epileptic drugs (AEDs). Most patients are on 2-3 different drugs and incur significant side effects. AEDs only work in seizure prevention/management for some fraction of epilepsy patients; the remaining patients are diagnosed with "medically refractory" epilepsy. There are two broad classes of epilepsy: 1) generalized, and 2) focal. In the case of focal medically refractory epilepsy, brain surgery is a possible treatment if the epileptic focus is not in a region of "eloquent brain" that would interfere with critical functionality including language, memory, and motor activity. Some more recent advances in treatment have involved electrical stimulation to prevent or mitigate seizures. Vagus nerve stimulators are implanted into the peripheral nervous system, and Neuropace implants implanted in cerebral cortex.
Diagnostics could clearly be improved with better source localization models. The physicians with whom I worked acknowledged that there are times when scalp EEG signals either do not detect epileptic signals because the sources of those signals are too deep, and also times when scalp EEG falsely localize or localize with insufficient spatial resolution. With better mathematical modeling of current spread through the brain ("forward modeling"), interpretability of scalp EEG signals could be improved. This is an area of research currently being pursued by my thesis lab.

Diagnostic & Therapeutic collaborations:  Although not part of my dissertation, members of my thesis lab are working on applying signal processing techniques for the localization of seizures, and forward modeling (modeling the flow of current from brain sources and how you’d expect those signals to distribute on electrodes). The data for this project is data that I acquired during my rotation at Mayo through the med-into-grad program.

Advice for new trainees--Autumn preparatory quarter:
1. I recommend volunteering as a teaching assistant for Dr. Kritchevsky's Basic Neurology course in the School of Medicine, during the spring quarter prior to starting your program. It is an excellent academic overview of clinical neurology, and will also facilitate review of neuroanatomy.
2. The earlier you can contact and establish a relationship with your advisors, the better. Meet with your mentor(s) at least one academic quarter before your rotation, if not earlier, to discuss and plan logistics. Find out what opportunities exist within the clinical setting and develop a detailed plan for engaging in those activities. Additionally, get feedback from your mentor on specific areas of academic preparatory training you should do before starting, e.g. which books to purchase or classes to take.
3. Try to expose yourself not just to the neurology of your particular disease, but also to neuroradiology and neurosurgery. You won't have a complete picture of the diagnostic and therapeutic process without some exposure to all three of these branches of medicine.
4. Establish a schedule that allows you to continue your thesis research at the same time as participating. Devoting half time over a two-quarter period might be better than full time/overtime in the clinic for one quarter. I think maintaining one foot in my lab while also working in the clinic would have fostered more discussions about clinical research.
5. Branch out and find additional people to interact with outside of your specific MSP group. Some of the highlights of my experience were observing an entire brain surgery start to finish, spending a day with a neuropsychologist who does functional mapping during neurosurgeries, and making contact with imaging (neuroradiology) people for prospective post-doc opportunities.

Advice for new trainees—Winter clinical training quarter:  My advice to new students is to take initiative. Define for yourself what you’d like to learn, what kinds of experiences you’d like to have, and then find a way to make it happen. There is no cookie-cutter approach that will suit all participants; this program is designed for creative and self-motivated students.

Take home perspective on Med-into-Grad at UCSD:  This is a great program for graduate students interested in clinical research, and for those who desire interdisciplinary training and would enjoy patient interaction not necessarily possible in a lab research setting. I highly recommend participating in this program!