Career forks

Most people pick a career by default: the path their lab, their friends, or their first internship put them on. Picking on purpose is the multiplier, and picking on purpose requires having seen the options. You do not have to decide until senior year. You do have to talk to one person on each path by the end of sophomore year, and here is what to ask about.

PhD and research

What it is. Five to six years of funded research toward a dissertation, then postdoctoral work or a research position in industry. The path into faculty jobs, into research scientist roles at BCIBrain-computer interface (BCI) A system that turns measured brain activity into control of something outside the body, from a cursor to a wheelchair to a speech synthesizer. Related: DecoderDecoder The algorithm that maps neural signals to intended actions, such as a classifier or a Kalman filter. , Closed loopClosed loop A system in which the output feeds back to influence the input, such as a user adjusting their brain activity while watching the cursor it controls. . Glossary entry companies, and into the parts of the field where the questions are still open.

What it needs by graduation. The research ladder climbed to a poster and ideally a publication; strong letters from two research mentors; a clear statement of what you want to study and why this lab. Grades matter less than research experience.

The person to talk to. A graduate student in a lab you like, about what their week actually contains, and a postdoc about what came after.

Industry R&D

What it is. Engineering at a device company: owning a board, a decoderDecoder The algorithm that maps neural signals to intended actions, such as a classifier or a Kalman filter. Related: ClassifierClassifier An algorithm that assigns each trial a label, such as left hand or right hand, after being trained on labelled examples. , Kalman filterKalman filter An estimator that blends a prediction from a motion model with a noisy measurement, weighting each by how much it trusts it; the decoder behind most intracortical cursor control. , Brain-computer interface (BCI)Brain-computer interface (BCI) A system that turns measured brain activity into control of something outside the body, from a cursor to a wheelchair to a speech synthesizer. . Glossary entry, a test system, a material. Fast, collaborative, constrained by regulation and manufacturing, and the place where the most devices reach the most patients.

What it needs. A spine, demonstrated: the characterized board, the reproduced decoderDecoder The algorithm that maps neural signals to intended actions, such as a classifier or a Kalman filter. Related: ClassifierClassifier An algorithm that assigns each trial a label, such as left hand or right hand, after being trained on labelled examples. , Kalman filterKalman filter An estimator that blends a prediction from a motion model with a noisy measurement, weighting each by how much it trusts it; the decoder behind most intracortical cursor control. , Brain-computer interface (BCI)Brain-computer interface (BCI) A system that turns measured brain activity into control of something outside the body, from a cursor to a wheelchair to a speech synthesizer. . Glossary entry, the fabricated electrodeElectrode The conductor that turns ion currents in tissue into electron currents in a wire; its chemistry and contact quality set the noise you live with. Related: Silver/silver-chloride (Ag/AgCl)Silver/silver-chloride (Ag/AgCl) The standard electrode material for biopotentials, because its chemistry produces a stable, low-noise, low-drift interface with salty gel. , Electrode impedanceElectrode impedance How hard it is to push current through the skin-electrode contact, typically 5 to 50 kilohms with gel and megohms dry, higher at low frequencies. , Half-cell potentialHalf-cell potential The DC voltage that appears wherever a metal meets an electrolyte; two electrodes with different half-cell potentials produce an offset that can swamp the signal. . Glossary entry. An internship or co-op. Fluency in design controlsDesign controls FDA's required process for developing a medical device: user needs, requirements, design, verification, validation, and the records of all of it. Related: Design history file (DHF)Design history file (DHF) The complete record of how a device was designed, showing every requirement traced to a test; writing a mock one teaches more than any lecture. , RequirementsRequirements Testable statements of what the device must do, such as "input-referred noise below 1 microvolt RMS from 0.5 to 40 Hz." , VerificationVerification Testing that the device meets its requirements; did you build what you specified. , ValidationValidation Testing that the device meets the user's need in real use; did you specify the right thing. . Glossary entry so you can start on day one.

The person. An engineer at a Salt Lake device company, via BioUtah or alumni, about what they build and what they wish new graduates knew.

Regulatory, quality, and clinical affairs

What it is. The people who get devices approved and keep them safe: writing submissions, running risk management, auditing quality systems, designing and running clinical trials. Underpopulated by engineers who understand the technology, well paid, and the path to leadership at device companies more often than R&D is.

What it needs. The Phase 5 documents: a mock DHFDesign history file (DHF) The complete record of how a device was designed, showing every requirement traced to a test; writing a mock one teaches more than any lecture. Related: Design controlsDesign controls FDA's required process for developing a medical device: user needs, requirements, design, verification, validation, and the records of all of it. . Glossary entry, an FMEAFMEA Failure mode and effects analysis: a table of every way the device could fail, how bad and how likely each is, and what you did about it. Related: ISO 14971ISO 14971 The international standard for medical device risk management that every company follows. . Glossary entry, a human factorsHuman factors Designing so the device fits how people actually behave, especially under stress and fatigue, so that errors are hard to make. Related: IEC 62366IEC 62366 The standard for usability engineering of medical devices, aimed at preventing use errors. , Use errorUse error A mistake made while using a device that the design invited, as opposed to random user carelessness. . Glossary entry observation, and the ability to talk about three real submissions. An internship in a quality or regulatory group, which are easier to get than R&D internships.

The person. A regulatory affairs or quality engineer, about a submission they worked on.

Clinical

What it is. Medicine, with the engineering as a foundation: neurology, neurosurgery, rehabilitation medicine, or the allied fields (audiology, prosthetics and orthotics, clinical neurophysiology). Or clinical engineering inside a hospital, running the technology.

What it needs. For medical school, the premedical coursework alongside the BME degree, clinical hours (the clinic thread), and the MCAT. For allied fields, their specific graduate programs. For clinical engineering, the BME degree and hospital experience.

The person. A physician-engineer (they exist and they are the most sought-after people in the field) or a clinical engineer at the university hospital.

Patent law

What it is. A law degree after the engineering one, then work as a patent attorney or agent writing and prosecuting patents for device companies. Uncrowded, lucrative, and dependent on exactly the technical understanding this site builds. The patent agent route (an exam, no law degree) is faster.

What it needs. The engineering degree, good writing, and an interest in the boundary between what is new and what is not. Reading a few neurotech patents and their claims is the way to find out whether this appeals.

The person. A patent attorney at a firm that represents device companies; they exist in Salt Lake and they take coffee meetings.

Founding

What it is. Starting a device company. Utah is a good place to do it: the corridor, the university’s Center for Medical Innovation, the entrepreneurship programs at Lassonde, and a culture of small device companies. Most founders in this field have done one of the other paths first, and the ones who start out of Bench to Bedside are the exception that shows it can work.

What it needs. A validated need (Phase 5), a team, and the willingness to spend years on regulation and fundraising before a patient is helped. The BiodesignBiodesign Stanford's process for medical technology innovation that starts from an observed clinical need rather than a technology looking for a use. Related: Needs findingNeeds finding Observing clinical work to catch the moments that are slow, awkward, or unsafe, before proposing any solution. , Need statementNeed statement One sentence, "a way to [do X] for [whom] that [outcome]," with no solution in it. . Glossary entry textbook’s second half is about this.

The person. A founder of a small device company, about what they wish they had known.

The one thing to do this year

Six conversations, thirty minutes each, over two years. Ask each person what their week looks like, what they wish they had done as a student, and who else you should talk to. Write five lines after each, as with a paper. By the end of sophomore year you will know which fork is yours, or you will know which two, and either is enough.