01 / SOFTWARE
Vendor-agnostic planning
Reads DICOM from any CT scanner. Pick entry and target; NeedleGPS computes the precise insertion angle.
IN COLLABORATION WITH
THE PROBLEM
Trajectories are planned in 3D.
Today they are executed by hand.
15,000/ day
CT-guided procedures performed every U.S. workday.
75%
of unassisted insertions miss a 10 mm target at depth.
> 10%
complication rate in complex cases, driven by needle reinsertions.
The angle the radiologist actually achieves rarely matches the angle they drew on the screen. Each reinsertion adds time, radiation, and the chance of a pneumothorax, bleed, or non-diagnostic biopsy.
THE SYSTEM
A bubble level, re-engineered for the CT suite.
Three components. One workflow. The device receives a target angle from the CT planning software or via manual entry, then guides the clinician to that angle with visual and haptic feedback in real time.
02 / DEVICE
Handheld bubble-level guidance
Reusable, palm-sized. The screen guides the clinician to the correct entry angle. Haptics confirm alignment before insertion.
03 / CONSUMABLE
Sterile single-use shield
A precision needle-guide interface in a sterile shell. Snaps on, keeps the field sterile, detaches once the angle is set.
THE DEVICE
Hand-held targeting with a bubble-level display, status LED, and haptic feedback. Coaxial needle mount accepts the sterile shield consumable. Detaches once the angle is set.
INSIDE THE DEVICE
Run the teardown sequence.
Reusable electronics. Sterile consumable. Single guidance needle. Each subsystem engineered for a clean handoff during the procedure.
PROTOTYPE EVIDENCE
Measured, not estimated.
Phantom insertion study. Experienced operators. Unassisted standard of care versus NeedleGPS-assisted.
25% → 90%
of insertions within 5° of the planned angle.
7×
reduction in needle reinsertions per procedure.
45%
faster in the toughest cases.
HISTORY
Built from a UCSF thoracic-radiology problem.
The team started with a single question: why do up to 1 in 5 CT-guided lung biopsies fail on the first pass? After hundreds of clinician interviews and four prototype generations through the UCSF Makers Lab, NSF I-Corps, and UCSF Innovation Ventures, NeedleGPS emerged as the simplest interface between a planning workstation and a clinician's hand. The company is now pre-seed with a working prototype and a prepared FDA pre-submission package.
TEAM
Built by operators and interventional clinicians.
NeedleGPS is led by Cristian Ionescu-Zanetti (CEO, prior founder and VP Product of a public diagnostics company) and Alan Czeszynski (CTO, multiple medical devices from 510(k) through commercial launch). Our clinician founders are practicing interventional radiologists and neuroradiologists at UCSF, with PhD training in physics and engineering at UC Berkeley, UC Santa Cruz, and MIT-affiliated programs. Clinical co-founders include Dr. Brian Haas, a thoracic radiologist at UCSF and San Francisco General Hospital (SFGH), and Dr. Lohith Kini, a UCSF neuroradiologist. Clinical development and clinical trials are run by Dr. Peter Hadar (Neurology, University of Pennsylvania and Massachusetts General Hospital).
EDUCATION & BACKGROUND
Harvard Medical School / MGH
MIT
Santa Clara University
SFGH
U.S. Naval Academy
UC Berkeley
UC Santa Cruz
UCSF
University of New Hampshire
University of Pennsylvania
FAQ
Common questions.
What is NeedleGPS?
A handheld targeting device, a sterile single-use consumable, and planning software that work together to guide the clinician to the correct angle on the first pass during CT-guided needle procedures such as biopsies, ablations, and injections.
Which CT scanners does it work with?
The planning software is designed to be vendor-agnostic and reads DICOM from any CT manufacturer. The handheld device receives the target angle wirelessly or accepts manual entry.
What is the consumable?
A single-use sterile shield with a precision needle-guide interface. It snaps onto the reusable handheld, keeps the field sterile, and detaches once the angle is set.
Is it FDA-cleared today?
NeedleGPS is currently an investigational device. Limited by United States law to investigational use. Not yet for sale. More information will be shared as we progress through early human testing.
Has it been tested?
Prototype testing has been conducted in bench and phantom settings with experienced operators. Early human testing is planned. Peer-reviewed publications will follow as the programme progresses.
Where can I find the IP?
A US patent application listing the founding team as inventors is pending. Once published, it will be searchable on Google Patents. More information will be shared as the programme progresses.
How can I get involved?
Clinicians interested in being design partners, investors, and potential strategic partners in imaging informatics can reach us via the contact section below.
GET IN TOUCH
Investors, clinicians, partners.
If you're a CT-guided procedure operator, an investor in surgical navigation, or a strategic partner in imaging informatics, we want to hear from you.