Digital Pathology AI
P

PreciseDx

PreciseDx sells PreciseBreast, a test that predicts whether early breast cancer will come back, using AI to read the cell patterns on a routine stained slide. Oncologists, breast surgeons and pathologists order it as an alternative to gene expression tests.

The AI grades the tumor from the slide, and the test combines that grade with tumor size, lymph node status and age. The result places a patient in a low or high risk group for recurrence. It is for early stage, hormone receptor positive, HER2 negative invasive ductal carcinoma with up to three positive lymph nodes. Results come back in hours, and PreciseDx says the test costs 30 percent of what gene expression tests cost.

PreciseDx runs the test in its own certified laboratory in New York and Florida, and bills it under its own billing code. It is a laboratory developed test, and no FDA clearance is listed. The method was validated on more than 1,600 Mount Sinai cases in a 2022 peer reviewed paper, and in an independent Dutch cohort in 2025.

PreciseDx was spun out of Mount Sinai. Its investors include Eventide Asset Management, Merck's Global Health Innovation Fund, Philips Ventures, Labcorp and Quest Diagnostics. No list price is published.

AI Health Index verifiedNovember 8, 2026
Compare PreciseDx with other vendors
Founded
2021
Headquarters
New York, New York, United States
Categories
pathology-ai, diagnostics-and-genomics
Indexed Products
PreciseBreast, OncoIntelligence platform
Buyer Segments
Oncologists, Breast surgeons, Pathologists, Health plans and payers
Assessment

Capability Axes

An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check. How grades read

AI Capability
AA on AI CentralityThe artificial intelligence is the product. Remove the model and there is nothing left to sell.
Vendor Published

The model is the product. PreciseBreast's score combines an AI grade, computed from cell and tissue patterns on the H&E slide, with tumor size, node status and age. Those clinical factors are already in the patient's chart, so without the AI grade there is no test to sell. PreciseDx offers no other product or service beside it.

CC on Autonomy and Oversight ModelAutonomy is claimed and oversight is asserted without a mechanism. Human in the loop appears as a phrase rather than a described control.
Vendor Published

Results go to an ordering clinician, and the controls around the score are not described. Oncologists, surgeons and pathologists order PreciseBreast and get it back as a laboratory result, so a clinician decides what the risk group means for treatment.

PreciseDx publishes nothing on whether a pathologist reviews the AI grade or the tumor outline before release, when a slide is rejected as unsuitable, how the cut point between low and high risk was set, or how a score near the cut point is reported.

BB on Model and Technology TransparencyThe approach or the suppliers are named without the version and update discipline behind them. Naming a supplier is the entry to this band both here and on Model Supply Chain Disclosure, which ask different questions of the same disclosure: who receives the data, and what produces the output.
Vendor Published

The method is named and in print, but no version or update practice accompanies it. The company describes the OncoIntelligence platform as outlining the tumor, deriving an AI grade from cell and tissue patterns on H&E slides, and combining it with tumor size, node status and age. The method was published in Breast Cancer Research in 2022.

No model version is published, and the training data is described only as Mount Sinai cases. Nothing describes how the model or the cut point is updated, or how ordering clinicians are told.

BB on Model Supply Chain DisclosureSubstantial partial disclosure, or a chain that is structurally short: an in house build, a cleared model that cannot be quietly swapped, or a deployment where the transfer does not occur at all. Naming the model provider exits the band below into this one; the axis rises from here on the completeness of the party list and on the terms that govern data once it arrives.
Vendor Published

PreciseDx developed its own model from Mount Sinai pathology research and runs it in its own laboratory, so no outside model provider sits between a slide and a result. The notice says business associates must protect patient information, without naming them. No hosting provider or subprocessor list is published.

BB on Clinical and Operational EvidenceNamed deployments with dated outcome figures and enough method to test them, or published research short of independent validation.
Peer Reviewed Publication

Peer reviewed validation started at the founding institution and now includes an independent cohort, with the newest results in conference abstracts. The method and its validation on more than 1,600 breast cancer cases from Mount Sinai were published in Breast Cancer Research (Fernandez and colleagues, 2022), by the founders. A study published in August 2025 externally validated PreciseBreast in an independent early stage cohort from the Netherlands and compared it with other risk models, and PreciseDx's page gives no figures from it.

A comparison with Oncotype in a gene expression tested cohort from Baptist Health Miami, with COTA, was presented at SABCS 2024 as showing equivalence. Three abstracts at SABCS in December 2025 covered risk stratification on biopsy specimens, triple negative disease, and a prospective evaluation at Lankenau Medical Center that PreciseDx describes as independent. PreciseDx also names a collaboration with NSABP on chemotherapy benefit, and reports 98 percent precision and reproducibility in tumor segmentation and less than 2 percent variation in the risk score.

The headline claims, equivalence with gene expression testing at lower cost and faster turnaround, rest on the conference comparison, not a peer reviewed or prospective study.

CC on AI Safety and PHI StewardshipGeneral assurances of privacy and security that do not answer the questions artificial intelligence raises: what is retained, what reaches a model, and what happens to it there.
Vendor Published

The privacy notices leave the questions the model raises unanswered. One of those notices allows research use of protected health information where required legal steps are taken. Nothing published says how long slides, images and results are kept, whether patient slides are used to train or recalibrate the model, or how training data was de identified.

The model was developed on Mount Sinai cases, and no data governance terms for that archive are published. No safety engineering for the model is published.

Regulatory and Compliance
BB on HIPAA and BAA PostureBusiness associate status is stated and supported by a substantive privacy document, with the agreement or its scope not fully published. For a vendor outside the United States, an equivalent regime documented to this depth grades here. A vendor whose product does not process protected health information grades here too when it says so plainly and explains the scope, with a service privacy document behind it: stating a position a buyer can rely on is the posture this axis grades, and the band above is closed to it because there is no agreement to publish.
Vendor Published

What is published is the privacy notice a laboratory covered by HIPAA issues to its patients. The Notice of Privacy Practices, effective 1 March 2025, describes how PreciseDx Inc. uses and shares protected health information and the rights patients have, including amendment. It covers research use where legal safeguards are met and the duties of business associates, and names a privacy contact. The notice shows PreciseDx's role as a covered entity, though it does not state it in words.

A separate consumer health data privacy policy and a website privacy policy are also published. No agreement for health system or payer partners is published.

DD on Security Certifications and Trust CenterControls are asserted with nothing independent behind them, or nothing is published. Read the note before concluding anything: this is the grade most often corrected on a second pass, because assurance material frequently sits on a parent domain or inside an old announcement rather than on the product pages.
Vendor Published

PreciseDx publishes no security audit or certification. It describes itself as HIPAA compliant and publishes privacy notices. No SOC 2 report, ISO/IEC 27001 certificate, HITRUST certification, penetration testing statement or trust center appears on its site.

CLIA certification, New York State certification and CAP accreditation cover laboratory quality, not information security.

CC on FDA and Regulatory StatusNo device claim is made and the product is scoped accordingly. Most administrative and operational products sit here and are not penalized for it, because this axis grades the appropriateness of the positioning rather than possession of a clearance.
Vendor Published

Sold as a laboratory developed test under state and federal laboratory rules, PreciseBreast carries no device claim. The company describes a CLIA certified laboratory and New York State Department of Health certification, both from 2022, and CAP accreditation. The FDA's databases list no PreciseDx device, and no FDA position is stated. State certification governs the laboratory itself.

The test is marketed as equivalent to gene expression testing for recurrence risk. The status behind that clinical claim rests on laboratory regulation alone.

CC on AI Governance and Bias DisclosureResponsible artificial intelligence is committed to in policy language with no evaluation behind it. Most of the index sits here.
Vendor Published

Analytical consistency is claimed, and results by group are not published. The company reports 98 percent precision and reproducibility in tumor segmentation and less than 2 percent variation in the risk score. It does not give the number of slides, scanners, operators or sites behind those figures. It reports validation in three cohorts, without results compared across age, menopausal status, node status, race or ethnicity, or scanner on its pages.

Breast cancer recurrence risk and tumor appearance differ by age and ancestry, which is where a model that reads tumor appearance is most likely to underperform. No governance process or monitoring after release is described.

BB on AI Liability and RecourseA published falsifiable commitment, or a real correction route the affected person can exercise against the vendor. A published error rate with its method and denominator grades here. A statutory right that runs to the covered entity rather than to the vendor does not reach this band on its own: every other route here asks something of the vendor, and being located in a particular jurisdiction is not conduct.
Vendor Published

PreciseDx reports the result as the laboratory, and its privacy notice gives the patient a route to correct it. The risk group is the AI driven result the laboratory reports. Under the Notice of Privacy Practices, a patient can ask PreciseDx in writing to amend information, and PreciseDx answers in writing whether it accepts or denies the request. No time limit is stated.

No error rate at the risk cut point, warranty, retesting commitment or indemnity is published.

Integration and Deployment
DD on EHR and Interoperability DepthNo integration evidence. A connector described as available on request grades here until one exists.
Vendor Published

There is no published integration. Clinics order PreciseBreast through PreciseDx's own online portal after registering, and the pages for oncologists cover billing and the patient assistance program. No record system, results interface, laboratory information system or standard is named, and how results reach the ordering clinician is not described.

CC on Deployment Model and Data ResidencyA single hosted option with location implied rather than committed.
Vendor Published

PreciseBreast is a laboratory service, and where data sits is implied, not stated. Slides go to PreciseDx's own laboratory in New York or Coral Gables, Florida, and results come back as a laboratory report, so the ordering practice runs no software.

Nothing published says where digital slide images and results are stored or processed, or which cloud or hosting provider is used.

Commercial
CC on Commercial TransparencyNo price is published and the posture is discoverable: a buyer can establish how the product is sold and what drives the cost before contacting the vendor. Most of the index sits here.
Vendor Published

No price is published, but how the test is billed can be worked out. PreciseBreast is billed per test under its own billing code, CPT 0220U. Insurance coverage varies by plan, and there is a patient assistance program and interest free payment plans of up to six months.

PreciseDx says the test costs 30 percent of gene expression testing, which its home page puts at more than $4,000, without stating its own charge.

BB on Setting and Specialty CoverageCoverage is named with validation behind part of it.
Vendor Published

The eligible patients are named precisely, with validation behind part of that group. PreciseBreast is for untreated, early stage, hormone receptor positive, HER2 negative invasive ductal carcinoma with no or one to three positive lymph nodes, at any menopausal status. It predicts recurrence over up to 8.8 years.

Validation covers a Mount Sinai cohort of more than 1,600 cases, a Baptist Health Miami cohort and a Dutch cohort. PreciseDx's pages do not show results separately for node positive patients or for tumor types other than invasive ductal carcinoma. Oncologists and surgeons order it as a prognostic laboratory test.

Alternatives

Alternatives to PreciseDx

Other products that do the same kind of work as PreciseDx, and what each one does.

All PreciseDx alternatives, with prices →

Stratipath Breast assigns early stage invasive breast cancers to low or high risk groups from routine H&E slides, as a faster, cheaper alternative to gene expression tests for pathology laboratories and breast cancer teams in Europe.

ArteraAI is the Los Altos, California company at artera.ai that builds multimodal AI cancer tests. It is unrelated to Artera of Santa Barbara (artera.io), which makes patient communication agents.

Imagene AI reads routine stained lung cancer slides and predicts which gene changes a tumor carries, without using more tissue for sequencing.

Waiv sells AI tests that read routine H&E slides, for pathology laboratories and oncology teams in Europe, and biomarker services for drug developers.

Veracyte sells genomic tests that pair a laboratory expression assay with a learned classifier, used by urologists, oncologists and endocrinologists to decide how intensively to treat a cancer or whether a thyroid nodule needs surgery.

Cleo Breast and Cleo Skin are AI tools for clinical pathology laboratories in Europe that detect and quantify cancer features in whole slide images.

Commercial

Pricing

Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.

Entry Price Pricing Basis BAA Tier Implementation Source
Not published
Per test laboratory billing under CPT 0220U; insurance coverage varies by plan; patient assistance program and interest free payment plans up to six months. No list price published. Covered entity laboratory: a HIPAA Notice of Privacy Practices (effective 1 March 2025) is published; no business associate agreement for partners is published. None for the ordering practice; the test is a laboratory service. Vendor Published

PreciseDx bills PreciseBreast per test under its own billing code and says it costs 30 percent of gene expression testing, which its home page puts at more than $4,000, without stating its own charge. Patients can apply to an assistance program or spread costs over up to six months. No list price or expected out of pocket cost under a plan is published.

PreciseDx pricing in full →