AI in oncology roadmap: pattern readers → foundation models → agents in the workflow

Software that reads scans alongside radiologists, catching cancers earlier and predicting who is at risk.

Low-dose breast X-ray used for screening. Newer 3D versions find more cancers with fewer false alarms.

The first AI for reading biopsy slides authorised by the FDA, which points pathologists to prostate cancer they might otherwise miss.

MSK spin-out with the first FDA-cleared AI pathology product and the Virchow foundation model.

Scanning microscope slides and letting software measure things a pathologist cannot see, including predictions of who will benefit from a treatment.

The scanners that turn glass slides into gigapixel images, and the software that stores and serves them, without which pathology AI cannot run.

Predicts a person's six-year lung cancer risk from one low-dose CT, even when no nodule is visible.

Reads a mammogram to estimate five-year breast cancer risk, consistently across races and devices.

Radiology AI company with the first FDA-cleared foundation-model triage platform (CARE, January 2026); oncology-relevant for incidental findings and workflow.

Predicts the 3D shape of proteins together with DNA, RNA, small molecules and antibodies, the starting point for much modern drug design.

Open-source structure models that match AlphaFold 3, with Boltz-2 also predicting how strongly a drug binds.

Structure and antibody-design models from Chai Discovery, with Chai-2 reporting high zero-shot antibody hit rates.

The tools that design entirely new proteins to bind a chosen target, now used for cancer binders and antibodies.

ESM3 is a generative protein model that designed a working fluorescent protein far from any natural sequence.

Using machine learning to pick targets, design molecules and antibodies, and predict which ADC will work.

Generative chemistry platform behind the first AI-discovered drug to reach phase 2, plus oncology candidates.

Generative-AI drug discovery company, listed in Hong Kong in December 2025, with a pan-KRAS candidate and a pan-TEAD inhibitor in the clinic.

Alphabet's AlphaFold-derived drug design company; its first AI-designed oncology candidate was cleared for human trials in January 2026 after a $2.1B raise.

Recursion is an AI-first biotech (merged with Exscientia in 2024) with a clinical oncology pipeline that includes an RBM39 degrader and a MEK inhibitor for familial adenomatous polyposis.

A model trained on billions of cell microscopy images to read what a drug or gene knockout does to a cell.

Launched in 2024 with over $1 billion to build AI-native drug discovery from Baker-lab protein design.

Google DeepMind built AlphaFold, AlphaMissense, AlphaGenome and Med-Gemini, the reference models for structure, variants, and medical multimodal reasoning.

Very large AI models trained on millions of slides or scans that can be adapted to almost any diagnostic question.

A pathology foundation model trained on millions of slides that can detect cancer and predict biomarkers from an ordinary H&E slide.

An open pathology model trained on 1.3 billion image tiles from a US health system, modelling whole slides at gigapixel scale.

Two open academic pathology models: UNI reads tissue images, CONCH links images with pathology text.

An open 1.1-billion-parameter pathology model from a French startup, among the strongest on public benchmarks.

TITAN is a model that summarises a whole slide, not just tiles, and can write a draft pathology report.

A model that reads slides and clinical text together to predict who will respond to immunotherapy.

A pathology model trained across 19 cancer types that predicts survival and mutations from slides.

The first AI tool cleared by the FDA to predict both prognosis and treatment benefit from a routine biopsy slide, in prostate cancer.

An FDA-cleared AI test (May 2026) that reads breast cancer slides to estimate recurrence risk in early hormone-positive disease.

First company with FDA-cleared AI pathology tests that predict treatment benefit (prostate 2025, breast 2026).

The first randomised trial of AI in breast screening found more cancers and cut radiologists' reading work almost in half without more false alarms.

Aidoc CARE is a single foundation model behind many FDA-cleared triage alerts in emergency radiology.

A model pretrained on 148,000 CT scans to segment organs and triage findings.

Merlin is a model trained on 15,000 CT scans with their reports that can find and describe hundreds of findings.

The first widely used transformer trained on millions of single cells, able to predict which genes matter in a disease.

A GPT-style model for single-cell data that predicts cell types, perturbation responses, and gene networks.

Predicts how cells will respond to a drug or gene knockout, trained on over 100 million perturbed cells.

Tahoe-100M is the biggest single-cell dataset ever released, built to teach AI how cancer cells respond to drugs.

The exam papers every pathology model is graded on.

Software that draws organs and tumours on scans automatically, saving hours per patient and making daily plan adaptation practical.

Limbus AI provides AI auto-contouring for radiotherapy, FDA-cleared and used across hundreds of centres.

TheraPanacea is a Paris AI company whose ART-Plan does auto-contouring and synthetic CT in radiotherapy.

Software, increasingly LLM-based, that reads a patient's record and finds trials or guideline options they qualify for.

Trial Library is software that helps everyday cancer clinics spot which patients might qualify for a clinical trial, refer them, and remove practical barriers like transport so more people, especially in under served communities, can join trials.

Massive Bio uses artificial intelligence to read a cancer patient's medical records and match them to clinical trials they may be eligible for, anywhere in the world, with doctors checking the results.

Google's medical versions of its Gemini models, able to reason over text, images, and long records.

A model trained on millions of hospital records that forecasts a patient's next diagnoses.

Training AI models across many hospitals without moving patient data, so the model learns from everyone while the data stay put.

French AI biotech using federated learning across hospitals; first CE-marked AI for MSI prediction from H&E.

Genomic testing plus one of the largest multimodal clinical datasets, used for AI models and trial matching.

Regular meetings where surgeons, oncologists, radiologists, pathologists and others review each patient's case together and agree a plan; mandatory in many countries and associated with more guideline-concordant care.

When an oncologist opens the order screen to prescribe a new line of treatment, the record would show the trials this patient may fit, with the nearest open site and a one-click referral.

Cancer AI models are usually tested on data from the same hospital they were built on. A registry of independent test datasets, and a rule that every model reports performance on at least one, would show which models really work.

Let validated AI make the first read on routine, high-volume samples like cervical smears and standard breast biopsy stains, so scarce pathologists spend their time on the difficult cases.

Hundreds of millions of CT scans are done each year for other reasons. Software could check each one for early lung, kidney, liver and pancreas changes, but only if a follow-up system exists.

Measuring tumours on scans for trials is slow, expensive and inconsistent between readers. Software that measures lesions and flags changes, checked by a radiologist, could make trial endpoints cheaper and more reliable.

Using a model of what would have happened to a patient on standard treatment, so fewer people have to be randomised to it.

The Imaging Data Commons is TCIA in the cloud, ready for large-scale model training.

Real-world evidence at scale: what happened to patients with a given genomic profile on a given treatment.

Thousands of cancer AI models are published; a handful are in clinical use, and fewer have shown they help patients.

Train one AI on scans, slides, genomics, and outcomes from many patients so it can predict, for a new patient, which treatment will work.

Models trained on Tempus's paired genomic, pathology, imaging and outcome data to predict response and prognosis.

Pathos AI uses very large artificial intelligence models trained on millions of cancer patients' records, scans and genetic data to pick which experimental cancer drugs to develop and which patients to test them in, and is building its own pipeline of such drugs.

Noetik builds artificial intelligence models of tumours from huge sets of tissue images and molecular data, to predict which patients will respond to a cancer drug and to find new targets.

Arc's growing library of cell data, the fuel for virtual cell models.

Simulating thousands of virtual patients on a computer can suggest which dose and schedule to test, so fewer real patients receive doses that are too high or too low.

How AI is moving from single-task readers of scans and slides towards systems that weigh everything about a patient, and what regulators and evidence still require.

The attempt to build a computer model of a cell good enough to predict what a drug or mutation will do before anyone runs the experiment.

Thousands of cancer AI models are published; a handful are in clinical use, and fewer have shown they help patients.

Records, scans, genomes and outcomes sit in separate systems that cannot talk. Every patient's experience is lost to the next.

We do not reliably know what happens to patients after approval, so we cannot tell which drugs deliver in practice.

The number of people with cancer is rising faster than the workforce trained to treat them.

Fewer than half of landmark cancer biology findings reproduce when someone else tries.

AI compute platforms are the GPUs, model libraries, and cloud services that pathology, radiology, and drug-design AI run on.

Supplies the GPUs and the BioNeMo framework most biological foundation models are trained on, and co-developed Evo 2 with Arc.

BioNeMo is the software stack many biology foundation models are trained and served with.