AI in Medicine — Cancer & Rare Disease — 2026-09-23 (Wed)

Overview

The most active movement is related to regulatory approval of diagnostic imaging and pathology AI (expansion of FDA approval). While there are multiple reports on prognosis prediction and preoperative evaluation in diagnostic areas such as breast cancer MRI and pancreatic cancer pathology AI, in the drug discovery area, the search for new candidates using AI agents and Iambic's IPO filing are notable. Specific data from ClinicalTrials.gov regarding trial status is not included in this input.

Breast cancer

- Reported that AI using breast MRI predicted the spread (progression) of cancer before surgery (Bioengineer.org). Subjects and evaluation level (single facility, prospective/retrospective) are not stated in the heading. - A report that multimodal AI can refine the decision to intensify postoperative drug therapy in early-stage high-risk breast cancer (Oncodaily). The stage/evidence level is unclear from the headline. - SimBioSys and Venture Café Phoenix will hold a panel on AI, innovation, and the future of breast cancer care (PR Newswire, event announcement).

lung cancer

The development stage and verification method are not stated in the heading (presumed to be in the preclinical stage, but not specified as the text has not been confirmed). - FDA clears AI triage software for 10 abdominopelvic CT findings (diagnosticimaging.com). It targets the abdominopelvic region rather than lung cancer. - A review article (CancerNetwork) discussing the evolution of AI in oncology and its impact on clinical trials, workflows, and outcomes.

Malignant brain tumor

- Reports that AI-assisted triage in the emergency department improved triage accuracy and patient flow (2 Minute Medicine). There is no mention of brain tumor specialization, but general emergency triage content. - AI-assisted brain surgery saves vision in Bedfordshire patient (Healthcare Today). This is a single case report, and generalizability cannot be determined from the headline.

Leukemia/lymphoma

- Reported that AI interpreted lymph node tissue images of lymphoma and showed accuracy close to that of pathologists (targeting canine lymphoma, Bioengineer.org). Extrapolation to humans is not mentioned in the heading. - Reported that AI automatically diagnosed PET/CT images of lymphoma and predicted survival prognosis (Bioengineer.org). No description of validation population or equivalent evidence level. - Reported that AI method has identified a candidate drug for leukemia treatment (Technology Networks). The development stage (preclinical, etc.) is unclear from the headline. - Utilization of AI is a hot topic at "Light the Night" (blood cancer awareness event) in Rochester (spectrumlocalnews.com, event related article).

Colon cancer/digestive organ

- Reported that AI scans pathology slides and predicts recurrence of pancreatic cancer (healthcare-in-europe.com). The number of target patients and whether the study was prospective/retrospective are not stated in the heading. - Mayo Clinic's AI detected pancreatic cancer several years earlier than diagnosis (Medical Economics, in a breaking news summary with other news). Detailed study design is not described. - No report on endoscopic detection specific to colorectal cancer is found in the heading.

Rare cancer

There are no direct reports on adenoid cystic carcinoma in the heading. Regarding rare diseases, there has been an announcement of a Korean ARPA-H joint research project (approximately 3 billion won, with an additional 800 million won) aimed at linking rare disease diagnosis AI to treatment development, but the cancer type has not been specified.

Diagnosis/Pathology/Image

- FDA allows expansion of abdominopelvic CT triage AI to support 10 findings (diagnosticimaging.com, a2z Radiology AI related itnonline.com). - a2z Radiology AI announces expansion of acute care suite by 10 new findings (Business Wire). Details of FDA approval (clearance classification, etc.) are unclear from the headline. - Reports an AI assistant that learns from radiology reports and points out abnormalities in abdominal scans (Medical Xpress). - Voicebrook announces digital pathology collaboration with Indica Labs to realize automatic AI biomarker reporting (PR.com, company announcement). - A market trends article (globenewswire.com) states that the automated microscopy market is expanding due to advances in AI image analysis.

Treatment selection/drug discovery

- Iambic Therapeutics (funded by Nvidia) has applied for an IPO on Nasdaq and plans to use the proceeds to develop cancer treatment drugs (Yahoo Finance, Axios, AllSci). This stage is the funding stage for drug discovery companies. - AbbVie and Iambic announce multi-year partnership in AI drug discovery (Healthcare Digital, Investing.com). Details of the stage (preclinical/clinical) are not listed in the heading. - An article discussing how AI can discover undiscovered targets for cancer immunotherapy (Oncodaily, opinion piece). - Article discussing the potential use of curation data and AI in regional drug discovery (Mexico Business News).

Regulations/Approval

- FDA allows AI triage software for abdominopelvic CT to expand support to 10 findings (diagnosticimaging.com, Business Wire). The approval classification of the item (510(k), etc.) is not specified in the heading. - Illinois Governor Pritzker signed an executive order (state-level executive order) establishing an "AI cabinet" to conduct AI risk assessments (Patch). A state-wide AI risk assessment framework, rather than regulations specific to cancer care. - There are reports that Goldman Sachs has proposed a target price of $75 for Tempus AI (TEM), but this is investment-related information and is outside the scope of this report (stock market prices are not covered).

Clinical trial implementation status

No relevant information (The current input does not include data on the specific trial name, target cancer type, phase, and registration status from ClinicalTrials.gov).

Japanese movements

- Iambic Therapeutics' application for listing on Nasdaq has been reported in Japanese media (BigGo Finance). - An article (d Menu News) that discusses the expansion of treatment possibilities through immunotherapy and AI drug discovery, asking, "To what extent can cancer, the number one cause of death in Japanese people, be cured?" There is no definitive description of a curative effect. - An article discussing the possibility of new drug development using AI to discover unexpected relationships between diseases (Newsweek Japan Edition). - Article dealing with the controversy surrounding AI's involvement in medical decisions and the "responsible entity" (BigGo Finance). - AI drug discovery partnership between AbbVie and Iambic is reported in Japanese (Investing.com).

Remarks from VIPs

The list of important people being tracked is blank, so no relevant information is available.

Latest news on AI and rare and incurable diseases

Overview

In the diagnostic field, there are mainly presentations on AI pathology and genome mapping, and there are no headlines indicating specific progress in therapeutic drugs or systems. Several companies announced the construction of genetic analysis infrastructure (dataset preparation). There are no headings showing individual trends in the 348 designated intractable diseases.

Diagnosis of rare diseases

MGI announced its AI-enabled pathology portfolio “VisiOmics” in Europe at ECP 2026

News-Medical

. Numerical values showing diagnostic accuracy and shortened time to diagnosis are not listed in the headline and are unknown. Google DeepMind maps 9 billion DNA mutations and shows its potential for use in elucidating rare diseases

WION

. This is the construction of a mutation map, and does not indicate clinical application to cases or improvement in diagnostic rate.

Gene analysis/genome

Precure selects Helix to provide multi-omics testing and AI tools [Genetic Engineering and Biotechnology News]. Pure Genetic Medical and Breakthrough Genomics launch AI-powered “Virtual Geneticist”

Select Science

, There are no specific numerical values such as mutation determination accuracy in the heading. Tempus AI is also building a large-scale whole-genome multimodal dataset

Bitget

Yahoo Finance

. All of these are at the stage of developing research infrastructure and datasets, and are not reports that demonstrate clinical ability to interpret mutations. The Mainichi Keizai article reports that DeepMind used AI to pre-analyze and map the approximately 9 billion single-letter changes (single nucleotide mutations) in human DNA, and appears to be referring to the same initiative as WION's report.

Development of therapeutic drugs

No applicable information.

Nerve/muscular diseases

No applicable information.

Immunology/Collagen disease

No applicable information.

Rare metabolic/endocrine diseases

No applicable information.

System/Patient Support

No applicable information.

Japanese movements

It was reported that Anthropic will establish a biology research institute and begin full-scale AI drug discovery efforts.

Infoseek News

. The target disease/stage (pre-clinical or not, etc.) is not stated in the heading and is unknown. Mainichi Keizai (Korean media) has reported on DeepMind's mapping of DNA mutations, but it does not indicate the movements of systems and institutions in Japan.

Remarks from VIPs

No applicable information.