AI in Medicine — Cancer & Rare Disease — 2026-09-25 (Fri) Evening News

Latest news on AI and cancer treatment

Overview

Mainly reports on the use of AI in the field of image diagnosis (CT/pathological images). In the case of lung cancer, research on predicting the risk of pneumonia associated with immunotherapy from CT images has been reported in multiple media. Although there has been movement in drug discovery and medical treatment infrastructure (EHR integration, molecular exploration), there is little primary information showing the specific progress of approvals and clinical trials.

Breast cancer

No relevant information (the headline "AI will change the way you choose snacks" is food industry news about Conagra and has nothing to do with breast cancer).

lung cancer

- A study conducted by MD Anderson Cancer Center in the US has reported that AI can analyze routinely taken CT images to identify patients at high risk of severe treatment-induced pneumonitis associated with immunotherapy. The same content was also reported on Bioengineer.org and News-Medical. The stage of development and validation (preclinical, single-center retrospective, prospective study, etc.) cannot be determined from the headline. - The NYT editorial, ``AI has made me better doctors, but made me worse medical students,'' is a general medical education theory that is not limited to lung cancer, and does not include findings regarding cancer types.

Malignant brain tumor

No relevant information (“AI improves brain aneurysm detection by 39%” relates to brain aneurysms = vascular disease, which is different from brain tumors).

Leukemia/lymphoma

- Dr. Encinas Mayoral comments on the benefits of the AI-designed BCMA/FcRL5-targeting TRiTE (T-cell inducing bispecific molecule) in OncLive regarding multiple myeloma (a type of blood cancer). The development stage (preclinical/clinical) is unclear from the headline.

Colon cancer/digestive organ

- Regarding pancreatic cancer (gastrointestinal cancer), a report using AI modeling has identified a hidden molecular switch that can activate "cold" tumors (tumors with poor immune response) (Bioengineer.org). The details of the research stage (pre-clinical model, etc.) are unclear from the headline.

Rare cancer

No applicable information.

Diagnosis/Pathology/Image

- Announcement that Leica Biosystems will integrate Tempus AI into the Aperio AI Store. This is a product integration related to a digital pathology platform across cancer types, and performance data is not listed in the headline. - Published a review paper on biomarker discovery using AI (possibility of application to diagnosis and treatment) in Nature magazine. This is a general article, and individual accuracy data is unclear from the headline. - (Reference) "AI improves cerebral aneurysm detection by 39%" is said to be a study based on actual clinical data, but it is not the main subject of this report because it targets cerebrovascular disease rather than cancer.

Treatment selection/drug discovery

- AI model that predicts immunotherapy-related pneumonia risk from CT (News-Medical, same study as lung cancer item). - Medium editorial (by Sergei Polevikov) "AI 0: Cancer 1" is a blog-style article that discusses the results of testing five cutting-edge AI models on cancer-related tasks. It needs to be treated as an individual's analysis and opinion, rather than peer-reviewed research. - Oracle launches AI-powered electronic health record (EHR) for oncology. Product announcements rather than clinical efficacy data. - Bioengineer.org published an article explaining the trend, titled ``AI from a laboratory rarity to a fundamental technology for tumor treatment.'' - The coverage of AbbVie's molecular search AI (search using 6,000 characteristics) relates to the early stage of drug discovery (candidate molecule search) and does not include information on the clinical stage.

Regulations/Approval

No relevant information (The Investing.com article mentions "Japan approval," but the specific product name and approval classification are not listed in the headline, and as the content deals with stock price trends, it is outside the scope of this report.)

Clinical trial implementation status

No applicable information (The information provided this time does not include data on the specific trial name, target cancer type, phase, and registration status from ClinicalTrials.gov).

Japanese movements

- A press release from Report Ocean Co., Ltd. announced that the diagnostic equipment market is predicted to reach US$73.5 billion in 2035 (CAGR 4.8%). The increase in chronic diseases and the spread of high-precision diagnostic equipment through AI and digitalization are cited as growth factors, but these are forecasts for the entire market, not figures specific to cancer diagnosis. - There have been reports of a 24-hour AI chat service called "VetCall" for veterinary hospitals (which responds to inquiries during closed hours), but this is a service in the field of veterinary medicine and has nothing to do with human cancer treatment. - The Investing.com article (with reference to diagnostic-related stocks selected by an AI model and approval in Japan) is about stock prices and investment, so we cannot confirm any concrete support for it as regulatory information.

Remarks from VIPs

- Dr. Encinas Mayoral (Title details not listed in headline, OncLive interview): Mentioned the benefits of AI-designed BCMA/FcRL5-targeted TRiTE in the treatment of multiple myeloma. - Sergei Polevikov (Medium contributor): Commentary on the results of testing five cutting-edge AI models on cancer-related tasks (personal blog post before peer review).

Latest news on AI and rare and incurable diseases

Overview

Only two headlines were provided, and it was not possible to confirm any concrete progress regarding diagnosis, therapeutic drugs, or systems. It concerns the business trends of corn breeding and cancer genome analysis companies, and has little direct connection to the field of rare and incurable diseases.

Diagnosis of rare diseases

No applicable information.

Gene analysis/genome

There are reports that Tempus AI is expanding its whole genome sequencing business, but the headline only mentions business evaluation, and does not indicate specific clinical results such as application to rare disease diagnosis or detection rate/diagnosis rate.

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

No applicable information.

Remarks from VIPs

No applicable information.

Furthermore, ``Agriculture companies collaborate to advance climate-resilient corn using AI and gene editing'' was excluded from the analysis because it is related to the agricultural field (breeding of climate-resistant corn) and has nothing to do with rare and intractable human diseases.