行業指南

腫瘤學中的人工智慧

AI in oncology uses machine learning to help find tumors on scans and slides, estimate risk, plan radiotherapy and organize information for treatment decisions such as tumor board reviews.

聆聽此指南AI 語音敘述 · ~3 min

本頁閱讀時間3分鐘
  1. 概述
  2. 深入探討
  3. 現實世界的實施
  4. The Future of AI in Oncology
  5. 付諸實踐
  6. 不斷探索
  7. 常見問題

概述

It matters because cancer care involves huge volumes of images and data, but so far the strongest evidence is for detection and workflow tasks, not for AI choosing which treatment a patient should get.

深入探討

Oncologists meet AI in three main places: detection, treatment planning and decision support. Detection is the most mature. In mammography, the Swedish MASAI randomized trial, first reported in 2023, found that AI-supported screening detected more cancers than standard double reading while cutting radiologist screen-reading workload by roughly 44 percent. In pathology, the FDA authorized Paige Prostate in 2021 as the first AI product to assist pathologists reading prostate biopsies. Research models can also estimate features such as microsatellite instability directly from routine stained slides, and MIT and Massachusetts General Hospital researchers built Sybil, which predicts future lung cancer risk from a single low-dose CT scan. Treatment planning is quietly widespread. Auto-contouring tools draft outlines of tumors and nearby organs for radiotherapy, a task that can take clinicians hours by hand. Clinicians still review and edit every contour. Treatment selection is where caution is most needed. IBM's Watson for Oncology was marketed to suggest therapies but faced reports of recommendations that disagreed with experts and some that were unsafe, and IBM later sold its Watson Health business. The lesson was that training on a small set of hypothetical cases and one institution's preferences does not produce reliable general advice. Tumor boards, where surgeons, oncologists, radiologists and pathologists discuss cases together, increasingly use software that assembles data and flags matching clinical trials or guideline pathways. Large language models are being tested to summarize records for these meetings, but hallucinated or outdated facts are a serious risk. The main limit of current evidence is that many studies are retrospective, measure accuracy rather than patient outcomes, and come from a few well-resourced centers. Finding more small cancers is only helpful if it improves survival rather than increasing overdiagnosis.

現實世界的實施

A breast screening program uses AI to triage mammograms so that likely-normal exams need one radiologist while higher-risk exams get two readers, cutting reading workload.

A radiotherapy team uses auto-contouring software to draw first-draft outlines of organs at risk on CT scans, which a dosimetrist and oncologist then check and correct.

A pathology lab uses an AI system to highlight suspicious areas on digitized prostate biopsy slides so the pathologist reviews those regions carefully.

A hospital's tumor board software gathers imaging, pathology, genomics and treatment history into one dashboard before the weekly multidisciplinary meeting.

The Future of AI in Oncology

Near-term growth is most likely in screening, pathology and radiotherapy workflow, where tasks are well defined and results can be checked. Longer follow-up from screening trials will clarify whether extra detections reduce interval cancers and deaths or add overdiagnosis. Multimodal models that combine images, genomics and records are an active research area, but they will need prospective trials before guiding treatment choices. Tumor boards may use AI summaries and trial matching more often, with clinicians verifying sources. Equity is a real concern, since many models are developed at large academic centers whose patients and equipment may not match community hospitals.

付諸實踐

為什麼這很重要

  • 背景與規則. 產業背景決定了人工智慧創意能否與現實接觸。
  • 品質管控. 領域約束會影響可接受的錯誤率和監督模型。
  • 配裝選擇. 成功的部署使技術能力與第一線工作流程保持一致。

留意

  • 監理要求可能會使原本強大的原型失效。
  • 歷史資料可能會編碼損害特定社區的偏見。
  • 遺留系統可能會造成整合瓶頸和隱性成本。

後續步驟

  1. 讓領域專家參與從問題框架到評估的整個過程。
  2. 在啟動前設計審計追蹤和文件。
  3. 儘早驗證合規性和安全義務。
  4. 分階段推出,並有明確的停止和回滾標準。

不斷探索

免費電子報

獲取每日 AI 簡報

每個工作日上午提供三則經過驗證的 AI 新聞,以淺顯易懂的英文撰寫。永久免費,無廣告。

每個工作日一封電子郵件。一鍵取消訂閱。我們絕不出售或分享您的地址。

測試一下自己

Take the AI in Oncology quiz

每個答案皆有即時回饋,通過課程後可獲得具備可驗證 ID 的可分享證書。

開始測驗

支持免費的 AI 教育。AI Understanding 是一個 501(c)(3) 非營利組織——絕無廣告,絕無付費牆。 捐款

常見問題

What is AI in Oncology?

AI in oncology uses machine learning to help find tumors on scans and slides, estimate risk, plan radiotherapy and organize information for treatment decisions such as tumor board reviews. It matters because cancer care involves huge volumes of images and data, but so far the strongest evidence is for detection and workflow tasks, not for AI choosing which treatment a patient should get.

瑞典 MASAI 試驗關於人工智慧支援的乳房 X 光檢查報告了什麼?

MASAI 是一項於 2023 年首次報告的隨機試驗,發現與標準雙讀相比,檢測率更高,放射科醫生的讀片工作量大幅下降。

IBM Watson for Oncology 的重要教訓是什麼?

與專家意見不一致的建議報告,其中一些建議不安全,顯示了其培訓方法的局限性。 IBM 後來出售了 Watson Health 業務。

FDA 於 2021 年授權哪種產品作為第一個協助病理學家進行前列腺活檢的人工智慧產品?

Paige Prostate 在數位化前列腺活檢幻燈片上突出顯示可疑區域,以供病理學家審查。

麻省理工學院和麻薩諸塞州總醫院的西比爾有何預測?

Sybil 使用一張低劑量胸部 CT 來估計一個人未來肺癌的風險。

大多数人工智能系统如何处理十亿像素全切片病理图像?

幻灯片太大而无法一次处理,因此对补丁进行编码和组合,通常会产生影响区域的热图。