本文へ
MMadical AI Use
ベンダー報告ベンダー

Quantitative assist for immunohistochemistry scoring

この言語の本文は未訳のため、英語で表示しています。

Vendors including PathAI and Ibex offer algorithms that estimate staining extent, such as a Ki-67 or PD-L1 percentage, on a digitized slide. A pathologist checks the region and the number before it enters the report. This card does not restate concordance percentages, which are assay-specific and easy to over-read. The signed score remains a pathology result, not a model result.

臨床の場

Immunohistochemistry sign-out in a digital pathology lab, where the stain already has a clinical scoring scheme the tumor board expects.

ツールとベンダー

  • PathAI
  • Ibex

出典が報告したこと

No concordance figure is copied here. Treat vendor white papers as vendor-reported until an independent study matches the assay you actually run.

出典

PathAI · 2025年2月1日 · 年 2025

索引への追加 2026年7月2日

出典を開く

タグ

  • 病理
  • 人が介在
  • コンピュータビジョン

教育と市場把握のための条目。医療助言でも推奨でもありません。

関連シナリオ

この領域のすべて

病理

規制

Adjunct detection on prostate biopsy whole-slide images

Paige Prostate marks a location on a digitized prostate biopsy slide that the model ranks as the most suspicious for cancer, after the pathologist’s own first look. FDA authorized it in 2021 as an adjunct. The output is not the diagnosis. The pathologist still signs out, using the full clinical context and any extra stains they choose to order.

Paige Prostate · Paige FullFocus

ニュース · 2021事例を読む

画像と放射線

規制

Autonomous diabetic retinopathy screening in primary care

IDx-DR, now marketed as LumineticsCore by Digital Diagnostics, was authorized to screen adults with diabetes for more-than-mild diabetic retinopathy without an eye specialist reading the photo first. A primary care team captures images on a specified retinal camera and the software returns a refer-or-rescreen result. It is not a complete eye examination. Pregnancy and several existing retinal conditions sit outside the labeled use.

LumineticsCore (IDx-DR) · Digital Diagnostics · Topcon NW400

ニュース · 2018事例を読む

画像と放射線

論文

AI-supported reading in the MASAI mammography trial

The MASAI trial in Sweden randomly assigned screening mammograms to AI-supported reading or standard double reading without AI. Transpara scored each exam and routed lower scores to a single reader and the highest score to double reading, with detection marks available on higher-risk exams. This card quotes the prespecified interim safety analysis, not a claim that every screening program should switch.

Transpara 1.7.0 · ScreenPoint Medical

論文 · 2023事例を読む
スポンサー

このインデックスの臨床読者に届ける。

スポンサー枠を見る