Global healthcare fairness: We should be sharing more, not less, data DOI Creative Commons
Kenneth P. Seastedt, Patrick Schwab,

Zach O’Brien

et al.

PLOS Digital Health, Journal Year: 2022, Volume and Issue: 1(10), P. e0000102 - e0000102

Published: Oct. 6, 2022

The availability of large, deidentified health datasets has enabled significant innovation in using machine learning (ML) to better understand patients and their diseases. However, questions remain regarding the true privacy this data, patient control over how we regulate data sharing a way that does not encumber progress or further potentiate biases for underrepresented populations. After reviewing literature on potential reidentifications publicly available datasets, argue cost—measured terms access future medical innovations clinical software—of slowing ML is too great limit through large databases concerns imperfect anonymization. This cost especially developing countries where barriers preventing inclusion such will continue rise, excluding these populations increasing existing favor high-income countries. Preventing artificial intelligence’s towards precision medicine sliding back practice dogma may pose larger threat than reidentification within datasets. While risk should be minimized, believe never zero, society determine an acceptable threshold below which can occur—for benefit global knowledge system.

Language: Английский

Deep learning in cancer genomics and histopathology DOI Creative Commons
Michaela Unger, Jakob Nikolas Kather

Genome Medicine, Journal Year: 2024, Volume and Issue: 16(1)

Published: March 27, 2024

Abstract Histopathology and genomic profiling are cornerstones of precision oncology routinely obtained for patients with cancer. Traditionally, histopathology slides manually reviewed by highly trained pathologists. Genomic data, on the other hand, is evaluated engineered computational pipelines. In both applications, advent modern artificial intelligence methods, specifically machine learning (ML) deep (DL), have opened up a fundamentally new way extracting actionable insights from raw which could augment potentially replace some aspects traditional evaluation workflows. this review, we summarize current emerging applications DL in genomics, including basic diagnostic as well advanced prognostic tasks. Based growing body evidence, suggest that be groundwork kind workflow cancer research. However, also point out models can biases flaws users healthcare research need to know about, propose ways address them.

Language: Английский

Citations

25
Demographic bias in misdiagnosis by computational pathology models DOI
Anurag Vaidya, Richard J. Chen, Drew F. K. Williamson

et al.

Nature Medicine, Journal Year: 2024, Volume and Issue: 30(4), P. 1174 - 1190

Published: April 1, 2024

Language: Английский

Citations

22
Artificial intelligence in drug development DOI
Kang Zhang, Xin Yang, Yifei Wang

et al.

Nature Medicine, Journal Year: 2025, Volume and Issue: 31(1), P. 45 - 59

Published: Jan. 1, 2025

Language: Английский

Citations

18
Deep learning identifies inflamed fat as a risk factor for lymph node metastasis in early colorectal cancer DOI
Scarlet Brockmoeller, Amelie Echle, Narmin Ghaffari Laleh

et al.

The Journal of Pathology, Journal Year: 2021, Volume and Issue: 256(3), P. 269 - 281

Published: Nov. 5, 2021

Abstract The spread of early‐stage (T1 and T2) adenocarcinomas to locoregional lymph nodes is a key event in disease progression colorectal cancer (CRC). cellular mechanisms behind this are not completely understood existing predictive biomarkers imperfect. Here, we used an end‐to‐end deep learning algorithm identify risk factors for node metastasis (LNM) status digitized histopathology slides the primary CRC its surrounding tissue. In two large population‐based cohorts, show that system can predict presence more than one LNM pT2 patients with area under receiver operating curve (AUROC) 0.733 (0.67–0.758) any AUROC 0.711 (0.597–0.797). Similarly, pT1 patients, or was predictable (0.644–0.778) 0.567 (0.542–0.597), respectively. Based on these findings, guide human pathology experts towards highly regions whole slide images. This hybrid observer approach identified inflamed adipose tissue as highest feature presence. Our study first proof concept artificial intelligence (AI) systems may be able discover potentially new biological progression. publicly available biomarker discovery setting. © 2021 Pathological Society Great Britain Ireland. Published by John Wiley & Sons, Ltd.

Language: Английский

Citations

65
Global healthcare fairness: We should be sharing more, not less, data DOI Creative Commons
Kenneth P. Seastedt, Patrick Schwab,

Zach O’Brien

et al.

PLOS Digital Health, Journal Year: 2022, Volume and Issue: 1(10), P. e0000102 - e0000102

Published: Oct. 6, 2022

The availability of large, deidentified health datasets has enabled significant innovation in using machine learning (ML) to better understand patients and their diseases. However, questions remain regarding the true privacy this data, patient control over how we regulate data sharing a way that does not encumber progress or further potentiate biases for underrepresented populations. After reviewing literature on potential reidentifications publicly available datasets, argue cost—measured terms access future medical innovations clinical software—of slowing ML is too great limit through large databases concerns imperfect anonymization. This cost especially developing countries where barriers preventing inclusion such will continue rise, excluding these populations increasing existing favor high-income countries. Preventing artificial intelligence’s towards precision medicine sliding back practice dogma may pose larger threat than reidentification within datasets. While risk should be minimized, believe never zero, society determine an acceptable threshold below which can occur—for benefit global knowledge system.

Language: Английский

Citations

55