Roxana Daneshjou, Kailas Vodrahalli, Weixin Liang, Roberto A Novoa, Melissa Jenkins, Veronica Rotemberg, Justin Ko, Susan M Swetter, Elizabeth E Bailey, Olivier… 展开作者
Roxana Daneshjou, Kailas Vodrahalli, Weixin Liang, Roberto A Novoa, Melissa Jenkins, Veronica Rotemberg, Justin Ko, Susan M Swetter, Elizabeth E Bailey, Olivier Gevaert, Pritam Mukherjee, Michelle Phung, Kiana Yekrang, Bradley Fong, Rachna Sahasrabudhe, James Zou, Albert Chiou
备注：
Machine Learning for Health (ML4H) - Extended Abstract
More than 3 billion people lack access to care for skin disease. AI diagnostic tools may aid in early skin cancer detection; however most models have not been assessed on images of diverse skin tones or uncommon diseases. To address this, we curated the Diverse Dermatology Images (DDI) dataset - the first publicly available, pathologically confirmed images featuring diverse skin tones. We show that state-of-the-art dermatology AI models perform substantially worse on DDI, with ROC-AUC dropping 29-40 percent compared to the models' original results. We find that dark skin tones and uncommon diseases, which are well represented in the DDI dataset, lead to performance drop-offs. Additionally, we show that state-of-the-art robust training methods cannot correct for these biases without diverse training data. Our findings identify important weaknesses and biases in dermatology AI that need to be addressed to ensure reliable application to diverse patients and across all disease.
Stanford School of Medicine（斯坦福大学医学院）；Stanford University（斯坦福大学）；Memorial Sloan Kettering Cancer Center（纪念斯隆凯特琳癌症中心）；Keio University School of Medicine（庆应义塾大学医学院）；Chan-Zuckerberg Biohub（陈-扎克伯格生物中心）
Access to dermatological care is a major issue, with an estimated 3 billion people lacking access to care globally. Artificial intelligence (AI) may aid in triaging skin diseases. However, most AI models have not been rigorously assessed on images of diverse skin tones or uncommon diseases. To ascertain potential biases in algorithm performance in this context, we curated the Diverse Dermatology Images (DDI) dataset-the first publicly available, expertly curated, and pathologically confirmed image dataset with diverse skin tones. Using this dataset of 656 images, we show that state-of-the-art …
How does the accuracy of deep neural network models trained to classify clinical images of skin conditions vary across skin color? While recent studies demonstrate computer vision models can serve as a useful decision support tool in healthcare and provide dermatologist-level classification on a number of specific tasks, darker skin is underrepresented in the data. Most publicly available data sets do not include Fitzpatrick skin type labels. We annotate 16,577 clinical images sourced from two dermatology atlases with Fitzpatrick skin type labels and open-source these annotations. Based on the…
DermaSynth: Rich Synthetic Image-Text Pairs Using Open Access Dermatology Datasets
Imperial College London（帝国理工学院）；Yildiz Technical University（伊尔德兹技术大学）；Usak Research and Training Hospital（乌萨克研究与培训医院）；Istanbul Medeniyet University（伊斯坦布尔梅德尼耶特大学）；Universität der Bundeswehr München（慕尼黑联邦国防军大学）；Istanbul Medicana Atakoy Hospital（伊斯坦布尔 Medicana Atakoy 医院）
DermETAS-SNA LLM: A Dermatology Focused Evolutionary Transformer Architecture Search with StackNet Augmented LLM Assistant
Data Science Program, Worcester Polytechnic Institute, Massachusetts, USA（沃斯特理工学院数据科学项目，马萨诸塞州，美国）；Computer Science Department, Worcester Polytechnic Institute, Massachusetts, USA（沃斯特理工学院计算机科学系，马萨诸塞州，美国）；PASE Advisory Group, New Jersey, USA（PASE顾问小组，新泽西州，美国）
DermINO: Hybrid Pretraining for a Versatile Dermatology Foundation Model
Microsoft Research Asia（微软亚洲研究院）；Peking University China-Japan Friendship School of Clinical Medicine（北京大学中国日友好临床医学学院）；Scientific-skincare Innovation Alliance (SIA)（科学护肤创新联盟（SIA））
DermaCon-IN: A Multi-concept Annotated Dermatological Image Dataset of Indian Skin Disorders for Clinical AI Research
Department of Artificial Intelligence, Indian Institute of Technology Hyderabad, India（人工智能系，印度理工学院海得拉巴学院，印度）；Indian Navy（印度海军）；Department of Dermatology, S R Patil Medical College, India（皮肤科系，S R Patil医学院，印度）；Department of Dermatology, S Nijalingappa Medical College, India（皮肤科系，S Nijalingappa医学院，印度）；Department of Dermatology, Sri Chamundeshwari Medical College, Hospital & Research, India（皮肤科系，Sri Chamundeshwari医学院，医院及研究，印度）；Indian Institute of Technology Hyderabad, India（印度理工学院海得拉巴学院，印度）
For the deployment of artificial intelligence (AI) in high-risk settings, such as healthcare, methods that provide interpretability/explainability or allow fine-grained error analysis are critical. Many recent methods for interpretability/explainability and fine-grained error analysis use concepts, which are meta-labels that are semantically meaningful to humans. However, there are only a few datasets that include concept-level meta-labels and most of these meta-labels are relevant for natural images that do not require domain expertise. Densely annotated datasets in medicine focused on meta-l…
Carnegie Mellon University Africa（卡内基梅隆大学非洲分校）；IBM Research（IBM研究院）
Recent advances in computer vision and deep learning have led to breakthroughs in the development of automated skin image analysis. In particular, skin cancer classification models have achieved performance higher than trained expert dermatologists. However, no attempt has been made to evaluate the consistency in performance of machine learning models across populations with varying skin tones. In this paper, we present an approach to estimate skin tone in benchmark skin disease datasets, and investigate whether model performance is dependent on this measure. Specifically, we use individual ty…
Stanford School of Medicine（斯坦福大学医学院）；University of Pennsylvania（宾夕法尼亚大学）
Artificial intelligence (AI) algorithms using deep learning have advanced the classification of skin disease images; however these algorithms have been mostly applied "in silico" and not validated clinically. Most dermatology AI algorithms perform binary classification tasks (e.g. malignancy versus benign lesions), but this task is not representative of dermatologists' diagnostic range. The American Academy of Dermatology Task Force on Augmented Intelligence published a position statement emphasizing the importance of clinical validation to create human-computer synergy, termed augmented intel…
SkinCaRe: A Multimodal Dermatology Dataset Annotated with Medical Caption and Chain-of-Thought Reasoning
School of Data Science, The Chinese University of Hong Kong, Shenzhen (CUHK–Shenzhen)（数据科学学院，香港中文大学（深圳））；Computer Science Program, CEMSE Division, King Abdullah University of Science and Technology (KAUST)（计算机科学项目，科学与工程学院，国王 Abdullah 科学技术大学）；Center of Excellence on Smart Health, KAUST（智能健康卓越中心，国王 Abdullah 科学技术大学）；Center of Excellence for Generative AI, KAUST（生成式人工智能卓越中心，国王 Abdullah 科学技术大学）；Department of Dermatology, Beijing AnZhen Hospital, Capital Medical University（皮肤科，北京安贞医院，首都医科大学）；Department of Dermatology, Tianjin Institute of Integrative Dermatology, Tianjin Academy of Traditional Chinese Medicine Affiliated Hospital（皮肤科，天津整合皮肤科研究院，天津中医药大学附属医院）；Department of Dermatology, Beijing Aerospace General Hospital（皮肤科，北京航天总医院）；Department of Dermatology, The First Affiliated Hospital, Shantou University Medical College（皮肤科，汕头大学医学院第一附属医院）；DermAssure, LLC（DermAssure 公司）；School of Medicine, New York Medical College（医学院，纽约医学院）；Capital Medical University（首都医科大学）；Department of Dermatology, Second Hospital of Jilin University（皮肤科，吉林大学第二医院）；Emergency Critical Care Center, Beijing AnZhen Hospital, Capital Medical University（急诊重症中心，北京安贞医院，首都医科大学）
An analysis of data variation and bias in image-based dermatological datasets for machine learning classification
Centro de Informática, Universidade Federal de Pernambuco (UFPE)（伯南布哥联邦大学信息中心）；Ambulatório de Dermatologia, Hospital das Clinicas - UFPE（伯南布哥联邦大学临床医院皮肤科门诊）
SynPAIN: A Synthetic Dataset of Pain and Non-Pain Facial Expressions
KITE Research Institute, Toronto Rehabilitation Institute, University Health Network（KITE研究 institute、多伦多康复研究所、大学健康网络）；Department of Computer Science, University of Toronto（多伦多大学计算机科学系）；Institute of Biomedical Engineering, University of Toronto（多伦多大学生物医学工程研究所）；Vector Institute（向量研究所）；Department of Occupational Science and Occupational Therapy, University of Toronto（多伦多大学职业科学与职业治疗系）；University of Regina（Regina 多伦多大学）
University of Toronto（多伦多大学）；Vector Institute（向量研究所）；Massachusetts Institute of Technology（麻省理工学院）
Machine learning systems have received much attention recently for their ability to achieve expert-level performance on clinical tasks, particularly in medical imaging. Here, we examine the extent to which state-of-the-art deep learning classifiers trained to yield diagnostic labels from X-ray images are biased with respect to protected attributes. We train convolution neural networks to predict 14 diagnostic labels in 3 prominent public chest X-ray datasets: MIMIC-CXR, Chest-Xray8, CheXpert, as well as a multi-site aggregation of all those datasets. We evaluate the TPR disparity -- the differ…
University of Edinburgh（爱丁堡大学）；Imperial College London（帝国理工学院）；University of Essex（埃塞克斯大学）
Magnetic Resonance Imaging (MRI) typically recruits multiple sequences (defined here as "modalities"). As each modality is designed to offer different anatomical and functional clinical information, there are evident disparities in the imaging content across modalities. Inter- and intra-modality affine and non-rigid image registration is an essential medical image analysis process in clinical imaging, as for example before imaging biomarkers need to be derived and clinically evaluated across different MRI modalities, time phases and slices. Although commonly needed in real clinical scenarios, …
