東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁 到查詢結果 [ null ]

切換: 標籤 | MARC模式 | ISBD

Augmenting Medical Image Classifiers...

Sagers, Luke William,

FindBook

Google Book

Amazon

博客來

Augmenting Medical Image Classifiers With Synthetic Data Across Populations /

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Augmenting Medical Image Classifiers With Synthetic Data Across Populations // Luke William Sagers.
作者:	Sagers, Luke William,
面頁冊數:	1 electronic resource (125 pages)
附註:	Source: Dissertations Abstracts International, Volume: 85-12, Section: B.
Contained By:	Dissertations Abstracts International85-12B.
標題:	Bioinformatics. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31296447
ISBN:	9798382776552

Augmenting Medical Image Classifiers With Synthetic Data Across Populations /
Sagers, Luke William,

Augmenting Medical Image Classifiers With Synthetic Data Across Populations /Luke William Sagers. - 1 electronic resource (125 pages)

Source: Dissertations Abstracts International, Volume: 85-12, Section: B.

Rapid improvements in the capabilities of generative artificial intelligence (AI) models to produce and interpret images have created new possibilities to address persistent challenges in medical machine learning including data scarcity, annotation costs, and model biases. This dissertation primarily explores the potential opportunities and limitations of using synthetic images created by large generative AI models to improve the performance and generalizability of medical image classifiers across populations. We also evaluate the ability of consumer-facing vision-language models to classify dermatology and chest X-ray images under different prompts.We designed an image-generation pipeline by fine-tuning diffusion-based models to create synthetic skin disease images using both generative fill and text-to-image methods. With this pipeline, we generated 500,000 synthetic dermatology images (which we publicly released for future research) representing 12 diseases across diverse skin tones. We then systematically evaluated the performance of AI disease classifiers when including or excluding synthetic images in model training. We show that in data-limited settings (in which there are few real images of a disease or skin-tone), synthetic data can improve classifier performance, but that these gains saturate once sufficient quantities of real images are available. We find that the biggest driver of model improvements is the quantity of real images. We also observed a correlation between the physician-assessed photorealism of synthetic images and gains in model performance. Collectively, these findings suggest that synthetic data presents a complementary tool to training disease classifiers and can be useful as an advanced augmentation method or a way to share features of a data distribution without sharing the data itself. However, efforts must still be focused on collecting more high quality, diverse, real data to train the next generation of fair, robust, and generalizable AI systems. We also evaluated the capabilities of consumer-facing and general-purpose vision-language AI models in interpreting chest X-rays and dermatology images. We found that these systems, which have not been specifically trained for medical image diagnoses, can perform at or near-human level on selected metrics, and that model performance and behavior can be influenced using the text prompt and task formulation. Our analysis suggests that evaluations of large language and vision-language models should carefully consider the prompt context and other inputs. Overall, this dissertation provides a systematic analysis of the opportunities, pitfalls, and open challenges regarding the use of synthetic data and generative AI for improving medical imaging across all populations.

English

ISBN: 9798382776552Subjects--Topical Terms:

553671
Bioinformatics.
Subjects--Index Terms:

Diffusion models

Augmenting Medical Image Classifiers With Synthetic Data Across Populations /
LDR:04333nmm a22004693i 4500 001 2432208
005 20250707081320.5
006 m o d
007 cr|nu||||||||
008 261011s2024 miu||||||m |||||||eng d
020 $a 9798382776552
035 $a (MiAaPQD)AAI31296447
035 $a AAI31296447
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Sagers, Luke William, $e author. $0 (orcid)0000-0002-5024-7314 $3 3821355
245 1 0 $a Augmenting Medical Image Classifiers With Synthetic Data Across Populations / $c Luke William Sagers.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (125 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-12, Section: B.
500 $a Advisors: Manrai, Arjun K. Committee members: Yu, Kun-Hsing; Lotter, William; Das, Shinjita; Plummer, Bryan.
502 $b Ph.D. $c Harvard University $d 2024.
520 $a Rapid improvements in the capabilities of generative artificial intelligence (AI) models to produce and interpret images have created new possibilities to address persistent challenges in medical machine learning including data scarcity, annotation costs, and model biases. This dissertation primarily explores the potential opportunities and limitations of using synthetic images created by large generative AI models to improve the performance and generalizability of medical image classifiers across populations. We also evaluate the ability of consumer-facing vision-language models to classify dermatology and chest X-ray images under different prompts.We designed an image-generation pipeline by fine-tuning diffusion-based models to create synthetic skin disease images using both generative fill and text-to-image methods. With this pipeline, we generated 500,000 synthetic dermatology images (which we publicly released for future research) representing 12 diseases across diverse skin tones. We then systematically evaluated the performance of AI disease classifiers when including or excluding synthetic images in model training. We show that in data-limited settings (in which there are few real images of a disease or skin-tone), synthetic data can improve classifier performance, but that these gains saturate once sufficient quantities of real images are available. We find that the biggest driver of model improvements is the quantity of real images. We also observed a correlation between the physician-assessed photorealism of synthetic images and gains in model performance. Collectively, these findings suggest that synthetic data presents a complementary tool to training disease classifiers and can be useful as an advanced augmentation method or a way to share features of a data distribution without sharing the data itself. However, efforts must still be focused on collecting more high quality, diverse, real data to train the next generation of fair, robust, and generalizable AI systems. We also evaluated the capabilities of consumer-facing and general-purpose vision-language AI models in interpreting chest X-rays and dermatology images. We found that these systems, which have not been specifically trained for medical image diagnoses, can perform at or near-human level on selected metrics, and that model performance and behavior can be influenced using the text prompt and task formulation. Our analysis suggests that evaluations of large language and vision-language models should carefully consider the prompt context and other inputs. Overall, this dissertation provides a systematic analysis of the opportunities, pitfalls, and open challenges regarding the use of synthetic data and generative AI for improving medical imaging across all populations.
546 $a English
590 $a School code: 0084
650 4 $a Bioinformatics. $3 553671
650 4 $a Dermatology. $3 829009
650 4 $a Medicine. $3 641104
650 4 $a Medical imaging. $3 3172799
653 $a Diffusion models
653 $a Generative AI
653 $a Image classification
653 $a Synthetic data
653 $a Vision-language models
653 $a Skin disease images
690 $a 0715
690 $a 0574
690 $a 0564
690 $a 0800
690 $a 0757
710 2 $a Harvard University. $b Medical Sciences. $e degree granting institution. $3 3815464
720 1 $a Manrai, Arjun K. $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 85-12B.
790 $a 0084
791 $a Ph.D.
792 $a 2024
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31296447