This is a republication of the abstract of the paper below, with the title above, preceded by the Key Messages, by the editor of the blog.
Assessing Methods and Tools to Improve Reporting, Increase Transparency, and Reduce Failures in Machine Learning Applications in Health Care
RSNA — Radiology, Artificial Intelligence
Christian Garbin, Oge Marques
Jan 26 2022
Key messages
by Joaquim Cardoso MSc
Digital Health and AI . Institute
March 30, 2022
What is the problem?
- Despite the remarkable progress (of AI applications), there are several examples of unfulfilled promises and outright failures.
- Therefore, datasets and models cannot be inspected in the same, direct way as traditional software products. Other methods are needed to detect failures in ML products.
How the study was conducted?
- This report investigates recent advancements that promote auditing, supported by transparency, as a mechanism to detect potential failures in ML products for health care applications.
- It reviews practices that apply to the early stages of the ML lifecycle, when datasets and models are created; these stages are unique to ML products.
What is the recommendation?
- Checklists can help increase the transparency of crucial stages of machine learning development, leading to early identification of issues that might impact the resulting products’ performance in real-life conditions.
Abstract
Checklists can help increase the transparency of crucial stages of machine learning development, leading to early identification of issues that might impact the resulting products’ performance in real-life conditions.
Artificial intelligence applications for health care have come a long way.
Despite the remarkable progress, there are several examples of unfulfilled promises and outright failures.
- There is still a struggle to translate successful research into successful real-world applications.
- Machine learning (ML) products diverge from traditional software products in fundamental ways.
- Particularly, the main component of an ML solution is not a specific piece of code that is written for a specific purpose; rather, it is a generic piece of code, a model, customized by a training process driven by hyperparameters and a dataset.
- Datasets are usually large, and models are opaque.
Despite the remarkable progress, there are several examples of unfulfilled promises and outright failures.
Therefore, datasets and models cannot be inspected in the same, direct way as traditional software products. Other methods are needed to detect failures in ML products.
This report investigates recent advancements that promote auditing, supported by transparency, as a mechanism to detect potential failures in ML products for health care applications.
It reviews practices that apply to the early stages of the ML lifecycle, when datasets and models are created; these stages are unique to ML products.
Therefore, datasets and models cannot be inspected in the same, direct way as traditional software products. Other methods are needed to detect failures in ML products.
Concretely, this report demonstrates how two recently proposed checklists, datasheets for datasets and model cards, can be adopted to increase the transparency of crucial stages of the ML lifecycle, using ChestX-ray8 and CheXNet as examples.
Concretely, this report demonstrates how two recently proposed checklists, datasheets for datasets and model cards, can be adopted to increase the transparency of crucial stages of the ML lifecycle, using ChestX-ray8 and CheXNet as examples.
The adoption of checklists to document the strengths, limitations, and applications of datasets and models in a structured format leads to increased transparency, allowing early detection of potential problems and opportunities for improvement.
The adoption of checklists to document the strengths, limitations, and applications of datasets and models in a structured format leads to increased transparency, allowing early detection of potential problems and opportunities for improvement.
Keywords: Artificial Intelligence, Machine Learning, Lifecycle, Auditing, Transparency, Failures, Datasheets, Datasets, Model Cards
About the authors
Christian Garbin, Oge Marques
From the College of Engineering & Computer Science, Florida Atlantic University, 777 Glades Rd, EE441, Boca Raton, FL 33431–0991.
Originally published at https://pubs.rsna.org
