News | Artificial Intelligence | August 26, 2022

With the increasing use of artificial intelligence (AI) in radiology, it is critical to minimize bias within machine learning systems before implementing their use in real-world clinical scenarios, according to a special report published in the journal Radiology: Artificial Intelligence, a journal of the Radiological Society of North America (RSNA). 

Example of how improper feature removal from imaging data may lead to bias. (A) Chest radiograph in a male patient with pneumonia. (B) Segmentation mask for the lung, generated using a deep learning model. (C) Chest radiograph is cropped based on the segmentation mask. If the cropped chest radiograph is fed to a subsequent classifier for detecting consolidations, the consolidation that is located behind the heart will be missed (arrow, A). This occurs because primary feature removal using the segmentation m

Example of how improper feature removal from imaging data may lead to bias. (A) Chest radiograph in a male patient with pneumonia. (B) Segmentation mask for the lung, generated using a deep learning model. (C) Chest radiograph is cropped based on the segmentation mask. If the cropped chest radiograph is fed to a subsequent classifier for detecting consolidations, the consolidation that is located behind the heart will be missed (arrow, A). This occurs because primary feature removal using the segmentation model was not valid and unnecessarily removed the portion of the lung located behind the heart. 

https://doi.org/10.1148/ryai.210290 © RSNA 2022 


August 26, 2022 — With the increasing use of artificial intelligence (AI) in radiology, it is critical to minimize bias within machine learning systems before implementing their use in real-world clinical scenarios, according to a special report published Wednesday, August 24, in the journal Radiology: Artificial Intelligence, a journal of the Radiological Society of North America (RSNA). 

The report, the first in a three-part series, outlines the suboptimal practices used in the data handling phase of machine learning system development and presents strategies to mitigate them. 

“There are 12 suboptimal practices that happen during the data handling phase of developing a machine learning system, each of which can predispose the system to bias,” said Bradley J. Erickson, M.D., Ph.D., professor of radiology and director of the AI Lab at the Mayo Clinic, in Rochester, Minnesota. “If these systematic biases are unrecognized or not accurately quantified, suboptimal results will ensue, limiting the application of AI to real-world scenarios.” 

Dr. Erickson said the topic of proper data handling is gaining more attention, yet guidelines on the correct management of big data are scarce. 

“Regulatory challenges and translational gaps still hinder the implementation of machine learning in real-world clinical scenarios. However, we expect the exponential growth in radiology AI systems to accelerate the removal of these barriers,” Dr. Erickson said. “To prepare machine learning systems for adoption and clinical implementation, it’s critical that we minimize bias.” 

Within the report, Dr. Erickson and his team suggest mitigation strategies for the 12 suboptimal practices that occur within the four data handling steps of machine learning system development (three for each data handling step), including: 

Data collection—improper identification of the data set, single source of data, unreliable source of data 

Data investigation—inadequate exploratory data analysis, exploratory data analysis with no domain expertise, failing to observe actual data 

Data splitting—leakage between data sets, unrepresentative data sets, overfitting to hyperparameters 

Data engineering—improper feature removal, improper feature rescaling, mismanagement of missing data 

Dr. Erickson said medical data is often far from ideally suited as input for machine learning algorithms. 

“Each of these steps could be prone to systematic or random biases,” he said. “It’s the responsibility of developers to accurately handle data in challenging scenarios like data sampling, de-identification, annotation, labeling, and managing missing values.” 

According to the report, careful planning before data collection should include an in-depth review of clinical and technical literature and collaboration with data science expertise. 

“Multidisciplinary machine learning teams should have members or leaders with both data science and domain (clinical) expertise,” he said. 

To develop a more heterogeneous training dataset, Dr. Erickson and his coauthors suggest collecting data from multiple institutions from different geographical locations, using data from different vendors and from different times, or including public datasets. 

“Creating a robust machine learning system requires researchers to do detective work and look for ways in which the data may be fooling you,” he said. “Before you put data into the training module, you must analyze it to ensure it’s reflective of your target population. AI won’t do it for you.” 

Dr. Erickson said that even after excellent data handling, machine learning systems can still be prone to significant biases. The second and third reports in the Radiology: Artificial Intelligence series focus on biases that occur in the model development and model evaluation and reporting phases. 

“In recent years, machine learning has demonstrated its utility in many clinical research areas, from reconstructing images and hypothesis testing to improving diagnostic, prognostic and monitoring tools,” Dr. Erickson said. “This series of reports aims to identify erroneous practices during machine learning development and mitigate as many of them as possible.” 

For more information: www.rsna.org 


Related Content

News | Point-of-Care Ultrasound (POCUS)

September 27, 2022 — Fujifilm Sonosite, Inc. – the world leader in point-of-care ultrasound (POCUS) solutions – has ...

Time September 27, 2022
arrow
Feature | Information Technology | By Jef Williams

I took my family to Blockbuster this weekend to pick out a movie, hoping there would be something good left to rent ...

Time September 21, 2022
arrow
Feature | Breast Imaging | By Christine Book

Findings from a clinical trial that used artificial intelligence (AI) in an effort to reduce false positives on breast ...

Time September 21, 2022
arrow
Feature | Digital Pathology | By Michael Valante

With digital pathology now being adopted by institutions around the globe, it is not surprising that we are seeing a ...

Time September 21, 2022
arrow
News | Coronavirus (COVID-19)

September 20, 2022 — Children and adolescents who have either recovered from COVID-19 or have long COVID show persistent ...

Time September 20, 2022
arrow
News | Digital Radiography (DR)

September 16, 2022 — 20/20 Imaging, a division of Konica Minolta Healthcare Americas, Inc., announces the FDA clearance ...

Time September 16, 2022
arrow
Feature | PACS | By Melinda Taschetta-Millane

According to a new report, “Global Picture Archiving and Communication System (PACS) Market Report and Forecast 2022 ...

Time September 16, 2022
arrow
Feature | Artificial Intelligence | By Christine Book

What follows is the second part of our coverage of the “Radiology: Artificial Intelligence Fireside Chat” conducted at ...

Time September 16, 2022
arrow
Sponsored Content | Whitepapers | Information Technology

Large-scale mergers and acquisitions are causing increased strain on IT teams responsible for consolidating and managing ...

Time September 15, 2022
arrow
News | Digital Radiography (DR)

September 14, 2022 — A revolutionary mobile imaging system that delivered the industry’s first collapsible column is ...

Time September 14, 2022
arrow
Subscribe Now