En Es

Industry News

Novel AI Algorithm for Mammography Interpretation Can Successfully Spot Breast Cancer Years Before Radiologists

By Medimaging International staff writers
13 Jan 2021

Image: DeepHealth`s AI identifies cancer in a patient one year earlier than detected in practice (Photo courtesy of DeepHealth)A novel artificial intelligence (AI) algorithm for mammography interpretation has demonstrated the ability to detect breast cancer a year or more earlier than current practice.

DeepHealth (Cambridge, MA, USA), a wholly owned subsidiary of RadNet, Inc. (Los Angeles, CA, USA), compared its AI to five full-time, breast-fellowship-trained expert radiologists reading the same screening mammograms. The software exhibited higher performance than all five radiologists, and the results suggest that the AI could help detect cancer one to two years earlier than standard interpretation in many cases.

Additionally, the software showed promising generalization capabilities, demonstrating strong performance when tested across clinical sites and populations that were not directly involved in training the AI algorithms. While AI holds tremendous promise for improving screening mammography interpretation, there remain substantial challenges in developing expert-level AI. The new study by DeepHealth demonstrates progress in resolving these challenges.

“Reaching world-class performance requires a new way of building AI,” said Gregory Sorensen, M.D., CEO, and co-founder of DeepHealth. “The brute-force methods that have worked so well in other domains, such as self-driving cars or game playing, where data is plentiful, have not translated effectively to many parts of medicine, where human data is often scarce. For example, to train the technology for better detection, AI algorithms must be developed from annotated data where the cancer status is known. Such data can be difficult to obtain. Then, to validate performance, the AI should be tested across different clinical sites and patient populations in different scenarios.”

“We have developed an approach that mimics how humans often learn by progressively training the AI models on more difficult tasks. By leveraging prior information learned in each successive training stage, this strategy results in AI that detects cancer accurately while also relying less on highly-annotated data,” said lead author Bill Lotter, Ph.D., CTO, and co-founder of DeepHealth. “Our approach and validation extend to 3D mammography, which is particularly important given its growing use and the significant challenges it presents for AI.”

Related Links:
RadNet, Inc.

E-mail Print
FaceBook Twitter Google+ Linked in

Additional news

26 Jan 2021
Roche Renews Partnership with Sysmex to Deliver Hematology Testing Solutions
Under the new framework of a Global Business Partnership Agreement (GBP) signed between Roche (Basel, Switzerland) and Sysmex Corporation (Kobe, Japan), both the companies have renewed their commitment to their long standing Distribution, Sales and Service (DSS) agreement, allowing Roche to continue to distribute Sysmex hematology products, including instruments and reagents.
Read More
25 Jan 2021
Nova Biomedical Adds Sample Retain Collector to BioProfile FLEX2 Cell Culture Analyzer
Nova Biomedical (Waltham, MA, USA) has added a Sample Retain Collector (SRC) for its BioProfile FLEX2 automated cell culture analyzer which measures up to 16 tests including pH, gases, metabolites, osmolality, cell density, and cell viability.
Read More
21 Jan 2021
New AI-Based Image Analysis Platform Enables Physicians to Analyze CT and MRI Scans in Half the Time of Traditional Methods
The US Food and Drug Administration (FDA) has granted 510(k) clearance for a new image analysis platform that enables physicians to analyze CT and MRI scans in half the time of traditional methods.
Read More
21 Jan 2021
New AI-Based Medical Image Segmentation and Measurement Tool Minimizes Human Errors in Clinical Analysis
A new general purpose, AI-based medical image segmentation and measurement tool for detecting objects of interest and their boundaries quickly and automatically could make surgical and diagnostic measurements easier and more accurate for better treatment decisions.
Read More
20 Jan 2021
Deep Learning Outperforms Standard Machine Learning at Brain Imaging Analysis, Finds New Study
Compared to standard machine learning models, deep learning models are largely superior at discerning patterns and discriminative features in brain imaging, despite being more complex in their architecture, according to a new study.
Read More
20 Jan 2021
Thermo Fisher Acquires Point-of-Care Molecular Diagnostics Provider Mesa Biotech
Thermo Fisher Scientific Inc. (Waltham, MA, USA) has entered into a definitive agreement to acquire Mesa Biotech (San Diego, CA, USA), a privately held molecular diagnostic company, for approximately USD 450 million in cash.
Read More
20 Jan 2021
Hillrom Acquires Bardy Diagnostics to Expand into Ambulatory Cardiac Monitoring Segment
Hillrom (Chicago, IL, USA) has reached a definitive agreement to acquire Bardy Diagnostics, Inc. (BardyDx; Seattle, WA, USA) for a cash consideration of USD 375 million.
Read More
20 Jan 2021
Philips Acquires Capsule Technologies to Expand Leadership in Patient Care Management Solutions for Hospitals
Royal Philips (Amsterdam, the Netherlands) has signed an agreement to acquire Capsule Technologies, Inc. (Andover, MA, USA), a provider of medical device integration and data technologies for hospitals and healthcare organizations.
Read More
19 Jan 2021
Artificial Intelligence Add-On for `World’s First` Portable MR Imaging System Receives FDA Clearance
Hyperfine Research, Inc. (Guilford, CT, USA) has received 510(k) clearance from the US FDA for its deep-learning image analysis software that measure brain structure and pathology in images acquired by the Swoop Portable MR Imaging System.
Read More
Copyright © 2000-2021 TradeMed.com. All rights reserved. | Terms And Conditions