Izotropic’s AI Breakthrough Positions IzoView to Redefine Breast CT Imaging Standards
- Proprietary algorithm positions IzoView to redefine global standards and expectations for image quality and safety in breast CT -
- Trained on 15 years of specialized breast CT data and protected as a trade secret, giving Izotropic a durable competitive edge -
- AI enhances IzoView’s image quality by reducing CT image noise without increasing radiation dose -
VANCOUVER, British Columbia, and SACRAMENTO, Calif., Aug. 28, 2025 (GLOBE NEWSWIRE) -- via IBN – Izotropic Corporation (CSE: IZO) (OTCQB: IZOZF) (FSE: 1R3) (“Izotropic”, or the “Company”), a medical device company commercializing innovative, emerging technologies and imaging-based products for the more accurate screening, diagnoses, and treatment of breast cancers, is pleased to announce the integration of its proprietary AI-based machine-learning reconstruction algorithm into its flagship device, the IzoView Breast CT Imaging System (“IzoView”).
The Algorithm
Artificial intelligence is reshaping medical imaging, and Izotropic is leading that shift with breast CT. In collaboration with The Johns Hopkins University School of Medicine, the Company has developed a proprietary machine-learning reconstruction algorithm designed to further improve IzoView’s image quality while maintaining low radiation doses, without the constraints that limit other AI methods.
Noise in CT Imaging
All CT imaging, including breast CT, naturally exhibits image noise, seen as a grainy texture in the reconstructed image. The less radiation used, the more visible image noise can become. Since CT systems are designed to minimize dose for patient safety, there is a trade-off between image quality and radiation exposure. Image noise is recognized as an inherent characteristic of all CT imaging and a key area of opportunity for technical innovation.
Existing AI Denoising Methods
Most traditional post-image-processing denoising techniques offer limited noise suppression and have led to the development of two modern denoising methods: Model-Based Iterative Reconstruction (“MBIR”) and Deep Machine-Learning Reconstruction (“DMLR”), both of which typically work to clean up image noise after the image has been reconstructed. While both have advanced the field, each comes with limitations that render them impractical for routine clinical workflows and applications where time and throughput are critical, as is the case in breast cancer screening.
MBIR improves image quality by simulating how X-rays travel through the body and refines the image through repeated calculations. While it can reduce noise, the method is extremely slow, often taking minutes to reconstruct a single image. It requires significant computing power, putting a drain on hospital and clinic IT infrastructures, and can produce images that look overly smooth, hiding subtle features that radiologists rely on for diagnosis. With breast CT producing up to 500 images in a single scan, MBIR is impractical for real-world use.
DMLR uses AI to remove noise, but most approaches require pairs of high and low radiation dose scans to train the algorithm, which is unrealistic and increases radiation exposure. Other variations have been developed that learn from single low-dose scans, but they tend to perform poorly in breast CT, where image noise tends to be correlated in ways that confuse algorithms and make it difficult to preserve fine anatomical details. As a result, current DMLR methods are not well-suited to breast CT.
IzoView’s AI Solution
Izotropic’s breast CT image reconstruction algorithm is a novel self-supervised deep learning method that produces superior denoising while preserving the natural texture of breast CT images. Unlike other methods that attempt to clean image noise after reconstruction, IzoView’s algorithm works earlier on the raw X-ray data captured before reconstruction. It mitigates long computer processing times, does not require paired image datasets for training, enables consistent image optimization at low radiation doses, and is made for real-world clinical workflows in breast cancer screening environments. This innovation reinforces the case for low-dose protocols in regulatory submissions and could enable IzoView to set the global standard for image quality and radiation dose expectations in breast CT imaging.
Protected as a Trade Secret
As general-purpose AI models become increasingly commoditized, the development of tailored, modality-specific algorithms becomes a major industry differentiator. IzoView’s algorithm is trained specifically on 15 years of breast CT data, a technically complex and specialized imaging domain, and is protected as a trade secret. As generic AI reconstruction and existing denoising methods underperform in breast CT, Izotropic’s algorithm represents a strong competitive edge over legacy breast CT devices currently on the market and would be challenging and costly for others to attempt to replicate.
The Future of IzoView & AI
The integration of Izotropic’s proprietary AI reconstruction algorithm positions IzoView at the forefront of imaging innovation and establishes a strong foundation for the future of intelligent imaging. With computer-aided diagnostics positioned as the radiologist support tool of the future, systems that generate clean, high-quality imaging data will be essential to unlocking their full potential. IzoView’s high-resolution, denoised images, generated in a way that prioritizes patient safety, offer an ideal dataset for training and deploying advanced AI diagnostic applications. As imaging continues to converge with AI and precision care, Izotropic is positioned to drive the next generation of imaging devices and define new standards in breast CT imaging.
About Izotropic:
More information about Izotropic Corporation can be found on its website at izocorp.com and by reviewing its profile on SEDAR at sedarplus.ca.
Forward-Looking Statements:
This document may contain statements that are "Forward-Looking Statements," which are based upon the current estimates, assumptions, projections, and expectations of the Company's management, business, and its knowledge of the relevant market and economic environment in which it operates. The Company has tried, where possible, to identify such information and statements by using words such as "anticipate," "believe," "envision," "estimate," "expect," "intend," "may," "plan," "predict," "project," "target," "potential," "will," "would," "could," "should," "continue," "contemplate" and other similar expressions and derivations thereof in connection with any discussion of future events, trends or prospects or future operating or financial performance, although not all forward-looking statements contain these identifying words.
These statements are not guarantees of performance and involve risks, including those related to capital requirements and uncertainties that are difficult to control or predict, and as such, they may cause future results of the Company's activity to differ significantly from the content and implications of such statements. Forward-Looking Statements are pertinent only as of the date on which they are made, and the Company undertakes no obligation to update or revise any Forward-Looking Statements to reflect new information or the occurrence of future events or circumstances unless otherwise required to do so by law. Neither the Company nor its shareholders, officers, and consultants shall be liable for any action and the results of any action taken by any person based on the information contained herein, including, without limitation, the purchase or sale of Company securities. Nothing in this document should be deemed to be medical or other advice of any kind. All images are for illustrative purposes only. IzoView has not yet been approved or cleared for sale.
Contacts:
Robert Thast, Interim Chief Executive Officer
Telephone: 1-604-220-5031 or 1-833-IZOCORP ext. 1
Email: bthast@izocorp.com
James Gagnon, International Communications
Telephone: 1-604-780-7576 or 1-833-IZOCORP ext. 2
Email: jgagnon@izocorp.com
General and Corporate Inquiries
Telephone: 1-604-825-4778 or 1-833-IZOCORP ext. 3
Email: info@izocorp.com
Corporate Communications
IBN
Austin, Texas
www.InvestorBrandNetwork.com
512.354.7000 Office
Editor@InvestorBrandNetwork.com
Legal Disclaimer:
EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
