AI in medical imaging: a picture of health?

The past few months have been incredibly busy for us, yielding a wealth of captivating projects. Below, we'll share a glimpse into some of these endeavors.??

While the summer vacation season is upon us, we're not slowing down; instead, we're gearing up for an upcoming The ESC Congress 2024 taking place in London in two months. The event brings together the world's leading cardiology experts and its 2024 theme is personalized cardiovascular care. This approach aims to give patients more control over their treatment by tailoring it to their specific needs, ultimately leading to better results and reduced health risks.?

A noteworthy detail is that our experts are attending this congress, and they'll be delivering a presentation and a poster. We look forward to sharing with you further details soon.?

In the meantime, we invite you to read this newsletter edition.?

Brain metastases meet AI??

Metastatic brain tumors are more frequent than those originating in the brain itself. These fast-growing tumors can appear in numerous regions of the brain at once. MRI scans reveal a range of signs of metastatic brain cancer. The tumors can vary greatly in size and number, from very small spots to large masses. The number of tumors can also vary significantly, with some patients having just a few and others having hundreds. Scans may also show areas over time where surgery removed tumors, cell death occurred (necrosis), or swelling (edema) appeared. This variation makes manual analysis of these scans challenging, subjective, time-consuming, and prone to interobserver variability.?

In one of the projects for a client, we developed a system to automatically detect and analyze brain metastases in MRI scans over time. This system merges several powerful deep learning techniques originally designed for various purposes. We tested it on over 275 scans from 87 patients across 53 locations, using meticulously verified manual annotations as a reference. The system, built with open-source tools for broad accessibility, precisely detects tumors and monitors disease progression. To guarantee the system's functionality with various data types, we devised a novel strategy for dividing the data into training and testing sets. We also developed new methods for evaluating the system's performance. Overall, this fully automated system can significantly aid clinicians and researchers by automatically tracking disease progression and treatment effectiveness.?

A paper outlining this project and its findings was recently published in Computerized Medical Imaging and Graphics.

AI and endoscopy image analysis?

In the previous edition of our newsletter, we mentioned a poster presented at the European Crohn's and Colitis Organisation (ECCO) congress in Stockholm that showcased the results of a project we were involved in. The poster was titled "P098 The Endoscopic Severity Score Map (ESSM): An Artificial Intelligence scoring system providing accurate, objective and localised measurements of endoscopic disease severity in ulcerative colitis". While AI is already widely used in endoscopy, this modality can still benefit significantly from the application of artificial intelligence algorithms. Today we’d like to invite you to our blog and a post Machine Learning in Endoscopy Image Analysis: Enhancing Diagnostic Accuracy and Efficiency.??

The article delves into specific applications across various subfields, including quality assurance, detection, segmentation, and classification within endoscopic imaging. We believe that for healthcare professionals and researchers interested in the intersection of technology and medicine, this content might provide valuable insights.?

Cybersecurity aspects in FDA submissions??

As an ISO 13485 certified medical software development company, we keep up to date with the latest requirements and regulations for medical device certification and global market access. Recently, we published an article on our LinkedIn page discussing the cybersecurity requirements mandated by the FDA. Plenty of recent industry publications highlighted that these changes are leading to an increase in application rejections by the FDA.??

If this topic interests you or you need help with it, we would be happy to schedule a meeting.?