Science In Silico: Medical Imaging & High Performance Computing
April 26, 2025
In this article, we look at 8 companies who are developing radiological diagnostic products, and see how HPC enabled, AI-enhanced diagnostic tools are enhancing physicians’ diagnostic workflows, and ultimately creating a more streamlined patient experience.
Medical diagnostics have long been at the forefront of adopting revolutionary new computer technologies. In the 1970s, the first computed tomography (CT) machines harnessed the power of computing to build advanced imaging systems, laying the groundwork for the industry’s embracing of ever improving HPC technologies that continues to this day.
As the availability and reliability of high performance computing (HPC) resources has improved over the years, so has the quality and speed of diagnostic tools. Magnetic resonance imaging (MRI) machines, for example, have seen an exponential increase in their magnetic strength in the past 50 years. The original MRI machine had a Tesla rating of 0.09, whereas the industry standard today is 3.0. This dramatic uptick in magnetic power means that modern scanners are able to produce extremely focused, high resolution images, and they need to rely on increasingly powerful computational resources to achieve them.
Whilst there are a number of different diagnostic areas and techniques, medical imaging is undoubtedly one of the most computationally heavy. In fact, HPC is so integral to their functionality that standard modern radiological technologies like CT and MRI could not exist without it.
The continued symbiotic improvements in HPC resources and diagnostic capabilities are visible in the recent generative AI boom, which has seen a multitude of companies developing HPC enabled, AI-enhanced diagnostic tools. These products rely on increasingly sophisticated algorithms and require significant computational resources to develop and run, but there are already countless examples where they are proving to be an important and effective part of front line medicine.
The Diagnostic Process
Brainscan’s diagnostic workflow, where the need for HPC resources is baked into every step.
Radiological techniques such as CT or MRI involve creating huge amounts of raw data, and this needs to be processed and reconstructed into a diagnostically useful display. This can be a very computationally complex process, especially when there is a need for extremely high resolution results.
Reconstruction algorithms will convert this raw data into DICOM files, which combine text, visual images with embedded patient data and are the standard industry format for medical imaging. These files are then transferred to a Picture Archiving and Communication System (PACS), a central repository where the images can be stored, managed and distributed as necessary. These may be stationed on-premises, at either the healthcare facility or at a third party data centre, or, as is increasingly the case, in the cloud.
Traditionally, these DICOM images will then be assessed by a radiologist. However, companies like the ones in this article are driving a push towards using computationally intensive techniques to digitally screen the images.
Diagnostic algorithms are extremely good at assessing large amounts of medical images, and finding areas of interest which may signify a certain pathology or disease. They are not, however, currently so nuanced that they can do the job of trained radiologists and diagnose a patient as ‘healthy’.
Medical imaging AI diagnostics companies are not generally attempting to position their product as something that will replace physicians, but rather as something that helps streamline their workflows, and give them a second opinion on their diagnoses. The terms augmented intelligence and intelligence amplification are increasingly being used in the industry to reflect their acknowledgement of the limitations of generative AI, and to reinforce the importance of having a trained human be the ultimate decision maker.
The Cohort
A study of 100 CE approved AI diagnostic products gives a good overview of the types of scan and body parts the tools work with.
In this article we focus on what are generally considered to be radiological diagnostic techniques, specifically looking at products that work with scans such as x-rays, CTs or mammograms. This means companies like Aiforia, who focus on microscopic histology technologies, or Ultromics, who offer an ECG analysis product, don’t make the cut. We will, however, be featuring alternative diagnostic methods in future articles.
Paris based AZMed provides an augmented intelligence solution which scans X-rays for potential bone fractures and prioritises radiologists’ workloads, presenting them with comparative DICOM images to support their diagnosis. The system has been successfully rolled out in over 1000 hospitals across the world since its 2021 EU MDR approval.
Red Dot is a computer aided diagnosis platform designed by London based BeholdAI. It has two core anatomical focuses, scanning chest x-rays and brain CTs to filter out normal results and alert radiologists to urgent cases. The products are currently being integrated across the NHS after successful trials at local trusts, and are also currently in operation at healthcare facilities in India and the US.
Brainscan is a Polish startup focusing on using AI to assess CT scans of the brain for abnormalities and pathological changes, before outputting an infographic for human review. Initially trained on over 250,000 scans, the platform gained CE approval last year and is currently being used in a handful of Polish healthcare institutions.
A spinout of the Medical University of Vienna, Contextflow’s software, focuses on the detection of chest disorders from CT scans. Their AI algorithms look for a number of different potential regions of interest, then deliver results through a series of visualisation tools which radiologists can then use to augment their own analysis.
Similar to AZMed, Gleamer is a French company offering a platform for assisting in the X-ray diagnosis of bone fractures. Spun out of the PSL Research University in 2017, they currently have four CE approved products, which process X-Rays and provide radiologists with DICOM overlays. Alongside their focus on skeletal data, they also have an algorithm which assesses chest scans, and a product in development that aims to interpret mammograms.
Image Biopsy Lab offers both an overarching musculoskeletal imaging analysis platform, as well as a number of CE marked and FDA approved products focusing on individual bones and joints. The Vienna based firm’s products analyse x-rays and CT scans for anomalies before presenting doctors with image and text based reports and their products are currently being installed in over 100 healthcare institutions.
With a wealth of scientific publications backing up its efficacy, Kheiron Medical Technologies flagship Mia platform is at the forefront of modern breast cancer diagnostics. Their CE certified suite analyses mammographic images to support radiologists in their decision making. The company has been working with the NHS for the past five years, incorporating the technology across the service.
Alongside a CE approved imaging instrument which works as a double reading aid to help radiologists verify their analysis of chest x-rays, the Lithuanian company Oxipit also offers a unique retrospective analysis tool. This service uses AI to audit doctors’ diagnostic reports, checking them for accuracy and validating their decisions.
Deeptech Deep Dive
The diagnostic products that the companies on this list have created fall under the umbrella of generative AI. This is where machine learning techniques generate new content based on a previously analysed dataset, and the process consists of two distinct stages: training and inference.
The initial training phase involves providing a model with large quantities of data, which for medical diagnostics, are usually publicly available scientific datasets. Certain parameters are then given which tell the model the desired output, in this case a potential anomaly in a radiological image. An algorithm then analyses the data and attempts to produce results which match the parameters, continually optimising and iterating itself until it reaches the desired level of accuracy.
Inference is where a new piece of data is given to the AI model, and a decision or prediction is made by comparing the new data to the knowledge it has boiled down from completing its training on the core dataset. In medical diagnostic terms, this will often mean analysing x-rays or CT scans to see if there are areas of interest which may indicate a pathological change. The bulk of scans will return as not showing any abnormalities, but if the algorithm finds something that a clinician needs to check, they will create infographics or additional DICOM layers alerting the doctor to their findings.
The initial training stage is much more computationally intensive than an inference. The training phase will generally require HPC clusters to be completed, but a single inference uses only a tiny fraction of the memory and computing power of a full training cycle, so can be performed on much less powerful machines. The majority of the total compute spend for a diagnostics AI product is in the training phase, but models continually need to be updated and refined as new data and techniques become available.
One of the more common imaging processes utilised by diagnostic AI companies is a convolutional neural network (CNN). This is a deep learning technique where the primary focus is processing and analysing visual data. This differs from something like a large language model (LLM) such as ChatGPT, which is more tuned to analyse and produce text and language based data.
The CNN will begin to process an image at an individual pixel level, identifying key elements of the layout as it goes. On its first pass, it may only identify very broad features, such as edges or textures, but as it moves through the process, additional convolution layers add more filters. This allows the CNN to build a more intricate analytic picture, eventually reaching a level of detail that results in the detection of a required output, for example identifying an anomaly in a CT scan that requires further attention.
The bulk of scans run through medical diagnostic tools actually come back as normal, and by using technology to filter these out and only provide consultants with images of interest, a huge time and cost burden is lifted from the shoulders of the healthcare provider.
Additionally, it is highly likely that once a diagnostic product has been successfully integrated into a healthcare institution, the HPC costs of running inferences will pass onto the institution. The software may run on top of a hospital’s internal infrastructure, or the data may be sent to a third party server for inferences, but it is likely that any additional costs will be factored into the total sale price of the service.
The exact time and cost of creating an HPC enhanced diagnostic tool can vary depending on the scale and complexity of the data and the level of validation required for regulatory compliance. Compared to a field like drug discovery, where getting a product onto the market can take $2 billion plus and over a decade, the process for getting a medical diagnostic ready for EU Medical Device Regulation (MDR) and FDA approval is much faster and cheaper. The majority of the companies on this list were able to successfully take a product from the bench to the bedside within 5 years with relatively modest fundraising efforts.
The publicly available financial data for this group of companies is too inconsistent to allow in depth analysis. However, it does appear that once a company has successfully obtained CE MDR and/or FDA approval for a product, they can quickly leverage this success to obtain further funding. Diagnostic AI companies also appear to have strong cases for securing grants to fund their work, with the European Innovation Council (EIC) in particular providing support to a number of companies in the sector.
Who’d Be a Radiologist?
In 2016, noted data scientist Professor Geoffrey Hinton famously quipped that “people should just stop training radiologists now. It’s just completely obvious that in five years deep learning is going to do better than radiologists”.
Whilst in 2024, there does still remain some conflicting opinions on these new technologies’ efficacy, there are also plenty of studies that indicate HPC powered algorithms can be just as accurate as trained radiologists in detecting abnormalities. And, due to the iterative nature of these techniques, continual increases in their precision could validate their widespread integration, and even help counter the growing radiologist talent gap.
There is currently a global shortage of trained radiologists, with those currently in the profession overworked, which inevitably leads to high levels of stress, and unfortunately, mistakes.
It’s estimated that there is a global real time diagnostic error rate of between 3-5%, meaning there are around 40 million diagnostic imaging errors happening every year. The results of these mistakes can be catastrophic, both for patients, and also for the doctors that make them. Staggeringly, over 70% of radiologists in the USA have been named in a malpractice lawsuit, which can only add to their levels of stress and the potential for future errors.
There is also a cost to the healthcare service, both reputationally and financially. The NHS paid out £71 million for radiology malpractice claims in 2021, so it’s no surprise to see the organisation keen to adopt new technologies which can help improve diagnostic accuracy.
Breached Whales
The medical diagnostic process produces massive amounts of extremely sensitive personal data, so it is imperative that diagnostic AI companies and the healthcare institutions using their products have a firm grasp on the regulatory compliance required to ensure the data is stored and managed securely.
There is a litany of recent examples of data breaches at radiology institutions, and whilst it can be very difficult to successfully navigate the technical safety requirements of regulatory guidelines like HIPAA or GDPR, it is absolutely essential for a company to ensure they are fully compliant. The repercussions of not adhering to these mandates extend far beyond reputational damage, with the resulting fines and lawsuits potentially costing in the millions.
Darwinist’s Beagle platform utilises Red Hat’s industry leading enterprise operating systems and identity management features to provide diagnostic companies with a turnkey solution which ensures that their HPC workflows and data storage processes are fully compliant.
References
https://pubs.rsna.org/doi/10.1148/rg.2018180021
https://www.rsna.org/news/2022/may/global-radiologist-shortage
https://phys.org/news/2017-02-real-time-mri-analysis-powered-supercomputers.html
https://www.cancerimagingarchive.net/collection/prostatex/
https://www.gov.uk/government/news/21-million-to-roll-out-artificial-intelligence-across-the-nhs
https://transform.england.nhs.uk/ai-lab/ai-lab-programmes/ai-in-imaging/ai-diagnostic-fund/
https://www.bmj.com/content/379/bmj-2022-072826
https://www.hipaajournal.com/hipaa-violation-cases/
https://link.springer.com/article/10.1007/s00330-021-07892-z