Posts Tagged :

medical imaging

H2020-Web

QUIBIM made huge strides in
AI-powered diagnosis thanks to EU H2020

Taking part in the EU’s H2020 initiative translated into huge strides in disease detection and diagnosis by QUIBIM, by enabling the development of tools powered by artificial intelligence (AI) that allow to extract patterns from medical images and significantly improve human performance.

The program, which came to a term on June 30th, culminated in QUIBIM PrecisionⓇ significantly expanding its network of customers and collaborators and an improved international brand awareness.  More recently the Chest X-Ray Classification tool, another solution in the QUIBIM PrecisionⓇ suite, also received CE clearance.

“We are very happy we’ve completed the program in time and line with all our objectives, allowing us to add one more stepping stone to the development of AI-boosted healthcare,” QUIBIM CEO Angel Alberich Bayarri said.

QUIBIM, a high-tech SME based in Valencia, Spain, received a €1.25M grant from the European Union’s Horizon 2020 research and innovation program in 2017, to help advance its machine learning and image processing algorithms to extract imaging biomarkers from images generated on CT, MRI, X-ray, US, DXA and PET scans.

With the EU’s support, the Spanish company could finalize the development, validation and automation of new imaging biomarkers in the areas of brain, lung and oncology, as well as achieve the integration of the new computing performance and data visualization frameworks for the QUIBIM platform.

Thanks to the automated analysis of imaging biomarkers, results are ready just within minutes with the best accuracy and reproducibility, through a medically certified, cost-effective and user-oriented solution, which is available to any physician using image-based diagnosis.

“QUIBIM allows physicians to make more accurate diagnosis by providing additional information extracted from the analysis of acquired imaging studies. Our technology uses algorithms that scout the image and extracts features as imaging biomarkers that can be compared to normative information in our database, based on patterns that are not easily visible to the human eye. The algorithms combine both AI data-driven techniques and model-driven approaches. As such, our products help to reduce costs of medical testing and misdiagnosis, especially from specialists,” Alberich said.

QUIBIM actively started in 2015 as a spinoff company of La Fe Hospital in Valencia, to help specialists including radiologists and pathologists make the most of AI in their everyday practice, by providing a one-stop-shop solution using imaging biomarkers and powerful algorithms.

More than 60 hospitals worldwide are currently working with QUIBIM tools, most of which have received CE marks and/or are FDA pending. The company has notably developed the new teaching platform of the European School of Radiology and has supplied its QUIBIM Precision® image analysis platform to the AI Precision Health Institute at the University of Hawai‘i Cancer Center in Honolulu.

QUIBIM has received €3.5m funding ever since its creation and has offices in Spain and the U.S. The company forecasts to generate revenues above €36 million and 150 direct jobs by 2023.

European Commision Flag QUIBIM Project

 

QUIBIM TEAM ECR 2019

ECR 2019 sets the trend in Artificial Intelligence

The city of Vienna was at the real core of Medical Imaging and AI at the 25th European Congress of Radiology (ECR). At QUIBIM we always enjoy being at this annual meeting because it is the perfect combination of a science and industry exhibition. In this edition, the congress reached the milestone of 30,000 attendees, and the numbers are expected to go up with each year.

QUIBIM was glad to be participating in both, the scientific and industry sessions. Our oral presentations were mainly addressing current challenges of artificial intelligence (AI) and convolutional neural networks (CNN) in clinical needs like metabolic disorder, prostate cancer and osteoarthritis. Ana Jiménez-Pastor, Rafael López-González and Fabio García-Castro, R&D Engineers at QUIBIM presented our new research in image processing pipelines aiming to perform a virtual dissection of the organs through an automated segmentation combined with features extraction.

Personally,  I was happy to give 3 lectures focused on the future of radiology: 3D Post-processing in 2019, Deep Learning in Medical Imaging and Start-up in Radiology. In the 3D Post-processing lecture I introduced, what I think is the main revolution of AI in our field the concept of Virtual In-Vivo Dissection (VIVID), a name coined by my team and I at QUIBIM, which is a strategy of isolating human body organs in medical images for  characterization through features such as imaging biomarkers. This has several applications, challenges and is difficult to solve by traditional computer vision algorithms like liver or cartilage segmentation in Magnetic Resonance Imaging but it has become a reality thanks to the use of CNN architectures such as U-Net combined with deep supervision. In the Deep Learning in Medical Imaging session, I  focused on using other presentation formats and I was glad to give a TED talk and share the podium with Dr. Wiro Niessen, who spoke about Machine Learning in a Pecha Kucha format. This session was organized by the European School of Radiology (ESOR) and was chaired by Prof. Dr. Valérie Vilgrain, and I must say the atmosphere was excellent and the room was really packed! Finally, in a session chaired by Prof. Dr. Elmar Kotter,  we shared our insights on how to create a start-up company in Radiology from scratch, how to get funding from investors and the main considerations when scaling-up.

QUIBIM was also invited by the ECR to give a presentation within the Artificial Intelligence Exhibition (AIX) sessions, a new space for innovative AI companies. During this session, moderated by Dr Hugh Harvey and Dr Wim Van Hecke, we presented the newest version of the QUIBIM Precision V3.0 platform launched at the last RSNA 2018.

boothecr

Meet us at ECR 2019

We are delighted to share that QUIBIM will attend the 25th European Congress of Radiology 2019 (ECR) from Wednesday, February 27th to Sunday, March 3rd, 2019. This year we have changed our standard location, now you can meet us at the AI Exhibition area (EXPO X1) booth #AI-15.

ECR attendees will have the opportunity to explore our latest version of the platform QUIBIM Precision® V3.0, focused on the Symbiosis of Radiology and Artificial Intelligence to seamlessly integrate imaging biomarkers into radiology workflows. Come and try our AI solutions and imaging biomarkers analysis!

Book a DEMO

Make sure not to miss our scientific contributions:

  • Wednesday, February 27:
    • 3D post-processing in 2019 – Dr. Ángel Alberich Bayarri.   |  Imaging Informatics, Artificial Intelligence and Machine Learning (Room N) – 16:00 – 17:30
    • Deeply supervised networks for the automated liver segmentation and quantification on MECSE-MRI – Ana Jiménez Pastor | EPOS
    • Stress testing a deep learning algorithm for normal/abnormal classification of Chest X-rays on a spectrum-biased abnormal – Rafael López Gónzalez  |  EPOS
      weighted dataset.
  • Thursday, February 28:
    • Functional imaging of the liver Chairperson’s introduction – Dr. Luis Martí Bonmatí. |  Abdominal Viscera, Contrast Media (Room M 5) – 10:30 – 12:00
    • AI PITCH – Dr. Ángel Alberich Bayarri.  |  Artificial Intelligence Exhibition (AIX) Theatre ( AIX Theatre) – 11:40
    • What to think about when writing a paper – Dr. Luis Martí Bonmatí. |  Education, General Radiology, Professional Issues (Room: C&T 3) – 14:00 – 15:00
    • Deep learning (DL) in medical imaging – Dr. Ángel Alberich Bayarri.   |  Education, General Radiology, Artificial Intelligence and Machine Learning (Room: M3) – 14:00 – 15:30
    • Quantification and evaluation of pre-post exercise femoral cartilage thickness and T2 changes in ultramarathon athletes – Fabio García Castro  |  Musculoskeletal, Imaging Methods (Room: O) – 14:00 – 15:30
    • How to manage critical reviews – Dr. Luis Martí Bonmatí. |  Education, General Radiology, Professional Issues (Room C&T 3) – 15:00 – 16:00
  • Friday, March 1:
    • ECR Academies: Radiology Leaders’ Bootcamp: Dream team Chairperson’s introduction –Dr. Luis Martí Bonmatí. |  Management/Leadership (Room M 2) – 10:30 – 12:00
    • Start-up in radiology – Dr. Ángel Alberich Bayarri.   |  Management/Leadership (Room M2) –  14:00 – 15:30
  • Saturday, March 2:
    • Automated Prostate multiregional segmentation in Magnetic Resonance using deeply supervised Convolutional Neural Networks – Rafael López González  |  Artificial Intelligence and Machine Learning, Oncologic Imaging, Imaging Informatics, Genitourinary, Physics in Medical Imaging (Room G) – 16:00 – 17:30
  • Sunday, March 3:
    • Automatic visceral fat characterization on CT scans through Deep Learning and CNN for the assessment of metabolic syndrome –  Ana Jiménez Pastor | Artificial Intelligence and Machine Learning, Abdominal Viscera, GI Tract, Oncologic Imaging, Imaging Informatics (Room D) – 14:00 – 15:30
    • Liver Case-Based Diagnosis Training – Dr. Luis Martí Bonmatí. |  Education, General Radiology, General Radiography (Radiographers) (Room E1) – 13:00 – 15:30

Join us at ECR 2019!

 For more information, get in touch with us at contact@quibim.com

QUIBIM_MIUC

MIUC
the new toolkit of QUIBIM Precision® platform to beat traditional workstations

Quibim has implemented a new toolkit named MIUC (Medical Imaging Universal Connector) to close the gap between hospital IT systems and the Cloud. Whereas the Cloud satisfies the processing requirements, Quibim Precision® handle the functionalities related to communications and management of DICOM objects among the hospitals and radiological centers.

Quibim Precision® allows users in hospitals and radiology departments to have a seamless integration of imaging biomarkers analysis within the radiological workflow, due to the MIUC capabilities combined by the Quibim Precision® Cloud computing environment and the interoperability features implemented in our system. Both image upload and data retrieval are fully automated and users only need to access the PACS when they are notified that a new biomarker report is available.

MIUC is placed inside the hospitals and clinics and it is responsible to establish all the required communications between the PACS and Quibim Precision®. During the analysis of imaging biomarkers, the study is anonymized, sent to the Cloud and analyzed. The final result is a one-page report which is sent back to the MIUC or can be directly visualized in the Quibim Precision® web interface. Furthermore, in clinical environments, the report is converted into DICOM objects and stored in the PACS as a new series within the original study. To identify the original study, the MIUC implements backward traceability in the client side to reidentify the anonymized studies.

Our platform is intended to be used by radiologists, either from a clinical environment, thanks to the MIUC, or as a final user using the web interface. In the clinical environment scenario, radiologists using Quibim Precision® do not have to worry about where the study or the report is. Instead, these issues are transparent to the user, who do not have to perform any action to launch a biomarker process, given that the MIUC rule engine does such work for them. The user will be notified by email when a new biomarker report is ready (and available in both the PACS and the Quibim Precision® web interface).

Nowadays, our Quibim Precision® platform is compliant with the DICOM standard at both communication level and data management and formatting level. Specifically, our platform receives imaging studies from hospitals, radiological centers or pharma companies. Then, the system analyzes the study and obtains quantitative measures, which are stored in a quantitative database and structured in a one-page report on a per-patient basis. Finally, this report is returned back as a result.  Quibim Precision® allows annotating biomarker reports using terms from RadLex and MeSH, enhancing the interoperability of its biomarker reports with other health information systems. In fact, the imaging platform is seamlessly integrated with the hospital PACS, being able to query and retrieve medical studies, processing them and storing the resulting biomarker reports as DICOM objects in the hospital PACS. On the other hand, the processing stage is performed on the Cloud, taking advantage of its benefits: high-performance computing and real-time hardware scalability on demand.

But, what has changed?

In previous updates of our platform, we improved the performance, capabilities and user settings view. With this new suite software QUIBIM Precision®- MIUC does query/retrieve the PACS, anonymizes the PACS responses and forwards them to Quibim Precision® in the Cloud. Furthermore, the MIUC leads our solution to a higher level of automation, given that it monitors the PACS querying for incoming studies. Once a new study reaches the PACS, the MIUC analyzes its header and determines whether a new biomarker analysis must be launched or not, depending on some DICOM elements in the study like imaging modalities, study description, series description or body part among others. For each biomarker analysis available in Quibim Precision®, there is a predefined set of rules that establishes which studies are susceptible to be processed by each analysis method. An incoming study matches a given analysis method whenever it fulfils the predefined set of rules for such analysis method. When this happens, the MIUC automatically sends the study to the Quibim Precision® Cloud processing platform, where it will be processed by the matching biomarker analysis pipeline. Once processed, a biomarker report is generated with the results and sent back to the MIUC. Finally, the MIUC stores the report in the PACS as a DICOM object, making the report available for the specialist who requested it. This way, the Cloud platform remains centralized and, at the same time, fully integrated with the hospital IT systems.

With the arrival of MIUC toolkit the need for conventional workstations with expensive licenses in radiology departments completely disappears. As in other business areas that are evolving from product to service, the Quibim image analysis technology was designed to be offered as the service that puts disruptive image analysis solutions at your fingertips.

SMEInstrument

QUIBIM granted with the H2020 – SME Instrument Phase II

QUIBIM team is proud to announce that the beneficiaries of the European Commission SME Instrument Phase 2 have been recently published, and our project has been awarded in the Open Disruptive Innovation (ODI) topic.

QUIBIM’s project “QUIBIM Precision” has been granted with 1.25M€ for the scaling and development of the company business plan through our cloud platform QUIBIM Precision®. This funding will boost and accelerate the strategic plan of the company, bringing new disruptive solutions to our platform, including dedicated High Performance Computing resources for our Machine Learning processes, incorporating new visual analytics tools and helping validate new imaging biomarkers that have been recently developed.

QUIBIM Precision® is an innovative imaging biomarker analysis platform in the cloud allowing for:

  1. Automated analysis of imaging biomarkers (results are ready just within minutes) with the best accuracy and reproducibility.
  2. Medically certified: QUIBIM is medically valid to scientifically support decision making.
  3. Open to any physicians: Optimized User Interface (UI), user experience (UX) and imaging analysis functionalities.
  4. Cost-effective: QUIBIM helps reduce costs of medical testing and misdiagnosis.

Our technology based on machine learning and image processing algorithms scouts the image and compares it to similar images in our database with known ground-truth diagnosis, based on patterns not obvious to a human eye.

About QUIBIM
QUIBIM is a biotechnology company dedicated to advanced medical image analysis. It was one of the start-ups selected in the 3rd edition of the acceleration program “LANZADERA” and it’s a spin-off company of the Medical Research Institute Hospital La Fe. Recently the company closed an investment round to boost its growth, including Tech Transfer UPV, AYCE Capital, and BioInfoGate, as well as some company promoters . QUIBIM’s business model is based on the extraction of quantitative information from radiological and nuclear medicine imaging using original and advanced computational processing techniques.

In 2016, QUIBIM was also selected as one of the disruptive companies by the EU and granted with the first round of Horizon 2020 SME Instrument Phase I Program 2016 under the call: “SMEInst-01- 2016-2017: Open Disruptive Innovation Scheme”

SME instrument – Horizon 2020 call
The SME Instrument is divided into 3 phases covering different stages of the innovation cycle. Phase 1 aims to cover the assessment of technical feasibility and market potential of new ideas. The project will be supported by an investment of 50,000€ and the typical duration should be no longer than 6 months.

Phase 2 aims to cover R&D activities with a particular focus on demonstration activities (testing, prototype, scale-up studies, design, piloting innovative processes, products and services, validation, performance verification etc.) and market replication encouraging the involvement of end users or potential clients. Project funding should amount to no more than 2,500,000€ and the typical duration of this phase should range from 12 to 24 months.

3D_Full_Low_Both

Assessing the bone quality of your patients is just a few clicks away

Quibim is proud to introduce its new advanced methodolgy to assess bone quality and fracture risk: QTS (Quality of Trabecular Structure). The need for a reliable approach for the characterization of trabecular microarchitecture is evident, as conditions and diseases related to trabecular bone quality and structure are becoming a focal point of precision medicine. Millions of dollars are spent in bone fracture care and prevention, which in many cases represent a grave danger to the patient health. In this scenario, can we rely upon available methods to predict bone fractures? Regrettably, the answer is no. Or so it was. Before we introduce QTS in detail, let’s first review some of the available methods to evaluate fracture risk.

DEXA Bone Mineral Density (BMD) limitations as a fracture predictor are widely known. However, despite its glaring shortcomings it’s still considered as the gold standard in clinical practice. The WHO developed the FRAX tool hoping to improve the low ratio of success of BMD as a predictor, and, while it improves sensibility and specificity assessing fracture risk, it’s still not satisfactory enough for it to be used as a stand-alone predictor in clinical practice or clinical trials.

In an attempt to fill this void in the search of a reliable predictor, new imaging biomarkers that guarantee to provide the needed sensibility and specificity have been developed. Trabecular Bone Score (TBS®) by Medimaps is a fracture risk predictor using DEXA as a source. TBS® performs a gray level analysis (textures) on the DEXA images to determine bone integrity. But, is DEXA adequate for this task? Unlike regular XR, in which it’s possible to assess the trabecular structure, DEXA’s low radiation is not enough for trabeculae differentiation. In addition, DEXA (like XR) represents a 3D structure as a projection onto a plane, losing spatial information.

Other medical imaging modalities are much more adequate for bone analysis and the subsequent fracture risk evaluation. Thanks to advanced computing models, Quibim has developed a new imaging biomarker that uses either magnetic resonance (MR), computerized tomography (CT) or X-ray imaging for a detailed characterization of the trabecular structure. QTS (Quality of Trabecular Structure) by Quibim introduces several advantages over other analysis methods:

  • Real extraction of the trabecular bone microarchitecture: bone volume, trabecular thickness, trabecular separation…
  • Information of the complexity of the structure: fractal analysis
  • 3D spatial information, not only in plane (only CT and MR)
  • 3D reconstruction of the trabecular bone (only CT and MR)
  • Optional mechanical analysis (only CT and MR, with QTS+)
  • QTS Score comprises all this information in a single score for a rapid and accurate characterization of the bone structure.

This groundbreaking analysis method is already available at our cloud web platform: Quibim Precision. Below we detail how to analyze your study in a few simple steps.

Quibim Precision allows the upload of studies in Dicom format. The upload process is easy, intuitive and 100% secure, guaranteeing the patient confidentiality thanks to Quibim’s anonymization and encryption system. The whole process is performed without the need of installing any additional software. To start, click on the green “Upload Study” button to the right of the website.

00

The platform will then ask for Dicom studies selection. With Google Chrome, we can drag and drop the study folder directly to the dotted box. With any other browser, we need to select the Dicom files to be uploaded.

00b

Once the Dicom folders have been selected, Quibim Precision allows the pre-visualization of the study to choose which sequences we wish to upload. In this case, we need a CT, an X-ray or a 3D T1 MR.

After the selection, the files will be anonymized and encrypted. Quibim Precision will ask for an encryption password before the upload process starts. The user needs to preserve this password, as it’s needed for patient traceability.

00c

After the upload process is completed, we click on the “Analyze Study” button of the study we want to evaluate.

01

This will take us to the detailed view of the study, where we can perform all actions related to it. First of all, we will choose among all the available sequences the one to be analyzed and its Quibim standard name. In this case, the name of the sequence is “Linear Attenuation [1/cm] (3035)”, which corresponds to the standard name “High Resolution CT”.

02

Next, we open the embedded Dicom viewer by clicking on “View”. Thanks to this viewer we can draw the ROI that will be used for the analysis. The viewer allows drawing 3D ROIs, first drawing a 2D rectangular ROI, and then selecting the slices on which it should be replicated. Clicking on “Apply” and “Save and go back” will store the ROI and it will be used for the analysis.

04b

Finally, we should choose which analysis method we’ll use among all the available apps in Quibim Precision. In this case we’ll click on the “Start Analysis” button of the “3D Bone microarchitecture – QTS Score” app. The analysis process will then start, and the user doesn’t need to perform any other action than checking the results.

03

The results of the analysis are available on the view of the study. Once the analysis is completed we can check them by clicking on the “View Study” button on the “Processed Biomarkers” section.

05

We can examine the 3D reconstruction images and the numeric results of the extracted imaging biomarkers on the results view. Furthermore, we can download them and the structured report in pdf format, which includes all the generated data in a compact, easy to read way.

06

07b

08

All the benefits of QTS are just a few clicks away. Create an account on Quibim Precision and start offering a real added value with your imaging studies.

Try us for free!

Metrology for imaging biomarkers

Will metrology labs take charge of Imaging Biomarkers certification?

As the reader might know, Imaging Biomarkers are parameters obtained objectively from medical images, which represent and quantify a tissue property (structural, functional or biological) extracted after applying computational models to images of specific medical imaging modalities like computed tomography (CT), magnetic resonance imaging (MR), conventional radiography (CR) and so on.

One of the main challenges for Imaging Biomarkers integration in clinics and in clinical trials is the need for standardized image acquisition protocols  and  pipelines for the analysis of imaging biomarkers, including the processing, analysis and reporting of the results.

As specified in this publication from QIBA (Quantitative Imaging Biomarkers Alliance) team: “Most Biomarkers require a computation algorithm, which may be simple or highly complex. While there is a rich history of the development of biomarker techniques, there has been comparatively little attention paid to the evaluation and comparison of the algorithms used to produce the biomarker results. Estimation errors in algorithm output can arise from several sources during both image formation and the algorithmic estimation of the biomarker. These errors combine (additively or non-additively) with the inherent underlying biological variation of the biomarker. Studies are thus needed to evaluate the imaging biomarker assay with respect to bias, and precision. A recurring issue is the lack of reported estimation errors associated with the output of the biomarker. Another challenge is the inappropriate choice of biomarker metrics and/or parametric statistics. For example, tumor volume doubling time is sometimes used in studies as a biomarker. However, it may not be appropriate to use the mean as the parametric statistic for an inverted, non-normal, measurement space. Since a zero growth rate corresponds to a doubling time of infinity, it is easy to see that parametric statistics based on tumor volume doubling time (e.g. mean doubling time) may be skewed and/or not properly representative of the population. In addition, there may be discordance between what might be a superior metric statistically and what is clinically acceptable or considered clinically relevant. For example, a more precise measuring method will typically better predict the medical condition, but only until the measurement precision exceeds normal biological variation; further improvement in precision will offer no significant improvement in efficacy. Finally, when potentially improved algorithms are developed, data from previous studies are often not in a form that allows new algorithms to be tested against the original data.”

A metrology-by-design strategy for imaging biomarkers design and implementation has still not been conceived, however, there is a clear need to fill this gap, since medical images are suffering a paradigm shift from being qualitatively interpreted by radiologists to become a powerful measurement instrument, that can provide a non-negligible and highly relevant information in many pathological processes.

Following the experience of our group and QUIBIM, with a model of success for the integration of quantitative image analysis in clinics, we published in 2012 our methodology for the integration of imaging biomarkers in clinics that was later supported by the European Society of Radiology in 2013, through the publication “ESR position on the stepwise development of imaging biomarkers” in Insights into Imaging. This paper was the former work of the ESR supporting the creation of the European Imaging Biomarkers Alliance (EIBALL) in 2015.

Now, several actions should be undertaken to allow for the implementation of biomarkers in clinical use and the evaluation of therapy response:

  • To determine the most significant sources of uncertainty for imaging biomarkers.
  • To establish a standard methodology for definition of the method and the protocol for image acquisition (Signal-to-noise ratio, contrast-to-noise ratio) for traceable measurements.
  • To harmonise calibration routes and methods for imaging biomarkers in European metrology institutes. This objective includes knowledge exchange and transfer and comparison measurements.
  • To provide metrology input and pre-normative research to the evolution of International (ISO, EN) standards concerning imaging biomarkers.
  • To engage with image device manufacturers to facilitate the take up of the technology and measurement systems developed, to support the development of new, innovative biomarkers, thereby enhancing the competitiveness of EU industry.

While researchers are dealing with these issues, there are specific entities that have been always in charge of supervision, validation and certification of measures: the National Metrology Institutes. The umbrella covering all National Metrology Institutes in Europe is EURAMET. Therefore, we should start to think in incorporating this metrology-by-design strategy in the process of creation, implementation and validation of imaging biomarkers. In QUIBIM we have already started this strategy, following QIBA recommendations and implementing new algorithms in a modular basis, identifying the sources of uncertainty in all the measurements that we perform with imaging biomarkers.

by Angel Alberich-Bayarri (@aalberich).

CEO & Founder of QUIBIM