Hot!

Other News

More news for your entertainment

How to Sell Machine Learning Algorithms to Healthcare Providers

The explosion of the computer aided diagnosis techniques is a reality in our days. There are thousands of new publications every year about new automatic measurements that turn into bio-markers. 

Image result for medical imaging
As we said in From Lab to Specialist (here), the challenge is put that technology in the medical specialist hands. Definitively they need to get involved because they know the science and the most important thing, the real state of the patient (for example, there is pain?) 

How can we continue developments and also help to have better treatments? If the technology is ready for validate drugs why it's not being used? I there is an FDA approval  or CE marked app, why is not used to choose the best treatment for the patient?



Image result for image biomarkers
Medimsight was created exactly for that, while others were working on exclusive plugins with expensive license models, Medimsight defined an open solution with a pay per use model so even a teenager in the basement and a senior designer in a big company, they have the same change. We want to stay in challenges, promote research and accelerate certifications. So clinicians get more innovation. It lowers costs. It increases creativity. Everyone can contribute. A catalyst for innovation.

Simon Harris wrote an article about what are the real options out there.

Here is a part of the article
One of the greatest commercial challenges for developers of medical image analysis algorithms is how to take their products to market. Most independent software vendors (ISVs) of image analysis solutions only offer a handful of algorithms for specific use-cases, e.g. coronary calcium scoring, bone age prediction, detection of lung nodules, etc. However, most generalist radiologists require a comprehensive “analytical tool kit” with a broad portfolio of algorithms that can detect a wide range of conditions for multiple body sites and across multiple modalities. Locating, evaluating and sourcing image analysis algorithms on a piecemeal basis from multiple vendors will be a cumbersome and time consuming process for healthcare providers. Not to mention the challenges associated with integrating the algorithms with the providers’ existing healthcare IT infrastructure. Whilst this may be a viable option for the larger academic hospitals and IDNs, most providers will not have the necessary resources for this and instead will prefer to deal with a small number of vendors, and ideally a single supplier.
And what they say about us:
Medimsight offers a cloud-based computer-aided diagnosis marketplace for biomarker quantification. The platform features 39 applications, including algorithms from LAIMBIO, FMRIB (Oxford Centre for Functional MRI of the Brain) and Martinos Center for Biomedical Imaging.  
You can read the whole one in Simon's page: Link 


About Simon HarrisManaging Director at Signify Research.
https://www.linkedin.com/in/simonharris2/

AI Powered Startups Around The World


AI is changing the way the interact with computers, internet and small things. Computer vision it's been always one of the most imporant tools to bring AI to real life problems. Aumented reallity, robotics, security, industrial quality controls and finally healthcare are some of the most promising uses.

It's an honor been included in this list of startup that are using these technologies to enhance healthcare systems and reduce people diseases.

Here is a part of the article
AI is becoming more and more pervasive. I've been discussing this with startup founders around the world who are developing software using AI. Startups are using machine learning, natural language processing, computer vision, facial recognition, speech recognition, pattern recognition in large data sets, and more to enhance capabilities and improve performance. Startups are using AI to fight human trafficking, to track terrorists, to detect and treat diseases, to help doctors quantify data, to match people with suitable employment, to automate businesses processes, to simplify travel and to accelerate research.
You can read the whole one in Margaretta's page: Link 

About Margaretta: Native San Franciscan with 30 years of experience working for software companies in Silicon Valley. She joined 6 early stage tech startups since 1992. She is building an ecosystem of entrepreneurs, engineers, researchers and investors with innovative ideas. Her interests include AI, machine learning, life extension, biotechnology, quantified self and precision medicine. She mentors startup founders in San Francisco, Los Angeles, New York, Boston, Austin, Washington D.C., Albuquerque, Barcelona, Bangkok, Nairobi, Seoul, Bosnia, Uganda, Morocco and The Russian Federation. 

How Computer Vision Is Transforming Medicine

Medimsight It's been highlighted in the list of companies transforming medicine thanks to the use of computer vision techniques.

There is a big number of companies building solutions around computer vision, especially using deep learning. It's been an incredible revolution and we can see that in most of the challenges to get the best biomarker values, deep learning techniques lead in term of results.

Here is a part of the article
Some of the most innovative biotech startups are developing technologies using computer vision to process medical images. These technologies help doctors detect malign changes such as tumors and hardening of the arteries and provide highly accurate measurements of organs and blood flow. With the exponential growth in computer performance which is increasing the speed at which computers process data, computer vision is becoming an increasingly valuable decision making tool in medicine.
You can read the whole one in Margaretta's page: Link 

About Margaretta: Native San Franciscan with 30 years of experience working for software companies in Silicon Valley. She joined 6 early stage tech startups since 1992. She is building an ecosystem of entrepreneurs, engineers, researchers and investors with innovative ideas. Her interests include AI, machine learning, life extension, biotechnology, quantified self and precision medicine. She mentors startup founders in San Francisco, Los Angeles, New York, Boston, Austin, Washington D.C., Albuquerque, Barcelona, Bangkok, Nairobi, Seoul, Bosnia, Uganda, Morocco and The Russian Federation. 

Increasing the power of image-based clinical trials using the cloud

Yes, there are a great number of medical image quantification methods out there, and improvements to the methods or new ones are published every month. Most of them never reach the clinic, and are never used by anyone other than the authors of the study reported in the paper. The medical image analysis community has been reinventing the wheel for a long time.





One of the reasons is the lack of availability of the methods. Most of the are never turned into a software platform that a clinician can use, and even those that are available, are too expensive or too specific for a clinical center to buy. Only centres acquiring a large number of certain studies find it worth to acquire a software platform. Thanks to Medimsight, this can completely change, as the developer of the method can simply upload his method to our platform, and the next day it is available for any clinician in the world with a simple user interface and a powerful visualization environment.

Another classical problem in medical imaging is the evaluation of the methods. Most methods, when first published, are tested only on a small sample of images. A classical sentence in the discussion section of these papers is ‘Further evaluation with a larger dataset is warranted in order to evaluate the clinical application of our method’. On of the reasons for this is the limited amount of subjects with a certain pathology that are scanned in a clinical center, specially if the pathology is not very prevalent. Interestingly, cloud systems such as Medimsight are a direct solution to this problem. When two independent clinicians upload a their image datasets to quantify them through one of the methods available in Medimsight, they could choose to notify the system about the description of their study. The platform can detect studies about the same pathology, or using the same analysis method, and could offer the clinician/researcher to participate in a meta-study with a much large of images. Simply, with the increasing use of Medimsight, studies of a method with a large cohort of multi-centric data will happen naturally. In the classical software platform based model, these would never happen.


This is great way to test a method, which is why developers are turning to Medimsight.


It is also a great way to compare data quality and quantification results among clinical centres. Which is why clinicians and researchers are turning to Medimsight.

From lab to specialist

There is a quirk in medical imaging software for analysis and quantification. There are many people working in this area, but many of the results have never been tested by specialists.

The research environment contributes to other problems to developing this software. For example, there is unequal funding, basic methodological development processes, prioritization of publication against the product. There is also another major problem. The software lacks feedback. 

If we developed a software program to automate the quantification of what can be obtained through any form of medical imaging, then we need to have a stronger collaboration with medical specialists. This can be difficult. There are reasons for this. Few people find the time to engage in research, and the healthcare sector isn't an exception. Medical specialists are faced daily with an increased number of patients.This means there is saturation of the service. As a result imaging services are collapsing. 

Interesting stats: In OECD countries between 2009 and 2014, the average MRIs per million population grows from 12.2 to 13.3 and CTs, from 22.8 to 23.6. 

And so on.

It is, therefore, easy to understand the problems in the system looking at the following table created using data from Google Scholar.






For practical purposes the following picture best describes what happens:


And all this, of course, not talking yet about CE or FDA, quality control within the pipelines, user interfaces, data security, monetizing or marketing. 

Is there a solution to this? Actually there is, but before reaching the goal, I would talk about the way to achieve this.

The Ad-Hoc software. An investment, a computer, a tool and a testing center where you plan to install. This model has been the most widespread in the past. Its limitations are obvious today. The software and the interface rarely see the light outside the center which is part of the project. Then, typically, the investment and the development team disappear.

Software as a plug-in. It seems an interesting evolution of the Ad-hoc as it seeks to simplify the problem of the specialist to work with dozens of user interfaces and add powerful and "easy" tools to install and to use. The problem is similar to the previous one. The basic software is not installed in the hospitals computers and even if it was,  90% of the hospitals don´t have enough power to perform certain tasks which can last hours.

Let someone else. This model is also quite prevalent today. I send my images (something specialists normally don´t like to do) and someone sends me the results. This model, although it may work, has several major problems. First, the security of data - At how many companies are going to be sending DICOM images? Once per image modality? Once per technique of analysis? In addition, there is something more important. What is the specialist involvement in the process? Do you really know what the companies are doing? The doctor or medical imaging specialist can, perfectly, understand the science of a publication in which software is based. So, why leave them out of the process? Why not let them choose which technique should be used?

Our solution is simple. A platform. 

Processing and storage in a single and secure location in the Cloud. A place where specialists and companies join into a market of methods. 

Easy to access for specialist to services continuously updated, no software downloads, no dependence on hardware and easy for the development team to integrate, distribute and monetize their analysis algorithm. 

That is precisely Medimsight.

Hello! Medimsight blog


MEDIMSIGHT LAUNCHES BLOG. WE ARE HERE, THE PLACE TO COMMUNICATE. MEDIMSIGHT NEWS AS WELL AS NEW PRODUCTS IN OUR CLOUD OF MEDICAL IMAGING, EVENTS AND ALL THOSE TECHNOLOGICAL INNOVATIONS RELEVANT FOR MEDICAL IMAGING.


On Medimsight we believe in communication as a way to keep up to date of scientific and technological developments affecting our specialties. That's why we love to share and listen. 


I invite you to be alert. Soon we will publish news. 


Greetings!