DeepTech News Briefs
Researchers at Iowa State University designed an AI system to create personalized prosthetic aortic heart valves. These customized valves can restore normal blood flow for people with aortic valvular disease. Over 90,000 people in the US have valve replacement surgery every year.
Traditional drug discovery is a very long and expensive process involving many tests to determine the safety and efficacy of each new drug candidate. AI is making the hunt for new drugs quicker, cheaper and more effective. Drug companies are already conducting clinical trials for drugs developed using AI, so although no AI for Drug Discovery companies have brought a drug to market yet, we could see the arrival of the AI industry’s first blockbuster drug as early as 2020.
An analysis of 47 biomedical unicorns found that most of the highest valued startups in healthcare have very limited or non‐existent participation in the publicly available scientific literature. The system of peer reviewed publishing, while imperfect, plays a role in validating innovative products and technologies in biomedicine. Healthcare products not subjected to peer review may be problemati
Although Pharma companies spend over $172 billion on research and development annually, over 90% of molecules discovered using traditional techniques fail in human clinical trials, and 75% of newly approved drugs are unable to cover the cost of development. BioPharma companies capable of building strong AI divisions and acquiring the best AI startups will dominate the biopharma industry and see their market capitalizations skyrocket.
On the entire continent of Africa, if you remove Egypt and South Africa, there are only 6 pediatric radiologists. In Liberia there are only 2 radiologists. Fourteen African countries have no radiologists at all. One hospital in Boston, Massachusetts General Hospital, has 126 radiologists. Researchers are working towards extending the reach of care outside of hospitals and clinics. AI will help increase access to care in places where radiologists are inaccessible.
Researchers in Korea analyzed literature evaluating 516 AI algorithms for medical image analysis and found that only 6% validated their AI and 0% were ready for clinical use. This lack of appropriate clinical validation is referred to as digital exceptionalism.
An interactive list of 100 innovative AI leaders who initiating data-driven transformation in the pharmaceutical and healthcare industries. This interactive list is based on assessment of the cumulative impact that each person is contributing to the advancement of AI in the areas of pharmaceutical and healthcare research. The overall success of AI transformation depends on highly skilled interdisciplinary leaders who have the ability to innovate, organize and guide others.
Although it's estimated that 65% of the hundred biggest innovations of our time are really based on small data, current AI developments focus mostly on big data, forgetting the value of observing small samples. This article is about the role of small data in the future of AI. Efforts have already begun in this direction. Although the current mantra of deep learning says “you need big data for AI”, more often than not, AI becomes even more intelligent and powerful if it has the capability to be trained with small data. Some AI solutions that rely only on small data outperform those working with big data
One of the sectors most impacted by DeepTech is AI for Drug Discovery. In this article we profile four top tier companies in this sector: Insilico Medicine, Cyclica, Healx, MindsAI. In contrast to other areas of human endeavor, drug discovery has not become faster and cheaper with time. Using AI for drug discovery could help change this.
Some of the industries most impacted by DeepTech include life sciences, biotech, biomedicine, and Longevity. DeepTech companies are fundamentally different from regular tech companies. Instead of being based on innovative business models, they’re based on cutting edge science and technology that have the potential to cause major disruption across industries. DeepTech companies profiled: The Buck Institute for Research on Aging, Leucadia Therapeutics, Oisin Biotechnologies, Elevian, Insilico Medicine.
This AI algorithm can diagnose 55 childhood diseases with 90 to 97% accuracy. This organ-based approach can outperform most doctors at detecting life-threatening conditions like meningitis and sepsis.
This white paper describes criteria the FDA proposes to use to determine when AI medical products will require FDA review before being commercialized. The White paper calls for proof in a clinical real world environment. AI algorithms will not be approved on the basis of computerized dataset analysis.
Scientists at the AI Precision Health Institute at the University if Hawai‘i Cancer Center are using AI to improve the diagnosis and treatment of cancer. They're collaborating with doctors in underserved communities, enrolling underrepresented populations in clinical studies, and making contributions to science that will benefit our generation and future generations. A prime example of DeepTech for Social Good.
A new global AI hub officially launched today at Stanford University. The Stanford Institute for Human-Centered AI seeks to become an interdisciplinary global AI hub and to fundamentally change the field of AI by integrating a wide range of disciplines and prioritizing true diversity of thought.
In India over 100 doctors and engineers have collaborated to develop 40 affordable medical devices to improve healthcare for a billion people. The Biomedical Engineering Technology Incubation Centre at Indian Institute of Technology, Bombay is a center of healthcare innovation.
This software that originated in a Stanford basement is now one of the top AI solutions. Arterys AI software analyzes cardiac MRI exams to show the heart in 7 dimensions: 3 in space, 1 in time, and 3 directions of velocity to show if blood is flowing through the heart the way it should or if there are anomalies that require surgery. Arterys is the first company to receive FDA clearance for an AI algorithm in the cloud.
These 100 AI leaders are initiating data-driven transformation in the pharmaceutical and healthcare industries. The overall success of AI transformation in healthcare depends on highly skilled interdisciplinary leaders who contribute to the advancement of AI in pharmaceutical and healthcare research and also have the ability to innovate, organize and guide others.
The level of sophistication used in due diligence should be on a par with the level of complexity in a given industry. Pharma AI companies are 100 times as complex as FinTech companies. Methodologies used to assess them should be 100 times as rigorous. Deep Knowledge Analytics uses multiple parameters. Early stage startups are assessed using 100 parameters. Advanced stage companies are assessed using more than 300 parameters.
Scientists have printed the first 3D living heart made of human tissue and biological molecules. Using tissues from the patient to make a personalized “ink,” a specialized printer was used to produce a working, pumping human heart complete with blood vessels. Because the engineered heart is made of the patient’s own tissues, transplanting the printed organ should keep the autoimmune system from rejecting the new heart.
This is the first time that the FDA has granted Breakthrough Designation for AI in cancer diagnosis. This designation will expedite product development and provides priority regulatory review for Paige.AI’s pioneering clinical-grade AI in pathology software.
New research suggests that loss of blood vessels in the retina could signal Alzheimer’s disease. On the left, the retina of a healthy person shows a dense web of blood vessels. On the right, the retina of a person with Alzheimer’s disease shows areas in blue where blood vessels are less dense. The study was published March 11, 2019 in the journal Ophthalmology Retina.
Deep Knowledge Analytics uses multiple parameters and applies quantified metrics to perform deep comparative analysis to differentiate levels of maturity, business development, scientific advantages, and technological levels in a very objective way. Early stage startups are assessed using 100 parameters. Advanced stage companies are assessed using more than 300 parameters.
AI is revolutionizing the drug industry by cutting development time. AI can save two years in R&D time says Alex Zhavoronkov CEO of Insilico Medicine. Insilico Medicine is turning its focus to China, as it moves its headquarters from US to Hong Kong.
AI algorithms are as accurate at analyzing slides as a human pathologist. That's good because the number of pathologists is declining and increasing disease in an aging population will lead to a deficit of pathologists.
Saeed Hassanpour PhD and his team at Geisel School of Medicine at Dartmouth developed an AI method that improves grading tumor patterns and subtypes of lung cancer and can classify lung cancer subtypes as accurately as a pathologist in less than a minute. They plan to apply this method to analyze images in breast, esophageal, and colorectal cancer.
Researchers at the AI Precision Heath Institute at the University of Hawaii Cancer Center are using AI to improve the diagnosis and treatment of cancer for hundreds of thousands of people who live in underserved communities on islands in the Pacific.
One big challenge with ML systems is keeping them accurate in production since they start deteriorating as soon as they are deployed in the real world. Although concept drift has been heavily studied for over 20 years, it is often ignored outside of academia. For example, ML models can change in different ways even when deployed in different buildings within the same hospital
AI researchers at Carnegie Mellon University developed a highly accurate active learning technique called MedAL that can diagnose disease using much less data. MedAL achieved 80% accuracy detecting diabetic retinopathy - using only 425 labeled images - which is a 40% reduction compared to random sampling. MedAL can also be used to detect skin cancer and breast cancer.
Samuel Finlayson and his team at Harvard Medical School fooled AIs into misclassifying images by altering a few pixels. Rotating images can also confuse AIs. Although AI promises to improve healthcare by quickly analyzing medical scans, there is evidence that it trips up on seemingly innocuous changes. The paper published in Science March 22, 2019 outlines motivations that various entities may have to use adversarial attacks and begin a discussion of what to do about them.
China-based pharmaceutical companies betting big and small biopharma able to quickly innovate will drive the use of AI for drug discovery – a market some analysts predict will reach a valuation of $20 billion by 2024.
This article is part of a DeepTech Series by Margaretta Colangelo and Dmitry Kaminskiy. Please click the subscribe button at the top of this article to have articles in our DeepTech series delivered directly to you each week.
Margaretta Colangelo, Managing Partner at Deep Knowledge Ventures, is based in San Francisco. Margaretta serves on the Advisory Board of the AI Precision Health Institute at the University of Hawai‘i Cancer Center. @DeepTech_VC
Dmitry Kaminskiy, General Partner at Deep Knowledge Ventures, is based in London. Dmitry is Managing Trustee of the Biogerontology Research Foundation.
Deep Knowledge Ventures is an investment fund focused on DeepTech. Investment sectors include AI, Precision Medicine, Longevity, and Neurotech. Deep Knowledge Ventures led Insilico Medicine’s seed funding round in 2014 and has remained a close advisor in the company’s journey towards becoming a global leader in the application of advanced AI, particularly deep learning and GANs.