Release date: 2017-10-27 What can I do in 2 minutes? Outside of the phone, email, and screen, you may not have thought about detecting the number of white blood cells in your blood. In the NVIDIA Inception Artificial Intelligence Startup Competition in May this year, Silicon Valley-based startup Athelas took over the hot emerging project from NVIDIA CEO Huang Renxun, who used spectroscopy and machine learning to create a portable Hardware, 2 minutes to complete the blood test, trying to change the rules of the game of traditional blood testing. Figure Athelas builds a blood test hardware with spectroscopy and machine learning (Source: DT Jun) After the strong return of artificial intelligence, the application is full of flowers. In addition to many companies investing in video and speech recognition, there is also a field that is also popular, that is, medical care, from image-assisted diagnosis, health management and monitoring, medical insurance evaluation and pricing, Many cancer companies have come up with good concepts and practices in cancer research and new drug development. According to research firm CB Insights, nearly 70 AI-centric healthcare startups were financed last year, less than 20 transactions in 2012, and it is clearly seen that this sector is heating up rapidly. In fact, the annual medical-related expenditure is about 7 trillion US dollars, accounting for about 10% of the global GDP. Not only the startups are eager to try, but the technology giants are not willing to miss it. Apple CEO Cook at the 10th anniversary of the Apple iPhone in September this year. Point out: "The medical monitoring market is much larger than smartphones. This will be Apple's future big thing." In addition to the acquisition of companies such as Gliimpse and Emotient, ResearchKit applications have been developed for researchers to make large-scale use of the iPhone. the study. CareKit allows researchers and organizations to monitor patients instantly with sensors and tools on the phone. Lock in the two most costly diseases in the United States: diabetes and heart disease. Subverting traditional blood test routines If you are a fan of The Lord of the Rings, you should remember the word Athelas, a herb derived from Middle-earth that has healing properties. In reality, Athelas is a startup company in Mountain View, Silicon Valley. When you walk into the office, you can see a hand-painted design on the wall and a product that looks a bit like Amazon's smart speaker Echo. But the function is very different, that is, they developed devices for blood diagnosis. Figure 丨 Athelas's office wall is hung with a hand-painted design of the hardware device (Source: DT Jun) “We are a company that develops portable blood diagnostics. There are definitely a few questions in your mind, such as what does this mean? What do you do?†Athelas co-founder Deepika Bodapati said with a smile, “We focus on all blood cells. Complete Blood Count (CBC)." When we do a health check, the nurse will often draw blood for us. Simply put, CBC is a hematological examination, which is a common and routine test, in which the white blood cell value is an indicator. The importance of white blood cells is that it is a battle against bacteria in the human body. If the number of white blood cells is too low, it may cause infection symptoms. In addition, some causes or diseases usually lead to abnormal white blood cells, such as chemotherapy, leukemia, inflammation, etc. . Generally speaking, when we want to test blood for hospitals or inspection agencies, consumers have to pay for the cheaper costs in addition to waiting for the test results. In particular, medical insurance in the United States is expensive, and most likely limit the policyholders within one year. The number of tests (which varies depending on how much the individual pays for the insurance) makes it difficult for the patient to grasp the changes or even worsen in the first time. "When we are not very comfortable, we are used to putting a thermometer into our mouths, because body temperature is an indicator of physical condition, but why can't we master blood changes as easily as body temperature?" Deepika Bodapati stressed. Therefore, they use machine learning and computer vision technology to create a device that is as simple to use as a thermometer. Just pin your fingers, drop the blood on the exclusive test strip, put the test strip into the device, and then open the APP on your phone. After 2 minutes, you can see the white blood cell count (WBC), lymphocytes (Lymphocytes), The number of moderate white blood cells (Neutrophils). Figure 丨 APP will show the number and status of white blood cells, lymphocytes, neutrophils, and quite intuitive (Source: DT Jun) After the advent of the blood glucose machine, let the diabetics monitor their blood sugar at home. Athelas thinks the same thing. Let the blood test enter the home environment, it can help the general public to monitor the blood and regularly control their health, such as chemotherapy and leukemia. Or patients with neutropenia (Neutropenia) have this need. In addition to improving the efficiency of blood testing, the comparison is mainly based on traditional counters, flow based cytometry, and the cost of new testing methods is greatly reduced. Now the blood test at the hospital is analyzed using instruments such as Coulter Counter and Sysmex Counter (Sysmex is a medical instrument label). The cost of purchasing these instruments is usually about $250,000. It can be designed by AI and computer. A plan for parity. In addition to entering the clinic, the Athelas product also provides a subscription service on the official website. After the user orders, the device will be sent to the consumer's home for a fee of $20 per month. At home, you can get test results, save transportation time, improve the efficiency of health monitoring, and at the same time cost a lot of people. Can it be accurate to rely on a drop of blood? Athelas hardware takes only 3 simple steps: drop blood, put it into the machine, open the app, you can see the results, save time and cost, and really have a chance to subvert the traditional rules, however, Deepika Bodapati may have seen the doubts in my eyes, and she then asked me a question: "I guess you will ask the question next? Is this something feasible?" In particular, the blood startup company Theranos broke the big scam in history, claiming that a small amount of blood test can test more than 200 test results such as hemorrhage, cholesterol, diabetes, pregnancy, etc. within 40 minutes. The company's market value once reached a high of 9 billion US dollars. However, it was exposed because it could not get the data. This incident also dragged down the startup company engaged in blood testing, which was questioned by the outside world and affected the fundraising. “So we have published the data on the website from the first day, and we continue to do a lot of research and clinical testing. It is the most important thing for us to cooperate with schools and medical institutions.†Deepika Bodapati emphasized. For example, Athelas works with hospitals with high information throughput, such as FEMAP family hospitals. In addition, a comparative study with traditional detection methods was also carried out. The nurses used the traditional method to test blood samples from patients. After one minute, another blood was taken using the Athelas method to test the results. The results from 170 patients showed Athelas. It meets the standards of the Gold Standard device. The US Food and Drug Administration (FDA) classifies medical devices into three categories based on their use and risk level: Class I equipment is low risk and is not used to maintain patient life and does not endanger the patient's health, such as dental floss, Lenses, bandages, etc. Class II equipment is at higher risk than the first category and requires more supervision and control to ensure its safety and effectiveness, such as condoms, tampons, hearing aids, sphygmomanometers, etc. Class III equipment has the highest level of risk, mostly to maintain the patient's life and health, or to implant in the body, such as heart valves, artificial hearts, and so on. Athelas hardware is currently under review at the FDA Class II. However, through research with hospitals, it has been used in small amounts in the clinic environment and home users. Currently, 80 patients use this device every day to monitor their condition. DT Jun also interviewed the opinion of a professional physician. A radiation oncologist who did not want to be named pointed out that the focus still depends on whether it is FDA-approved. “If it is passed, it means that the test equipment is equivalent to the existing blood cell counter. "But from the perspective of business model, he is quite optimistic about Athelas's opportunity. "It should be successful in the United States, mainly because the United States is far-reaching. It is very early to go to the clinic and open the prescription, and the cost of the consultation is quite high." Expensive, this is a hard demand. The future 'detection' will be recognized as a formal medical record, just a matter of time," he said. The concept of convenience and low cost proposed by Athelas is indeed good, helping patients to manage themselves, and more effectively mastering the physical condition, coupled with the use of deep learning, telemedicine and other technologies to meet the trend, let Athelas in August this year A $3.7 million seed round funded by Sequoia Capital was completed, while Athelas' angel investors included Elged Gil, co-founder of genetic testing company Color Genomics, and Richard Socher, chief scientist of cloud services company Salesforce. Figure 3 shows the number of white blood cells in 3 steps Used to assist patients with cancer and cardiovascular disease The Athelas team is quite young, Deepika Bodapati and another founder, Tanay Tandon, are only 22 or 20 years old, and complement each other with different professions. Tanay Tandon is a member of the Stanford Artificial Intelligence Lab (Stanford AI Lab), specializing in machine vision, machine learning. In addition to AI, Stanford University is also well-known for its use in molecular imaging research for cancer diagnosis. Deepika Bodapati is a researcher from the Stanford Multimodality Molecular Imaging Lab (MMIL). Figure 丨Athelas two co-founders Tanay Tandon (left), Deepika Bodapati (right) (Source: DT Jun) In fact, Tanay Tandon first founded the company in 2014, when he was only 17 years old, when his idea was to design a special microscope set to be attached to the phone lens, using machine learning and computer vision algorithms from the blood. I recognized the first place in the YCombinator Hackathon competition, but it was difficult at the time, and he thought that it should make the equipment and services more professional, so I adjusted the direction a year and a half ago. Switch to a stand-alone device and work with healthcare organizations. For the key elements of the design of the device, Tanay Tandon pointed out that the core of the hardware is the actuation system, using the Gaussian edge autofocusing algorithm to establish a standard method of capturing cells, while at the same time, the US disease control The data collected by the Center for Prevention and Prevention (CDC), the blood data collected by researchers at Stanford University and the University of California, San Francisco (UCSF), were compiled into training data sets, and computer vision and deep learning were used to identify and classify cell types. Deepika Bodapati is responsible for developing a rapid staining mechanism. When contacted with blood cells, it not only allows the cells to be imaged, but also highlights their type and makes the machine "understand". Unlike traditional counters and flow cytometers, Athelas uses deep learning as the core of the system (Source: DT Jun) In addition to working with oncologists, Athelas is using equipment to help cancer patients monitor the progress of chemotherapy, and in the future it is more likely to be used in the prevention of heart disease. In the United States, more than 2,000 people die of heart disease or stroke every day. Cardiovascular disease is the number one killer in the United States. In the past 10 years, a large number of studies have shown that heart attacks are not entirely due to high cholesterol. The main culprit in which fat blocks the blood vessels and causes the largest blockage of the arteries is inflammation. "This is not our imagination. Many literatures point out that heart disease is associated with white blood cells." Tanay Tandon pointed out that two or three days before the outbreak of heart disease, the body's inflammatory condition will come to a peak, which is a warning, and heart disease is often more common in the home than in the hospital, so if the user can check the blood at home The number of white blood cells in the clinic can actually detect that the body has begun to go out of the situation, and must go to the hospital for examination and medication, so as to prevent heart attacks, strokes and other regrets. For Athelas' future plans, Tanay Tandon points out that they are working hard to integrate more features into the system, such as inflammation tracking, urinary tract infections, more cell counts, etc., which need to work with existing clinical medicine communities. One goal is to bring equipment to more care stations, homes, and hospitals. Source: DeepTech Deep Technology (micro signal mit-tr) Video Doorbell,Intercom Video Doorbell,Sd Card Video Doorbell,Camera Doorbell Shenzhen Zuomi Technology Co., Ltd. , https://www.bkvis.com