InkWood Research estimated the size of the artificial intelligence market in the healthcare industry at around $1.21 billion in 2016. We surveyed more than 50 executives from healthcare companies previously and laid out the state of AI applications in the healthcare space. It seems AI-based healthcare innovations have made their way into Asia, led by developments in China, India, Japan, and South Korea. Numerous companies claim to assist healthcare professionals in Asia with aspects of their roles, including assisting in diagnostics, remote caregiving, and improving a patient’s ability to manage their health using data from wearable devices.

We researched the space to better understand where AI comes into play in the Asian healthcare industry and to answer the following questions:

  • What types of AI applications are currently in use in the healthcare industry in Asia?
  • What tangible results has AI driven in the healthcare industry in Asia?
  • Are there any common trends among these innovation efforts? How could these trends affect the future of the healthcare industry in Asia?

This report covers vendors offering software across three applications:

  • Personal Health and Disease Management
  • Computer Aided Diagnostics (CAD)
  • Patient Fall Detection

This article intends to provide business leaders in the Asian healthcare space with an idea of what they can currently expect from Ai in their industry. We hope that this article allows business leaders in the Asian healthcare industry to garner insights they can confidently relay to their executive teams so they can make informed decisions when thinking about AI adoption. At the very least, this article intends to act as a method of reducing the time business leaders in the Asian healthcare industry spend researching AI companies with whom they may (or may not) be interested in working.

Personal Health and Disease Management

iCarbonX

iCarbonX is a Shenzhen, China-based company with 76 employees. The company offers an application development platform, which it claims can help software developers at healthcare businesses create health monitoring apps using machine learning.

iCarbonX seems to be taking an Amazon-for-healthcare approach and claims their platform can be used by healthcare businesses to provide their customers with wellness and health management apps. iCarbonX also offers their own health management application for individuals called Meum.

Individuals using Meum can monitor and record health information, such as diet, exercise, sleep, and emotional state. For instance, a user might record voice memos or photos of what they ate for a particular meal, and the software might automatically analyze nutrition intake and provide proactive advice in cases of nutritional imbalances.

The company is collecting medical information, such as DNA profiles and Fitbit-style wearable device activity data, frequent blood tests, heart data, and patient medical history, through alliances with healthcare startups like SomaLogic and HealthTell. iCarbonX claims it uses machine learning algorithms to find patterns in that healthcare data, such as the best preventative measures for a particular disease, or to create a personalized medical profile for each individual user. Developers can use the iCarbonX platform to deliver useful health tips and recommendations to customers. Meum users can receive personalized tips on fitness, skincare, and nutrition.

The company does not provide a video demonstrating how its software works.

Due to the nature of their platform, which is intended to help developers create health management and monitoring apps, we could find no case studies for iCarbonX in which businesses reported success with the company’s software.

iCarbonX does not list any clients on their website, although they have invested nearly $400 million in initial companies that form their alliance ecosystem to collect healthcare data. The alliance includes companies such as PatientsLikeMe, SomaLogic, HealthTell, AOBiome, GALT, Imagu, and Robustnique. The company has raised over $200 million in funding so far and is backed by prominent firms such as Tencent Ventures.

Jun Wang is founder and CEO of iCarbonX. He was previously the founder of the Beijing Genomics Institute (BGI). Wang earned a bachelor’s degree in artificial intelligence and a Ph.D. in bioinformatics from Peking University.

Computer Aided Diagnostics (CAD)

12 Sigma Technologies

12 Sigma Technologies is a Chinese company with 21 employees and offices in San Diego. The company offers medical data analysis software, which it claims can help hospitals and physicians make medical diagnoses using deep learning.

The company has partnered with several hospitals in Beijing, Suzhou, and Shanghai to gather labeled images that are used to train their deep learning image recognition software. Physicians can input an X-ray or CAT scan image into the software and scout for indications of many diseases simultaneously. The software analyzes the image for symptoms of diseases, such as lesions, by comparing the image to several thousands of scans from patients who have been diagnosed with certain diseases by human doctors in the company’s database. The system then provides a list of possible diseases and the areas of concern identified in these scans that might be due to a disease.

Below is a short 6-minute video demonstrating how 12 Sigma Technologies is using AI to detect lung cancer nodules:

12 Sigma Technologies signed a partnership with Intrasense to develop their deep learning-based Computer Aided Diagnostics (CAD) system, aimed at supporting physicians in analyzing and diagnosing lung cancer nodules from patient CT scan images. The partnership sees 12 Sigma Technologies using Intrasense’s Myrian Studio to develop the σ-Discover Lung, which the company claims may accurately detect possible diseases from patient scans. According to 12 Sigma Technologies, the company’s software could help physicians detect tiny lesions in patient body scans and generate structured reports.

We could find no robust case studies or reported success with 12 Sigma’s software. We were also unable to find any mention of enterprise-level companies on 12 Sigma Technologies’ website nor in any of their press releases, but they have raised $29 million in funding and are backed by SB China Venture Capital.

Dashan Gao is Co-Founder and CTO at 12 Sigma Technologies. He holds a PhD in Electrical Engineering from the University of San Diego. Previously, Gao served as Computer Scientist for  at GE Global Research and as Senior Staff Engineer at Qualcomm.

VoxelCloud

VoxelCloud is a Los Angeles-based company with offices in Shanghai and Suzhou. The company offers a medical image analysis software, which it claims can help physicians make more accurate clinical decisions using computer vision.

VoxelCloud claims its software uses a machine learning model trained on large datasets labeled by healthcare experts. Physicians can first upload a patient’s scans into the software. The software compares these scans to those within the software’s training dataset of images labeled as containing anomalies that indicate lung cancer, heart disease, and other conditions. Then, the software flags features on the imported CT scans as anomalous for a human doctor to review.

From 3:30 to 4:45 in the video below, CEO Xiaowei Ding explains how VoxelCloud could automate medical image analysis from computed tomography images and digital color images of the retina:

VoxelCloud does not make available any case studies reporting success with its software.

We were unable to find any mentions of enterprise level clients on VoxelCloud’s website. The company has raised over $28.5 million in funding and is backed by venture firms such as Tencent and Sequoia.

Xiaowei Ding is the Chief Executive Officer at Voxelcloud. He is a medical image analysis and machine learning researcher and holds a Ph.D. in Computer Science from UCLA in 2015.

Lunit

Lunit is a Seoul-based company with 44 employees. The company offers a software called Lunit INSIGHT, which it claims can help hospitals and doctors identify abnormalities in chest radiographs using computer vision.

Lunit claims doctors can use its INSIGHT software to detect lung cancer nodules from chest radiographs. The company claims the machine learning model behind its software was trained on more than 70,000 clinical test cases. Lunit INSIGHT compares the nodules in the radiograph to those of cancerous nodules in the training dataset to see if they match. The system then shows the location of the nodules in the form of a heat map and creates an abnormality score for each detected nodule.

Below is a short 2-minute video from CEO Anthony Paek demonstrating how Lunit Insight works:

Our research yielded no results when we tried to find case studies for the software.

We were unable to find any mention of enterprise-level companies on Lunit ’s website nor in any of their press releases, but they have raised $20.5 million in funding and are backed by Softbank Ventures Korea.

Anthony Seungwook Paek is CEO and co-founder at Lunit Inc. He holds a PhD in Electrical and Electronics Engineering from Korea Advanced Institute of Science and Technology.

Patient Fall Detection

FRONTEO Healthcare

Fronteo Healthcare is a Tokyo-based company founded in 2015. The company is developing a software tentatively called the Fall Prediction System, which it claims can help hospitals identify which patients might be more likely to fall down using machine learning.

With a growing population of people over the age of 65 in Japan, nursing homes and hospitals seem to have an urgent need to reduce instances of patients falling down and injuring themselves.

Fronteo Healthcare claims users can install their software in hospital computer systems. The software uses a patient’s historical electronic medical records to inform a predictive model based on the company’s own AI platform, called the Concept Encoder. The software then provides nursing staff with a priority list of patients that might need extra care due to their higher likelihood of falling down

According to Fronteo Healthcare, the software is currently undergoing trial runs in hospital settings at several medical institutions in Japan. These trial runs will inform the machine learning model behind the software.

We were unable to find any demonstrative videos for the company’s software or their AI platform.

We were also unable to find any mention of enterprise-level companies on Fronteo Healthcare’s website nor in any of their press releases, but they have raised $3 million in funding so far.

In addition, Fronteo Healthcare doesn’t seem to have any executives on its C-team with AI experience. We caution business leaders to be wary of companies claiming to do AI without AI talent on their executive teams.

Takeaways for Business Leaders in the Asian Healthcare Industry

AI vendors selling into the Asian healthcare industry seem to most commonly provide solutions for medical image analysis and diagnostic assistance. The existence of large troves of data in the form of images such as CAT scans and radiographs and the advancement in image recognition technology have likely fueled the increase in AI adoption across Asian healthcare companies.

iCarbonx seems to have the most traction of the companies discussed in this report in terms of funding, having raised over $200 million from various high profile investment firms, including Tencent and China Bridge Capital. The company seems to have attracted the most capital due to its big-picture vision to mine and gather detailed patient information through alliances and tie-ups in both China and the US.

Healthcare businesses and hospitals in Asia might expect to see many more AI applications in the next two to five years mainly targeted at improving the capabilities of doctors and physicians. Diagnostic assistance might be the most commonplace application for AI due to the presence of historical diagnostic records that perhaps make it easier to create databases of labeled images on which AI software can be trained.

Business leaders in healthcare could also expect to see AI assist in the analysis of medical data, such as electronic medical records and genetic or blood test results. That said, integrating AI software to automate the analysis of this type of data might be a long and arduous process involving cleaning and scrubbing a hospital’s data. Integration could take anywhere between three and six months, as well.

 

Header Image Credit: Enterprise Innovation