AI TAM

Part
01
of eight
Part
01

Global TAM, AI/Machine Learning in Healthcare

The current (2019) global TAM for AI/Machine learning in healthcare is between $3.15 billion and $3.16 billion. We have entered the detail in row 2, column B of the attached spreadsheet as requested.

GLobal and Machine Learning

  • The market size for AI and machine learning in healthcare in 2018 was $2.1 billion, according to market research by Report Linker.
  • Report Linker notes that the market will grow at a CAGR of 50.2% between 2018 and 2025 to reach $36.1 billion. That means that the market will be worth $3.15 billion (Calculation was done using an online CAGR Calculator) in the current year.
  • Machine Learning will account for the biggest market share in the healthcare AI/Machine Learning market.
  • However, Research and Market estimates that the market will be worth $27.6 billion in 2025, growing at a CAGR of 43.5%. This means that the market was worth $2.2 billion in 2018 and $3.16 billion in 2019 (calculation was done with the aid of a Reverse CAGR Calculator).
Part
02
of eight
Part
02

US TAM, AI/ML in Healthcare

Recent data from MarketWatch indicates that the healthcare artificial intelligence market had a 2016 value of $320 million; that market is estimated to grow at a CAGR of 38%. Using this data, the estimated current TAM for healthcare artificial intelligence and machine learning in the United States is $849.98 million. The entire project is here.

US TAM, AI/ML in Healthcare

  • The North American segment of the healthcare artificial intelligence (AI) market dominated the overall industry, with a 2018 value of over $653.9 million in 2018. Global Market Insights expects a lucrative CAGR through 2019-2025 in the North American segment.
  • A report from Markets and Markets echoes the findings of Global Market Insights. The former expects that the wide-scale of adoption of AI in healthcare, especially in the United States, will be a driving factor in the growth of the North American AI healthcare market.
  • The strong presence of key AI hardware and software providers (e.g., NVIDIA, Intel, Xilinx, Microsoft, AWS, Google, IBM, Johnson and Johnson, and GE) complement North American market growth. The North American AI healthcare market is also seeing active participation from start-up organizations; such as, CloudMedx, Imagia Cybernetics, Precision Health AI, and Cloud Pharmaceuticals.
  • According to data from MarketWatch (provided by Market Study Report), the AI and machine learning healthcare market had a 2016 value of $320 million in the United States. Market Study Report expects that market to see a CAGR of 38% over the forecast period.
  • Using this data, the calculated current United States TAM for AI and machine learning in healthcare is $840.98 million.

Research Strategy

We began our inquiry with an investigation into databases and industry sites (e.g., the Financial Post, The Economic Times, The New York Times, PR Newswire, et al.) looking for industry reports, market research, and news releases. Our investigation uncovered the dominance of the North American healthcare AI market and its 2018 value ($653.9 million). Much of the data we uncovered focused on the global marketplace, but we did find data on the United States market. According to that data (provided by MarketWatch), the AI healthcare market size in the United States had a 2016 value of $320 million; the market was expected to see a CAGR of 38% over the forecast period. We used this data to calculate the 2019 TAM using a reverse CAGR calculator. Our starting value was 320,000,000; the CAGR was 38%; and the number of periods was 3. This yielded a value of 840,983,040.
Part
03
of eight
Part
03

Past US TAM, AI/ML in Healthcare

The estimated 2014 TAM for AI and machine learning in healthcare in the United States was approximately $123 million. Our findings on this topic have been added to column C of row three of the attached spreadsheet.

U.S. TAM FOR AI AND MACHINE LEARNING IN HEALTHCARE (2014)

  • North America is the region of the world that is likely to witness the highest CAGR between the years 2019 to 2025
  • The United States is expected to be a significant contributor in the AI-in-healthcare market within North America.
  • In 2016, the U.S. TAM for AI and machine learning in healthcare was $320 million.
  • Between 2014 and 2016, the U.S. TAM for AI and machine learning in healthcare is expected to grow at a CAGR of about 38%.
  • In 2014, the estimated U.S. TAM for AI and machine learning in healthcare was up to $123 million. (Calculated)

Research Strategy:

Our research began by checking statistical sites, research firms, and media sites for any reports on the U.S. TAM growth for AI and machine learning in healthcare in 2014. Thus, we explored research firms such as Research Gate, Docs Player, Technavio, Industry Daily Observer, Grand View Research, Business Wire, and Globe-News Wire for relevant information. We also checked statistical and consulting websites, including Statista, Deloitte, KPMG, and McKinsey. Also, we reviewed media websites like Reuters, Forbes, and The NYT.

On the website Pymnts, we discovered that North America is leading the global market of AI and machine learning in healthcare with the United States serving as one of the major contributors to the region’s AI-in-healthcare market. Additionally, while exploring Market Watch, we found that the U.S. market size for AI and machine learning in healthcare was valued at around $320 million in 2016, and the market was anticipated to grow at a compound annual growth rate (CAGR) of about 38% until 2024. We decided to utilize this data to calculate the 2014 TAM, assuming that the market had grown at the same CAGR between 2014 and 2016 (38%).

To calculate the 2014 market size, we used the following equation:
  • Market size for previous year = market size of next year - (market size of next year x CAGR)
Thus,
  • 2015 market size = 320,000,000 - ($320,000,000 x 38%) = $198,400,000 or $198.4 million
  • 2014 market size = 198,400,000 - ($198,400,000 x 38%) = $123,008,000 or $123 million
Therefore, we estimate that the TAM for AI and machine learning in healthcare in the United States in 2014 was roughly $123 million.
Part
04
of eight
Part
04

Past Global TAM, AI/ML in Healthcare

The global TAM (or market size) for artificial intelligence (AI) in healthcare was $600 million in 2014. We have entered it into row 2, column C of the attached spreadsheet.

2014 Global TAM for AI in Healthcare

  • Machine learning is an application of AI. So to find the global TAM, we found the market size for AI only, since machine learning would be included.
  • The 2014 global TAM for AI in healthcare was $600 million.
  • At the time of this 2017 report, the expected CAGR through 2021 was expected to be 40%. This likely changed with more recent reporting.

Research Strategy

We began our research by getting an understanding of what Total Addressable Market (TAM) is. We discovered that TAM is very personalized for each company based on the product they are offering, the price they will be charging, and the specific market they are trying to reach. Since we are looking at the global TAM, not in terms of a particular product offering, we are using market size as a stand in for TAM. To further confirm the validity of this strategy, Divestopedia states, "Total addressable market is also known as global total market, even if the proposed business is unable to serve some parts of this market." Global total market is used in this reporting.
Part
05
of eight
Part
05

Projected Global TAM, AI/ML in Healthcare

The projected global TAM for AI and machine learning in healthcare for the next five years (2024) is between $19.21 billion and $24.11 billion. We have entered the detail in row 2, column D of the attached spreadsheet as requested.

GLOBAL AI and MACHINE LEARNING in healthcare

  • The global market size for AI and machine learning in healthcare in 2018 was $2.1 billion, according to market research by Report Linker.
  • Report Linker notes that the market will grow at a CAGR of 50.2% between 2018 and 2025 to reach $36.1 billion. That means that the market will be worth $24.11 billion in 2024 (calculation was done using an online CAGR Calculator).
  • However, Research and Market estimates that the market will be worth $27.6 billion in 2025, growing at a CAGR of 43.5%. This means that the market was worth $2.2 billion in 2018 and will be worth $19.21 billion in 2019 (calculation was done with the aid of a Reverse CAGR Calculator).
Part
06
of eight
Part
06

Projected US TAM, AI/ML in Healthcare

The projected US TAM for AI and Machine learning in healthcare in 2024 is about $4.2 billion. We have entered the detail in row 3, column D of the attached spreadsheet.

US AI/MLTAM Projected

  • The US market size for AI and Machine Learning in healthcare was worth about $320 million in 2016.
  • The market is expected to grow at a CAGR of 38% until 2024.
  • If the market was worth $320 million in 2016 and was growing at a CAGR of 38%, it will mean that the market will be worth about $4.21 billion in 2024 (calculation with the aid of an online CAGR calculator).
Part
07
of eight
Part
07

Leading Bespoke AI/Machine Learning Consultancies

Six of the top consultancies that offer AI and machine learning bespoke support have been selected, and their services described. Their website has also been included as a link in their title. The six companies are given below:

six of the top AI/machine learning consultancies:

  • InData Labs
  • Arcanys
  • Eleks
  • MicroscopeIT
  • Sigmoidal
  • LeaderGPU

1. InData Labs

  • InData Labs is an AI consultancy that uses machine learning to assist companies in developing AI solutions and products.
  • The company provides services such as design, implementation, and integration of AI solutions into the customer's ecosystem.
  • It has extensive expertise in machine learning, natural language processing, computer vision and predictive modeling.
  • Aside from Data Science and AI Solutions, InData Labs offers also services of Data Strategy, BI Implementation, AI software development and Big Data Solutions.
  • AI software development includes building a custom AI solution for the client from end to end. This means working with the client through various steps from discovery, project setup, development, live release and support.

2. Arcanys

3. Eleks

  • Eleks' services are divided in three categories: advisory, engineering and optimization.
  • The company has expertise in Data Science, using various methods such as AI and machine learning to help clients solve problems.
  • Services offered in Data Science include analysis and consulting, roadmaping, modeling and deploying.
  • Other services include product design, application development, PoC development and application re-engineering.

4. MicroscopeIT

  • Microscope IT use their expertise combined with AI and machine learning to process data from the client and deliver products and services, automation and enhanced processes, and valuable findings for research.
  • It is possible for them to provide unique technologies for clients.
  • The company is specialized in custom software development, AI and cloud consulting.
  • Other services include the ability to conduct R&D projects, in such fields such as machine vision, machine learning and microscopic image processing.

5. Sigmoidal

6. LeaderGPU

  • LeaderGPU is a company that provides GPU-power to satisfy needs from powerful computing and machine learning.
  • However, the company offers consulting services, with the possibility to build, develop and deploy machine learning technology that can assist businesses in finding solutions to complex data issues.
  • They can develop custom AI solutions for clients and satisfy their machine learning requirements.
  • This company can deliver bespoke AI projects and help clients choose what kind of AI suits them for their specific issue.
  • LeaderGPU can build specific AI solutions according to clients machine learning strategies.
  • It can also provide support solutions that can help with ongoing analytics.

RESEARCH STRATEGY

To satisfy the criteria of top defined as well-known in the industry, we gathered various rankings of top consultancies in the field of AI/machine learning. Six rankings were been found, from six different sources, listing companies. The companies that appeared the most in those six rankings have been picked.





Part
08
of eight
Part
08

AI/ML Use Cases

Artificial intelligence (AI) and machine learning (ML) technologies have seen rapid growth in various sectors of the global healthcare industry. AI and ML technologies are helping healthcare providers in several facets of the industry from cutting costs on health operations to improving the overall quality of patient care. AI technologies are currently making great headway in boosting efficiency, improving outcomes, influencing healthcare innovations, and impacting every aspect of healthcare. Some key healthcare AI/ML use cases include research and clinical decision support, medical imaging, clinical trial matching, and medical knowledge management.

Use Case #1: Clinical Decision Support

  • It is known that medical diagnosis is a highly collaborative and intricate process that is susceptible to inconsistencies and errors affecting everyone involved, from hospital staff to patients. To help reduce the number of diagnostic errors and improve the overall process of diagnosis, AI-powered tools, programs, and platforms are being used by clinicians and physicians. The primary aim of using AI in clinical decision support is to provide better care plans and deliver better treatments by reducing diagnostic errors.
  • According to the Director of Research Strategy and Operations at MG/BWH Center for Clinical Data Science, Dr. Katherine Andriole, AI can be a valuable asset in helping physicians make accurate decisions based on patterns identified from a large collection of digital data. AI systems can analyze vast amounts of data in seconds and organize them in a coherent manner to help physicians quickly connect the dots in the treatment of patients, especially in the ICU.
  • In AI-based clinical decision support systems, the platform (computer) behaves as a knowledge base that generates patient-specific recommendations and assessments which are then presented to researchers or clinicians for further consideration. The AI system can offer support and provide necessary recommendations in real-time at the location where the healthcare profession is making the clinical decision. This helps the clinician to provide the patient with the best treatment possible.
  • Cognitive computing systems that utilize machine learning, natural language processing, and other AI technologies, are innovative AI systems that synthesize data to provide a simple answer for a particular question. These systems are mainly used in the field of oncology where AI recommendations are always evaluated by a physician. One such cognitive computing system is IBM’s Watson Oncology.
  • Watson Oncology was designed and developed by IBM in collaboration with Memorial Sloan Kettering Cancer Center as a cognitive computing application. The product is trained with over 13 types of cancer.
  • During a pilot test, IBM Watson was provided with an algorithm that searched across note sections, clinical records, and various other data to learn the key predictors and markets for congestive heart failure. Watson was able to analyze 21 million patient records in a span of six weeks and developed a code that achieved a risk identification accuracy rate of over 85%.
  • Various companies operating in the healthcare industry use AI in clinical decision support such as IDx, Apple, and Aidoc.
    • IDx is a Coralville-based healthcare firm that has designed an AI diagnostic system enabling physicians to identify diabetic retinopathy by analyzing a batch of retina images. In a single minute, the AI tool identifies the eye condition and provides recommendations for the physician. The tool has been approved and used in the University of Iowa Hospitals and Clinics.
    • Apple has developed AI sensors that are connected to its flagship wearable to help identify atrial fibrillation in customers. This AI-based application is used by customers under the age of 55.
    • Aidoc, located in Israel, has designed an AI-powered brain solution that detects cases of internal bleeding. This clinical decision support system utilizes AI and deep learning to assist radiologists in cases of acute intracranial hemorrhage.

Use Case #2: Medical Imaging

  • Medical imaging is found to be one of the most complex sources of patient information. AI-powered tools that are integrated with medical imaging provide physicians with the required support to precisely diagnose specific conditions. According to Dr. Katherine Andriole, imaging analytics is a common field of AI application to help enhance physicians’ decision-making capabilities.
  • It is found that even the most experienced medical professionals have difficulties in manually analyzing high-resolution images in large numbers. AI-driven systems in medical imaging can compute a vast number of images and provide analysis from CAT scans, X-rays, MRIs, and various other testing modalities, thereby, aiding clinical professionals in improving their accuracy in identifying conditions and providing treatments.
  • Algorithms are written to accurately and quickly diagnose conditions by reviewing millions of images. For example, an algorithm for a medical imaging tool can be written to look through a vast number of chest films, with malignancies and without, in order to learn to spot malignancies. It is to be noted that these AI tools do not provide the end solution or make decisions but provide recommendations for the physician to make the final call.
  • Some specific uses of AI in medical imaging are to identify cardiovascular abnormalities, to detect fractures and other musculoskeletal injuries, and to aid in the diagnosis of neurological diseases.
    • Identifying cardiovascular abnormalities: AI medical imaging tools are used as an initial screening tool to detect abnormalities in imaging tests that are commonly ordered such as x-rays. In this case, algorithms are written to automatically develop reports, highlighting the different measurements that are considered abnormal, thereby, saving time for clinical professionals.
    • Detecting fractures and injuries: AI imaging tools use algorithms that review various images, particularly in trauma patients, to help identify and account for every single injury. This AI application helps reduce patient risk, false negative risks, and medical-legal risks for radiologists.
    • Aiding neurological disease diagnosis: In this case, AI tools use algorithms that streamline the entire diagnosis process by identifying images that show evidence of amyotrophic lateral sclerosis (ALS). These algorithms can be written to provide automatic reports to reduce the workflow burdens on healthcare providers.
  • According to a report published by Signify Research, the AI market in medical imaging is set to experience rapid growth and is expected to generate $2 billion in revenue by 2023. This particular market is driven by specialist software developers and start-ups that are using machine learning and AI technologies to further develop medical imaging. Some notable companies operating in the healthcare industry that use AI-powered medical imaging tools are Tencent (China), Alibaba (China), and IQRAA Hospital (India).

Use Case #3: Clinical Trial Matching

  • In the field of oncology, cancer clinical trial coordination is found to be extremely difficult and most often the numbers are not in favor of the patients or the medical staff. In the United States, less than 5% of those diagnosed with cancer participate in clinical trials and 20% of these trials are “stymied due to lack of patient recruitment”.
  • There is a considerable disconnect in the recruitment of patients in clinical trials due to the daunting process of data-matching. Identifying patients with the correct clinical attributes and who meet the right eligibility criteria is a lengthy and complex process. Manual recruitment process, imprecise criteria, and unstructured healthcare data are the main reasons for the “cumbersome clinical trial identification”.
  • To solve this problem, AI-powered screening processes are used to help increase the speed of identifying accurate clinical trial matches. In this case, AI-powered matching solutions are able to compute vast amounts of structured and unstructured data pertaining to patient EHR information, trial data, medical literature, and eligibility criteria. These AI tools are developed using natural language processing which allows the AI system to examine all the gathered information in a short period of time.
  • Clinical trial matching AI systems can take structured or unstructured, lengthy, and inexact criteria information and learn to interpret the requirements of the clinical trial, specific to each patient. Therefore, an AI-based system will be able to quickly understand the exclusion or inclusion criteria of the clinical trials and the medical record’s clinical data. In a particular study conducted by IBM Watson Health, AI-powered identification reduced the screening of patients for the eligibility of a clinical trial by 78%.
  • With the use of AI matching/identification systems, clinicians will be able to identify trials that can be considered as treatment options and can continuously analyze available patients for clinical trial eligibility. In the United States, the adoption of AI-based tools in clinical trial recruitment is found in research and academic medical centers.
  • According to Forbes, The Opportunity Project (TOP) team created an AI-based system that matches patients to clinical trials, this year (2019). This initiative is a collaborative effort of a number of nonprofits and technology providers and was headed by the Presidential Innovation Fellows and the United States Health and Human Services Office of the Chief Technology Officer.
    • The first stage of this process includes a chatbot that answers questions asked by patients about their cancer treatment. The chatbot provides necessary background information on the subject through a voice interface of conversational text.
    • In the second stage of this process, the patient agrees, through the chatbot, to email his/her doctor for further medical information that is needed for a clinical trial. A clinician enters the provided information into the TOP app by following a guided process. The app is designed to find clinical trials that the patient is eligible for.
    • In the final stage of this process, the patient obtains the information in the form of an interactive map that shows where the clinical trial is located and allows the patient to read more about the clinical trial by clicking on the map.
  • London-based company, Antidote, and California-based firm, Deep 6 AI, are two companies that have developed AI-based patient recruitment systems.

Use Case #4: Medical Knowledge Management

  • Numerous companies, healthcare firms, hospitals, clinics, and medical universities are leveraging AI technologies to develop better healthcare data management and medical knowledge management systems. A popular form of medical knowledge management powered by AI is virtual reality and augmented reality platforms that are designed to educate doctors, nurses, surgeons, and medical students in conducting surgical procedures.
  • In Japan, there are 12 hospitals that use VR modules from technology provider, HoloEyes, to view three-dimensional models of organs like kidneys and livers. The VR modules allow medical professionals to accurately plan the finer details of critical surgical procedures. Another Japanese company, Intuitive Surgical, has developed a VR simulation training module for medical professionals to help train on certain procedures.
  • Johnson & Johnson Institute, an Ohio-based establishment, has developed a VR platform that helps train surgeons, doctors, nurses, and students in a large selection of surgical procedures. This platform utilizes deep learning technology that allows medical professionals to experience and learn real-world scenarios without actually being present in a physical training room.



Sources
Sources

From Part 04
From Part 07