Artificial Intelligence (AI) and Machine Learning (ML) are revolutionizing the industry and are radicalizing conventional workflows. The abundance of data is the key to AI and ML to work effectively. Medical science is producing a large amount of data every day from research and development (R&D), physicians and clinics, patients, caregivers etc. These data sets can be used to feed Machine Learning models to develop promising solutions in medical science worldwide.
Use of such advanced technologies in health and medical sciences is recasting prevention, diagnosis and treatment processes.
The future seems bright as many companies, institutes and startups have started working and investing into technologies like AI and ML to devise solutions and cures, even for life-threatening diseases like cancer which still remains uncontrollable beyond a certain stage.
Recent case studies conducted over Xlpat are a testimony to the same.
XLPAT is a self-learning patent research tool which uses AI and ML algorithms to automate the standard workflow of IP searches.
A thorough search conducted on XLPAT highlights the futuristic scope of AI technologies in Cancer research by extracting deep insights from raw patent and non-patent data. While analyzing data, we curiously verified some futuristic concepts of Cancer treatment on Novelty Checker (an AI-based app to verify Novelty of Ideas) and were amazed to see that many are already putting the effort in this direction.
The IP and Market trends in this sector are acting as a catalyst to motivate the development of innovations and are also a driving force for both, emerging players and experts to invest into ML centric cancer research specifically.
The efforts and developments currently being done in this direction can be summarized in 5 segments.
Breakthrough Innovations in cancer research:
Machine learning and related technologies are being broadly applied to the field of oncology to identify malignancy causing tumours.
With collaborations between research institutes and tech companies, this field still awaits to be explored.
However significant innovations are being seen around the world and IP market trends are a testimony to the same.
As seen from the patent data, implementation of concepts such as Machine Learning, Deep Learning and Neural Networks for Cancer Research and Therapy picked up pace 2015 onwards.
XLPAT – the Patent trend in AI technologies in Cancer Treatment
The above trend indicates that the recent developments in Machine learning and an increase in data processing power of the machines helped organizations to foresee the scope of AI technologies in cancer treatment.
During market research, a few of the future concepts to diagnose malignant tumours came into limelight. We inquisitively tried one of the concepts of “A mobile application which uses artificial intelligence to detect malignant tumour with a single photograph” on Novelty Checker (NC).
The automated report generated by Novelty Checker found some overlapping patents with the above-mentioned concept which confirmed that the organizations are keen to tackle un-catered problems with emerging technologies.
- KR101628990B1 Mobile Breast Cancer Tactile Detection Device, System and Method – Keimyung University
- US9241663B2 Portable Medical Diagnostic Systems and Methods Using A Mobile Device -Jana Care Inc
- IN201823028816A A Digital Device Facilitating Body Cavity Screening and Diagnosis – Ramasubramanian
- CN106570848A Computer-aided Detection Method for Breast Carcinoma Calcification Point Based On Local Binary Pattern And Support Vector Machine – University Zhejiang
Fig 1: Novelty Checker – Sub technology areas related to “AI in malignant tumour cancer”
Fig 2: Novelty Checker – Top organizations working in “AI in malignant tumour cancer”
Market search also indicated that a South Korean AI company “LUNIT” started by university students uses high-level AI for cancer imaging. https://lunit.io/
In India, Oncostem has developed a novel product “Can Assist Breast” focusing specifically on breast cancer. https://oncostem.com/
Such ideas and concepts hint at a plethora of innovations which can be possible in the near future.
Tech and Health Care Collaboration
AI-empowered cancer research involves and demands convergence of efforts and ideas of medical science and technology.
Hence many big players of both technology and health industry are collaborating to devise innovations in the process of cancer research and treatment.
One pioneer example of such a collaboration is Envision.
Envision Better cancer care is a partnership between Varian Medical Systems and Siemens. This partnership brings together Varian treatment products with Siemens cutting edge imaging solutions.
GE Healthcare, Vanderbilt University Medical Center Partner for Safer, More Precise Immunotherapy Cancer Treatment – Artificial Intelligence (AI)-powered applications will help predict how individual patients will respond to immunotherapies in advance of treatment. Click Link
Collaborations and partnerships between health and tech companies bring the brightest minds and best of both areas together. This opens door to new avenues in the direction of AI-based cancer research.
Universities contribution to cancer research
Prestigious institutions such as the Zhejiang University, Fudan University, the University of California and Case Western Reserve University are few of the universities globally which have broadened the application scope of AI in Cancer Diagnosis and Treatment.
By patenting solutions which use concepts such as Neural Networks and Deep Learning, these Universities have developed remarkable solutions particularly in the field of Breast and Lung cancer.
A remarkable example and which has come forward in XLPAT Novelty Checker is that of the ZJU.
It has developed a respiration detecting device for diagnosis of early lung cancer by processing image data and extracting patterns from it by means of a neural network to diagnose eleven kinds of lung cancer. (CN200510049236A)
Another noteworthy innovation observed in the XLPAT database is that of the University of California.
It outlines the development of methods and kits which use machine learning algorithms for determining the prognosis and progression of breast cancer in a subject. (WO2018009703A1).
Zhejiang DE Image Solutions Company and ZJU worked with Intel to train DL algorithms to identify thyroid nodules and classify them as malignant or benign in ultrasound images. The AI-based medical image inferencing solution proved to be 10% more accurate than radiologists.
Many universities are investing in cancer research and it is a good sign for big companies to collaborate with universities who are aiming to develop technologies in this sector.
Some Indian Startups –
The Indian Med-tech Startup, AIstra has filed a patent for their AI-powered cervical detection system, CervAstra that detects the disease in its early stages, and measurably increases the odds of survival. CervAstra, therefore, makes Cervical cancer screening affordable, reliable and accessible.
The Bangalore-based 2016 startup, Niramai build a machine learning software that helps to detect breast cancer at a much early stage and can be the key to saving lives. In its patent, Niramai describes a Thermography-based breast cancer screening system which determines whether hot spots, as seen in a thermal image of both breasts, can be classified as being possibly malignant based on a determined measure of symmetry.
Challenges amid Opportunities
AI and ML technologies in cancer research and treatment have opened doors to many innovations and possible refinements, however, some vital questions need to be addressed and should be taken into consideration.
Some questions and facts which should be taken into account can be as follows:
• How to expedite AI work in Cancer research?
• How to consolidate innovations and come up with concrete integrated solutions?
• How to channelize R&D to take existing innovations forward?
• Ethical challenges: What if a smart algorithm misses a cancerous nodule on a lung X-ray or comes up with a false prediction?
• Governing bodies to regulate and standardize artificial intelligence implementation are practically non-existent.
• It is vulnerable to serious issues like data breach and identity theft.
• Data sets that are relevant & accurate for training deep learning algorithms and artificial machines are scarce.
Automation and machine learning technologies being applied to cancer research seem to be quite promising however it brings with itself own sets of challenges.
In the end, it is statistical computational operating data which is always prone to miscalculations and discrepancies.
One supporting example which outlines and reflects loopholes in this aspect is that of a report on IBM Watson.
IBM called Watson a revolution into the area of cancer research but however recent reports suggest otherwise. It was reported that the solutions given by Watson were not up to the mark and were sometimes completely opposite.
Here are few links for the same:
Hence, after thorough case studies and analysis, it can be inferred that the scope of Artificial Intelligence and Machine Learning technologies in cancer treatment is a behemoth, however, it is still a long way to go and a lot of possibilities to be explored.