|Year : 2019 | Volume
| Issue : 2 | Page : 182-189
Artificial intelligence in healthcare in developing nations: The beginning of a transformative journey
Abhishek Mahajan1, Tanvi Vaidya1, Anurag Gupta1, Swapnil Rane2, Sudeep Gupta3
1 Department of Radiodiagnosis and Imaging, Tata Memorial Centre, Tata Memorial Hospital; Homi Bhabha National Institute (HBNI), Mumbai, Maharashtra, India
2 Department of Pathology, Tata Memorial Centre, Tata Memorial Hospital; Homi Bhabha National Institute (HBNI), Mumbai, Maharashtra, India
3 Department of Medical Oncology, Tata Memorial Centre, Tata Memorial Hospital; Homi Bhabha National Institute (HBNI), Mumbai, Maharashtra, India
|Date of Web Publication||20-Dec-2019|
Department of Radiodiagnosis and Imaging, Tata Memorial Hospital, Dr E Borges Road, Parel, Mumbai - 400 012, Maharashtra; Homi Bhabha National Institute (HBNI), Mumbai, Maharashtra
Source of Support: None, Conflict of Interest: None
Introduction: Artificial intelligence (AI) is the fascinating result of the convergence of various technologies, algorithms and approaches. Its role in early detection and diagnosis will be a boon to society in the next few years and should be seen as 'Augmented Intelligence.' Issues like ethical considerations and collaborations need to be sorted out to ensure apt implementation in healthcare. This article reviews the current status of AI in developing nations like India and highlights some insights that could provide new directions and opportunities for AI in healthcare, especially in the field of radiology.
Implications for Patient Care: (1) AI will aid in nation-wide research and cooperation that will provide an impetus for the development of imaging science and decentralization of medical services. (2) AI may help to bridge the gap for need of specialized medical personnel in the peripheral areas in developing countries like India. (3) Government initiatives, ethical considerations and joint public private sector collaborations will ensure smooth transition and implementation of AI in healthcare especially in radiology.
Summary Statement: This manuscript highlights the existing gap between the developed and developing nations regarding implementation and acceptance of AI in healthcare and at the same time brings out insights in regard to the government initiatives to ease the progress of AI in healthcare in developing countries like India.
Keywords: Artificial intelligence, developing country, health system reform, healthcare, India, radiology
|How to cite this article:|
Mahajan A, Vaidya T, Gupta A, Rane S, Gupta S. Artificial intelligence in healthcare in developing nations: The beginning of a transformative journey. Cancer Res Stat Treat 2019;2:182-9
|How to cite this URL:|
Mahajan A, Vaidya T, Gupta A, Rane S, Gupta S. Artificial intelligence in healthcare in developing nations: The beginning of a transformative journey. Cancer Res Stat Treat [serial online] 2019 [cited 2021 Apr 23];2:182-9. Available from: https://www.crstonline.com/text.asp?2019/2/2/182/273674
| Introduction|| |
Artificial intelligence (AI) is the result of the convergence of various technologies, algorithms and approaches. AI refers to the ability of machines and systems to acquire and apply knowledge, so as to perform a variety of cognitive tasks (e.g., sensing, processing oral language, reasoning, learning, and making decisions), basically mimicking human behavior. This, if coupled with collection and data analytics, would enhance our understanding of the way things work. In the United States (US) and other developed countries, the use of AI in the healthcare sector is already revolutionizing the industry with AI being used in various spheres of medicine, such as diagnostics, personalized medicine, and the pharmaceutical industry to mention a few for example, in a recent study, the AI system was shown to have comparable efficacy as compared to radiologists in screening of breast cancers (area under the curve was 61.4% higher than that of the radiologists). The adoption of AI in the healthcare sector can have far-reaching implications in terms of augmenting accessibility to healthcare services through early detection, diagnosis, decision-making, and treatment and is expected to see an exponential increase in the next few years. Multinational companies are also collaborating to draw upon clinical insights and expertise in the field, to synchronize AI applications with imaging informatics solutions.,, This article reviews the current status of AI in developing nations like India and highlights some insights that could provide new directions and opportunities for AI in healthcare, especially in the field of radiology.
| Artificial Intelligence in Radiology|| |
Radiology has been one of the most rapidly evolving fields of medicine. AI is not new to radiology, and AI-based radiology consult systems such as the Missouri Automated Radiology System (MARS), ICON (a computer-based expert system being developed to help radiologists), MARSII, and Phoenix are testaments to the existence of AI in radiology since the early 1970s., Since then, AI has made its presence felt in various radiology applications. Machine learning algorithms such as Naive Bayes, Genetic Algorithms, Neural Networks, and Support Vector Machines (SVM) have been used for decades for detection, diagnosis, classification, and risk assessment of breast cancer. In the test data of the artificial neural network (ANN) model that was implemented as a result of these analyses, disease prediction rate was 90.5%. It is seen from these high predictive values that the ANN model is fast, reliable and without risks and therefore can be of great help to physicians. In the recent years, development of accurate deep learning algorithms for breast cancer diagnosis has emerged as a key tool that can be applied to reduce the interpretation time required by radiologists, thus improving clinical efficiency. Some notable radiology projects involving AI technology include the 'Knee OA staging' project, which is currently underway at Stanford University, and aims to automatically quantify the severity of knee osteoarthritis from X-ray images. Various machine learning techniques like SVM and region convolutional neural networks are being explored to achieve the same. Recently, several researchers have tried to apply machine learning methods to neuroimaging data to assist with timely stroke diagnosis. Rehme et al. used SVM in resting-state functional magnetic resonance imaging (fMRI) data, by which endophenotypes of motor disability after stroke were identified and classified (endophenotype refers to an internal characteristic of a person which renders the individual more susceptible to a disorder). Griffis et al. tried naïve Bayes classification to identify the stroke lesion in T1-weighted MRI and found that the results were comparable with human expert manual lesion delineation. AI technologies incorporating multifeature analysis using diffusion-weighted imaging, magnetic field correlation, fMRI, and volumetrics have also shown promise in the accurate detection and classification of traumatic brain injury in emergency settings. Given the growing volumes of clinical imaging data, developing a data-driven segmentation model has assumed great clinical importance to avoid tedious manual processing and reduce inter-observer variability in the analysis and characterization of lung nodules. AI-based healthcare solutions such as these could enhance productivity and improve healthcare outcomes; and the time is ripe for healthcare providers to test and embrace this ubiquitous AI technology and help the country to overcome present and future challenges.
From the time of patient entry into hospital until reporting the result to the clinician, AI has the potential to do wonders. It can automatically link prior patient records to the current imaging and schedule the patient accordingly, based on the need of the patient as well as automatically detect emergency findings, which would only need to be confirmed by radiologists once, thus leading to effective treatment [Figure 1].
| Ethical Framework for Artificial Intelligence in Radiology|| |
The ethical framework for AI applications in radiology should consist of biomedical ethics – autonomy, beneficence, justice, nonmaleficence, and explicability [Figure 2].
Autonomy: Patients have the right to make their own decisions as images contain not just pixel information, but also threaten privacy and data ownership is questionable.
The Beneficience , 'do good' and Nonmaleficence, 'do no harm,' lay emphasis on avoiding loss to the patient's well-being medically or commercially (such as targeting a subset of patients for a product marketing).
Justice implies the fair distribution of medical goods and services without any discrimination, ensuring shareable benefits, and preventing any new harm that can arise from implicit bias.
Explicability consists of transparency and accountability.
Transparency: There should be satisfactory logical explanations for the decision-making process of AI which is accessible to the user/patient in case of any discrepancy.
Accountability: There could be medicolegal issues regarding responsibility for these decisions; if an autonomous system makes a mistake, who is responsible – its developer or the user?
| What Opportunities Can Artificial Intelligence Bring to Developing Countries?|| |
Strengthening national competitiveness
“Everything that civilization has to offer is a product of human intelligence; we cannot predict what we might achieve when this intelligence is magnified by the tools that AI may provide, but the eradication of war, disease, and poverty would be high on anyone's list. Success in creating AI would be the biggest event in human history. Unfortunately, it might also be the last.” These words of Stephen Hawking, one of the greatest minds of the century emphasize that AI, as the foremost cutting-edge technology of this generation, can potentially accelerate the pace of innovation, enhance the productivity of a country and strengthen national competitiveness. AI promotes growth in three ways: The intelligent automation of the workforce, augmentation of labor and physical capital and creation of opportunities for new skills, business ideas and services. The widespread use of AI in education, medical care, environmental protection, urban operations, judicial services, and other fields will greatly improve the level of precision in public services. It is likely that AI technologies will be able to accurately sense, forecast, and provide early warning of major situations for infrastructure facilities and social security operations; and take desired actions, which will significantly elevate the capability and level of social governance, playing an important role in effectively maintaining social stability.
In recent times, another application of AI starting to make its presence felt in cancer care in some low and middle-income countries is 'Watson for Oncology (WFO),' developed by IBM in collaboration with the Memorial Sloan Kettering Cancer Center (MSKCC). WFO is a cognitive computing system designed to provide treatment recommendations after analyzing data obtained from published medical literature, treatment protocols, patient charts, test cases, and guidelines that have been selected by the experts from MSKCC. The WFO/Cota 'recommended' option was similar to the selection by breast cancer experts in 175 of 223 cases (78.5%) and the 'for consideration' option was selected in 21 cases (9.4%);
Supplementing areas of need
Healthcare systems in developing countries are endemically short of medical workers. AI applications have the potential to fill this gap. For instance, an important application involves the decentralization of diagnostic testing through AI-based diagnostic technology. AI applications have been developed that substitute and complement highly trained and expensive expertise by analyzing medical images. For example, AI has shown its capability in classifying skin cancer with a level of accuracy comparable to dermatologists. Thus, mobile devices outfitted with deep neural networks can potentially extend the reach of dermatologists outside of the clinic. Another example is the use of introducing natural language processing for extracting pneumonia-related concepts from chest X-ray reports which would then assist the antibiotic assistant system to alert physicians to the possible need for anti-infective therapy., Clinical medicine requires doctors to handle enormous quantities of data, pertaining to human physiology, biochemistry, imaging and increasingly genetic profiling. The ability to assimilate and analyze the data in a holistic manner is crucial for decision-making. Machine learning may become an invaluable complementary tool for clinicians in the practice of personalized medicine.
Education and employment
AI can potentially provide customized teaching and create job opportunities for AI-related work in the IT sector.
Untapped potential and challenges: A SWOT (Strengths/Weaknesses/Opportunities/Threats) analysis of artificial intelligence for developing nations
We present a critical review of the current status of AI in developing nations through a SWOT analysis [Figure 3]. This key management tool is used by organizations to focus on their strengths, minimize threats, and take the maximum advantage of opportunities available.
The main strengths of India to become a leader in AI adoption are:
- Its vast and emerging talent pool: With universities offering ample courses on data sciences and machine learning, the nation's growing engineering talent is being directed toward the development of AI-based solutions
- Freedom from legacy assets: The exponential growth in computing power in recent years, coupled with significant progress in software, provides developing nations with a superior ecosystem in their pursuit of AI-based AI-based solutions, bypassing some of the older inefficient systems and processes
- Availability of big data: The most dependent factor of AI is the availability of data. AI functions are directly proportional to the amount of data and the kind of data fed into the system and it is said, “If AI is the new electricity, data is the new charcoal.” The Picture Archiving and Communication System (PACS) revolutionized radiology and 'big data' are readily available for developing AI algorithms. Considering that 90% of the world's population is in the developing nations and 95% of the patients require some form of imaging during the process of their treatment, there are huge amounts of data housed in developing nations and there is disparity in regard to development in the field of AI in these nations. These data serve as a gold mine of valuable information for the development of AI-based healthcare applications. However, India still lags behind developed markets like the US and the United Kingdom in terms of innovation and has not yet developed a robust AI-based ecosystem.
Before deploying AI in a worldwide manner, there are several barriers that face developing nations that require to be addressed such as
- Lack of trained personnel and expertise for AI development
- Nonstandardized data curation; the need for 'intelligent data mining and management'
- High resource cost and nonavailability of tools in developing nations
- Lack of awareness for adoption and collaborative approaches for implementation
- Quality control, data privacy, and data security
- The 'digital divide': Although billions of people around the world have reaped the benefits of the internet, many billions have been left behind. For example, internet access per 100 inhabitants in the Asia-Pacific region is still far lower than Europe and North America; the widening gap among sub-regions in Asia and the Pacific is an alarming trend, considering that the introduction of AI on a global scale is only possible with broadband infrastructure. While developed countries, with the most expansive and high-speed broadband networks, are embracing and investing in AI at astonishing rates, developing countries have bigger mountains to climb before they can hope to achieve this goal and are still getting left behind.
There is growing interest, however, as India starts to invest additional resources and deploy new AI applications in various sectors. Amidst the 'Group of Twenty International Forum' (G20) countries, India ranked third in 2016 largely due to an increase in the number of AI start-ups since 2011 at a Compound Annual Growth Rate of 86% – higher than the global average. However, it is dominated by American firms like Accenture, Microsoft and Adobe, which have their innovation centers here in India. Government organizations such as National Institution for Transforming India (NITI) Aayog and Department of Biotechnology are playing key roles in understanding the current situation of AI in India and are implementing solutions for rapid deployment of AI in healthcare. NITI Aayog is a policy think tank of the Government of India. It aims to achieve Sustainable Development Goals and enhance cooperative federalism by involvement of State Governments of India in the economic policy-making process via a bottom-up approach. Apart from this, startups and public private collaborations are also taking the lead in developing novel AI algorithms customized to healthcare needs, especially radiology.
- Lack of acceptance; a potential concern of technological unemployment: It is believed that widespread adoption of AI may make some jobs redundant. However, even in AI-advanced countries, an international survey of 3000 AI-aware executives across ten countries concluded that most firms did not expect AI to significantly reduce the size of their workforce. Thus, contrary to the general perception, the role of AI would basically be to complement rather than substitute the human taskforce
- Ethical considerations: Ethical considerations with respect to privacy and security, including lack of formal regulations with regards to anonymization of data, balancing privacy and accessibility of data present important ethical dilemmas. How governments acquire and manage data, now and in the future, will be crucial. Striking the right balance between privacy, ownership and transparency is an uphill task.,
| India Leading the Way to Development Via Artificial Intelligence|| |
Advances in AI and data analytics are boosting innovation in many parts of the world. While the US government is investing only $1.1 billion in nonclassified AI research, its private sector is spending billions in fields such as finance, healthcare and defense, which is transforming their economy. India, being the fastest growing economy with the second largest population in the world, is likely to have a significant stake in the AI revolution in the near future. As of now, India devotes only 0.6% of the Gross Domestic Product (GDP) to research and development (R&D), well below the 2.74% in the US. Its limited investment has certainly slackened innovation and put the country at an economic disadvantage. A PricewaterhouseCoopers report estimates that AI will contribute $15.7 trillion to the world economy by 2030 – more than the combined current output of China and India. An Accenture report, ReWire for Growth, forecasts that AI will boost India's annual growth rate by 1.3 percentage points of gross value added (GVA) by 2035. This will amount to an addition of $957 billion, or 15% of current GVA (a close approximation of GDP), to India's economy in 2035, compared with a scenario without AI., In India, as healthcare solution providers are looking for ways to achieve the maximum productivity and efficiency, AI can have a far-reaching impact on the healthcare situation in a more constructive manner. India has been at the forefront of bringing recent advances in the field of medicine to clinical practice. The concept of bench to bedside molecular functional imaging and research work originating from collaboration between medical and engineering fraternity in the field of AI in radiology are testimony to this.,, Thus, it is becoming increasingly important for Indian healthcare providers to leverage this technology as the country is facing a serious health crisis due to growing burden of diseases and poor doctor–patient ratio. With a combination of new technologies such as AI, Internet, and Big Data, India can bring a new generation of healthcare solutions which will be a game changer for the sector.
| Government Initiatives to Strengthen Artificial Intelligence in Healthcare in India|| |
Recognizing AI's potential to transform India's economy, the Government of India has authorized an organization named 'NITI Aayog' to address the national strategy on AI and other emerging technologies. In pursuit of the above, NITI Aayog has collaborated with several leading AI technology companies to implement AI projects in critical areas such as agriculture and health. NITI Aayog CEO, Amitabh Kant stated that “AI will be bigger than the advent of the Internet or the harnessing of electricity.” and “India must embrace it (AI) with all its might.” As India embarks on a journey to lead the developing world in AI research and application; with the motto of 'AI for all' (#aiforall); there is a need to address the challenges of access, affordability, shortage, inconsistency of skilled expertise as well as to capitalize on the opportunities that manifest during this journey. One such initiative by NITI Aayog is the Tata Memorial Center Imaging BioBank.
| Tata Memorial Center Imaging Biobank|| |
AI-based Radiomics project by NITI Aayog in collaboration with Tata Memorial Centre Imaging Biobank: (Machine learning and Artificial Intelligence Database and Tumor Radiomics Atlas Project for Cancer unit) is currently underway, which will allow the generation of imaging biomarkers for use in research studies, support biological validation of existing and novel imaging biomarkers and in the long run, provide an unprecedented opportunity to improve decision support in cancer treatment at low cost, [Figure 4].
| Recommendations by Niti Aayog for Promoting Artificial Intelligence Adoption in India|| |
For the rapid development and progress of AI in India, NITI Aayog has proposed a two-tiered structure to address India's AI research aspirations:,,,
- Centre of Research Excellence focused on developing a better understanding of existing core research and pushing technology frontiers through creation of new knowledge
- International Centers of Transformational AI (ICTAI) with a mandate of developing and deploying application-based research. Private sector collaboration is envisioned to be a key aspect of ICTAIs. The Tata Memorial Center Imaging Biobank will be one of the first initiatives of the ICTAI.
Certain recommendations have been laid down by NITI Aayog to promote AI adoption in the country. These recommendations could go a long way in the attainment of 'AI for all', an ambitious enterprise that endeavors not only to promote development of AI in various sectors but also tackle the responsibilities that accompany it, such as ensuring adequate data privacy, security and balancing ethical considerations with the need for innovation. The five recommendations include:
| Creating a Multistakeholder Marketplace– the National Artificial Intelligence Marketplace|| |
The development of any AI-based product is a tedious process requiring a multitude of specialized processes that are necessary for final delivery. Thus, it is extremely difficult for start-ups to venture into the AI industry. In order to encourage the development of sustainable AI solutions for sectors such as health and education, it is necessary to create a level playing field to facilitate the participation of all players in the value chain. The acquisition of raw material or data is the starting point for entering the realm of AI-based solutions. This is a major barrier to enter the industry as very few players have access to usable data. Thus, it is crucial to address information asymmetry and promote effective collaboration between the various stakeholders in the AI ecosystem in order to ensure a level playing field. This may be possible by creating a 'marketplace' which could:
- Enable access to the required AI component, be it data or business models, and services, such as data annotation, and enable rating of these assets
- Serve as a platform for execution and verification of transactions.
Such a marketplace model would have a following advantage:
- Reducing asymmetry of information: Transactions on a large scale would provide incentives to both data owners and AI model creators and encourage participation of new players
- Facilitate creation of novel business models in a targeted manner: Efficient availability of raw materials required at different time points of the AI solution development cycle would allow effective segregation of tasks which is essential to tackle specific problems and find appropriate solutions
- Unraveling new sources of data and facilitating more efficient use of computational and human resources: It is estimated that only 1% of data today is analyzed due to lack of awareness and availability of AI experts. For instance, several medical imaging centers are collecting valuable data; however, these databases are not analyzed since AI models cannot be created without computational infrastructure and trained personnel. In the presence of a formal marketplace, diagnostic centers would have an incentive to collect these data and provide access to the data in the market with requisite security measures in place
- Provide an opportunity to address ethical concerns regarding data sharing: Currently data sharing occurs in an informal manner with no safeguards in place. Creation of a formal marketplace for data transactions would ensure the development of data security measures to prevent misuse of valuable information.
| Facilitating Creation of Assorted Big Data Sets|| |
The lack of annotated data in the domestic context has emerged as a major hurdle in development of AI solutions for both startups and research projects. NITI proposes the creation of 'Big Data' sets which can be readily accessible to them, akin to a 'plugin' mode which would enable easy access to data customized to the needs of Indian AI developers. Availability of such a large corpus of annotated data will certainly spur research and innovation in the field of AI and machine learning. Given the magnitude of this task, creation of these data sets will require the availability of financial resources as well as international expertise to focus on problems in the Indian context.
| Partnerships and Collaboration|| |
AI being a highly collaborative domain requires a stratified approach, for promoting development of AI tools as well as adoption of AI in different spheres of life. A stratified approach to partnerships is proposed as follows:
- Low transfer: informal interactions such as conferences and social networking
- Medium mobility: training of industrial employees, internship programs, and academic entrepreneurship
- High relationships: sharing infrastructure and interorganizational arrangements for pursuing R&D.
Such a multipronged approach would promote the implementation of the “AI+X” paradigm, where AI researcher works in close collaboration with the researchers in other sectors (termed as X). For instance, a radiologist and computer scientist could work together to develop solutions and enhance knowledge in real time. Industry–Research collaboration is also essential to improve the final output based on feedback from the end user. Finally, collaboration with trade bodies and venture capitalists could be beneficial by sharing information about the common problems, arriving at possible solutions, identifying new international markets for AI-based products/technologies and negotiating transaction between national and international bodies.,
| Spreading Awareness of the Advantages of Artificial Intelligence|| |
A major hurdle to the adoption and promotion of AI is the lack of awareness with regards to the work being done across the country. This can be overcome by creating an online portal such as an AI Database for registered people to access and obtain information. This database could serve as a reliable source of information for experts and projects being implemented. Such a portal could also become a forum for sharing discussion related to research collaboration and finding trained personnel to deliver on AI projects. More importantly, there is a need to create awareness among officials in government agencies and public sector undertakings, regarding the various benefits that AI offers, by organizing workshops and live demonstrations of AI applications and how their implementation can help augment human labor, rather than displace it.
| Promoting Startups|| |
It is a must to provide an impetus to the economy of the nation. Availability of financial support and adequate infrastructural facilities is important to ensure their participation in AI projects.
| Conclusion: Roadmap to Bridge the Gap for Artificial Intelligence and Healthcare in Developing Nations|| |
The future for AI in Radiology and in healthcare is bright. It has the potential to augment the care we provide and should be seen as 'Augmented Intelligence.' As the available database grows exponentially, human beings will find it increasingly difficult to process this information to make meaningful medical decisions. AI may play a key role in utilizing individual data to diagnose disease, thereby enabling the creation of personalized treatment plans. The time is ripe for developing countries like India to join the race to lead the AI revolution, which is still in infancy. The world's developed nations have long been the front-runners in this competition; which truly cuts across all spheres of national power, considering that leadership in AI will enable global dominance. Though developing nations like India lag the superpowers in fundamental research and resources, they enjoy advantages in the form of a vast engineering workforce, a burgeoning startup scene, and a large pool of data waiting to be tapped. More importantly, these nations have the entrepreneurial spirit to help businesses derive value from real-time data and the ambition to carve a niche for themselves in an increasingly AI-driven world. The onus is on the national and international radiology associations, premier teaching institutes and government organizations to bridge the gap of AI development. Strategic positioning, ethical considerations, and joint public–private sector collaborations will ensure smooth transition and implementation of AI in healthcare, especially in radiology.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Russell SJ, Norvig P. Artificial Intelligence: A Modern Approach. Malaysia: Pearson Education Limited; 2016.
Sharma A, Harrington RA, McClellan MB, Turakhia MP, Eapen ZJ, Steinhubl S, et al.
Using digital health technology to Better Generate evidence and deliver Evidence-based Care. J Am Coll Cardiol 2018;71:2680-90.
Saritas I. Prediction of breast cancer using artificial neural networks. J Med Syst 2012;36:2901-7.
Obermeyer Z, Emanuel EJ. Predicting the future – Big data, machine learning, and clinical medicine. N
Engl J Med 2016;375:1216-9.
Verghese A, Shah NH, Harrington RA. What this computer needs is a physician: Humanism and artificial intelligence. JAMA 2018;319:19-20.
Saracci R. Epidemiology in Wonderland: Big data and precision medicine. Eur J Epidemiol 2018;33:245-57.
Lehr JL, Lodwick GS, Nicholson BF, Birznieks FB. Experience with MARS (Missouri automated radiology system). Radiology 1973;106:289-94.
Swett HA, Miller PL. ICON: A computer-based approach to differential diagnosis in radiology. Radiology 1987;163:555-8.
Rodriguez-Ruiz A, Lång K, Gubern-Merida A, Broeders M, Gennaro G, Clauser P, et al.
Stand-alone artificial intelligence for breast cancer detection in mammography: Comparison with 101 radiologists. J Natl Cancer Inst 2019;111:916-22.
Burt JR, Torosdagli N, Khosravan N, RaviPrakash H, Mortazi A, Tissavirasingham F, et al.
Deep learning beyond cats and dogs: Recent advances in diagnosing breast cancer with deep neural networks. Br J Radiol 2018;91:20170545.
Rehme AK, Volz LJ, Feis DL, Bomilcar-Focke I, Liebig T, Eickhoff SB, et al.
Identifying neuroimaging markers of motor disability in acute stroke by machine learning techniques. Cereb Cortex 2015;25:3046-56.
Griffis JC, Allendorfer JB, Szaflarski JP. Voxel-based Gaussian naïve Bayes classification of ischemic stroke lesions in individual T1-weighted MRI scans. J Neurosci Methods 2016;257:97-108.
Lui YW, Xue Y, Kenul D, Ge Y, Grossman RI, Wang Y. Classification algorithms using multiple MRI features in mild traumatic brain injury. Neurology 2014;83:1235-40.
Wang S, Zhou M, Liu Z, Liu Z, Gu D, Zang Y, et al.
Central focused convolutional neural networks: Developing a data-driven model for lung nodule segmentation. Med Image Anal 2017;40:172-83.
Kahn CE Jr. From images to actions: Opportunities for artificial intelligence in radiology. Radiology 2017;285:719-20.
Berlyand Y, Raja AS, Dorner SC, Prabhakar AM, Sonis JD, Gottumukkala RV, et al.
How artificial intelligence could transform emergency department operations. Am J Emerg Med 2018;36:1515-7.
D'Antonoli TA, Weikert TJ, Sauter AW, Sommer G. Ethical considerations for artificial intelligence implementation in radiology. p. 4.
Atieh Graham DM, McNamara DM, Waintraub SE, Goldberg SL, Norden AD, Hervey J, et al
. 1589P Are treatment recommendations provided by cognitive computing supported by real world data (Watson for Oncology with Cota RWE) concordant with expert opinions?. Ann Oncol 2018;29(suppl_8):mdy297-031.
Philip CC, Mathew A, John M J. Cancer care: Challenges in the developing world. Cancer Res Stat Treat 2018;1:58-62. [Full text]
Hamet P, Tremblay J. Artificial intelligence in medicine. Metabolism 2017;69:S36-40.
Massat MB. Artificial intelligence in radiology: Hype or hope? Appl Radiol 2018;47:22-6.
Miller DD, Brown EW. Artificial intelligence in medical practice: The question to the answer? Am J Med 2018;131:129-33.
Esteva A, Kuprel B, Novoa RA, Ko J, Swetter SM, Blau HM, et al.
Dermatologist-level classification of skin cancer with deep neural networks. Nature 2017;542:115-8.
Fiszman M, Chapman WW, Aronsky D, Evans RS, Haug PJ. Automatic detection of acute bacterial pneumonia from chest X-ray reports. J Am Med Inform Assoc 2000;7:593-604.
Mohan CS. Artificial intelligence in radiology – Are we treating the image or the patient? Indian J Radiol Imaging 2018;28:137-9.
] [Full text]
Chen JH, Asch SM. Machine learning and prediction in medicine - beyond the peak of inflated expectations. N
Engl J Med 2017;376:2507-9.
Paiva OA, Prevedello LM. The potential impact of artificial intelligence in radiology. Radiol Bras 2017;50:V-VI.
Liew C. The future of radiology augmented with artificial intelligence: A strategy for success. Eur J Radiol 2018;102:152-6.
Char DS, Shah NH, Magnus D. Implementing machine learning in health care – Addressing ethical challenges. N
Engl J Med 2018;378:981-3.
Yeung S, Downing NL, Fei-Fei L, Milstein A. Bedside computer vision – Moving artificial intelligence from driver assistance to patient safety. N
Engl J Med 2018;378:1271-3.
Mahajan A, Goh V, Basu S, Vaish R, Weeks AJ, Thakur MH, et al.
Bench to bedside molecular functional imaging in translational cancer medicine: To image or to imagine? Clin Radiol 2015;70:1060-82.
Sapate SG, Mahajan A, Talbar SN, Sable N, Desai S, Thakur M. Radiomics based detection and characterization of suspicious lesions on full field digital mammograms. Comput Methods Programs Biomed 2018;163:1-20.
Singadkar G, Mahajan A, Thakur M, Talbar S. Automatic lung segmentation for the inclusion of juxtapleural nodules and pulmonary vessels using curvature based border correction. J King Saud Univ Comput Inf Sci 2018. Doi: https://doi.org/10.1016/j.jksuci.2018.07.005
Chavan SS, Mahajan A, Talbar SN, Desai S, Thakur M, D'cruz A. Nonsubsampled rotated complex wavelet transform (NSRCxWT) for medical image fusion related to clinical aspects in neurocysticercosis. Comput Biol Med 2017;81:64-78.
Noronha V. Making a case for cancer research in India. Cancer Res Stat Treat 2018;1:71-4. [Full text]
[Figure 1], [Figure 2], [Figure 3], [Figure 4]