• Users Online: 289
  • Print this page
  • Email this page


 
 
Table of Contents
LETTER TO EDITOR
Year : 2020  |  Volume : 3  |  Issue : 3  |  Page : 622-624

Mining artificial intelligence in oncology: Tata Memorial Hospital journey


1 Department of Radiodiagnosis and Imaging, Tata Memorial Hospital, Homi Bhabha National Institute, Mumbai, Maharashtra, India
2 Department of Radiodiagnosis and Imaging, Tata Memorial Hospital, Mumbai, Maharashtra, India

Date of Submission04-Mar-2020
Date of Decision20-Apr-2020
Date of Acceptance20-Apr-2020
Date of Web Publication19-Sep-2020

Correspondence Address:
Abhishek Mahajan
Department of Radiodiagnosis and Imaging, Tata Memorial Hospital, Dr E Borges Road, Parel, Mumbai - 400 012, Maharashtra
India
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/CRST.CRST_59_20

Get Permissions


How to cite this article:
Bothra M, Mahajan A. Mining artificial intelligence in oncology: Tata Memorial Hospital journey. Cancer Res Stat Treat 2020;3:622-4

How to cite this URL:
Bothra M, Mahajan A. Mining artificial intelligence in oncology: Tata Memorial Hospital journey. Cancer Res Stat Treat [serial online] 2020 [cited 2020 Oct 21];3:622-4. Available from: https://www.crstonline.com/text.asp?2020/3/3/622/295524



The last decade has witnessed the upsurge of artificial intelligence (AI) in the field of medicine, specifically in medical imaging. The enormous research work has led medical imaging to reach the vanguard of technological transformation. Lately, the advances in hardware (graphics processing unit) and mathematical algorithms (neural networks) have made AI the most sought after topic for research, and inputs from the experts in various fields of medicine and information technology have led to a boom in its development globally. Various large vendors such as IBM Watson, Butterfly Network Inc., Arterys, Google, and Zebra Medical Vision have been applying AI in medical imaging and propelling it for innovative solutions.[1] To this date, physicians know little about AI and machine learning, except for those who actively work with it. For a majority of physicians, it is a future prospect that will take ample time and investments to reach its full potential, whereas for some others, it is a vision that now seems attainable, is welcomed, and has been incorporated for the betterment of the current health-care facilities in the developed world.[1]

Till date, most of the AI applications could perform, at their best, at subhuman levels, but with newer advances and deeper learning, AI has shown an efficacy equal to that of humans and even surpassed humans in certain purpose-specific applications.[1],[2],[3] However, most of the published literature has emphasized the use of AI for enhancing the working capacity and potential of the health-care professionals, for example, for segregating normal clinical reports from those with positive findings, such as radiographs with fractures, or calling attention to clinical findings indicating life-threatening conditions, such as pneumothorax and pulmonary embolism, which need urgent attention. Therefore, it is more augmenting in nature rather than being purely artificial. Currently, AI is in the second phase, “ peak of inflated expectations,” of the Gartner hype cycle, and Gartner has singled out AI as one of the top three megatrends in technological development, owing to its excellent computational command, infinite data, and newer and novel advances in machine and deep learning networks.[1],[2],[3]

Radiomics, a niche area in machine learning, is a method enabling the extraction of a large number of features from medical imaging using data characterization algorithms. It has witnessed the description of shape, texture, and intensity pertaining to medical imaging and is currently implicated in the characterization of various features with clinical relevance to reach a diagnosis. Oncology, in particular, utilizes radiomics for the identification and stratification of risk in various malignancies;[4],[5] the spectrum includes prediction of distant metastases, histological tumor profile, gene expression, recurrence, as well as survival.[6],[7],[8] AI can be used for making diagnostic predictions in brain imaging for the detection of cancerous lesions and differentiating between benign and malignant or primary and metastatic lesions.[9]

One such example is the indigenous AI algorithm for brain tumor segmentation (BraTS) that uses a three-dimensional, deep, convolutional neural network. The algorithm was developed by the departments of radiodiagnosis and imaging, pathology, medical oncology, and neurosurgery at the Tata Memorial Center, Mumbai, India, in collaboration with the Department of Electronics and Telecommunication at the Shri Guru Gobind Singhji Institute of Engineering and Technology, Nanded, India. It can improvise the segmentation in medical imaging and predict the overall survival in the patients with radiomic features ciphered for the given tumor morphology; it was awarded the 3rd best prize in the International BraTS 2018 Challenge.[10] For its contribution to the development and upgradation of AI pertaining to cancer imaging, the cancer imaging archive is worth mentioning. This platform hosts a myriad of collection of anonymized data, specifically related to cancer imaging, which is readily available in the Digital Imaging and Communications in Medicine format after processing, and is categorized according to the type of disease and the modality on which it was studied.[11]

Other recent advances in the field of AI from India include research on the open-source BI-RADS data sets conducted at the All India Institute of Medical Sciences, New Delhi, to assess the accuracy of prediction of histology with and without a radiologist's opinion for breast cancer imaging.[12] A content-based image retrieval system for pulmonary nodules using integrated AI and computer-aided detection was developed by Dhara et al. at the Post Graduate Institute of Medical Education and Research, Chandigarh.[13] Furthermore, the incorporation of AI in the imaging of cancers of the prostate gland, breast, lung, and brain constitutes other such examples of recent advances.[14],[15],[16],[17],[18],[19],[20],[21],[22],[23]

The ongoing research and developments in biobanks have been enriched by the acceptance from the Department of Biotechnology (Government of India) under the guidance of the National Institution for Transforming India Aayog. In their article, “AI for all,” they address the strategy on AI and other emerging technologies and have allocated a tidy sum of research grant for its upshot.[1],[24] The project is in collaboration with the Tata Memorial Center Imaging Biobank – the Machine Learning and AI Database and Tumor Radiomics Atlas Project for carving data on oncopathology and oncoimaging under the supervision of Dr. Sudeep Gupta, Dr. Abhishek Mahajan, and Dr. Swapnil Rane.[1],[25]

For unknown reasons, in developing countries such as India, there is still a lack of the incorporation of AI in health-care facilities. In such nations, ethical issues need to be addressed with appropriate rules and regulations to enable its smooth functioning in the near future. Lately, concerns regarding the privacy and security of subjects have been raised, as the digital data tapped into the system could be used to reconstruct the face of the subject, which can certainly be misused.[1] To rectify this issue, various concerned regulatory bodies have laid down guidelines regarding feeding the data into the system. The guidelines recommend that the data should be fed clearly without the superficial details, such as facial features (anonymization), from the computed tomography scans and magnetic resonance imaging of the head-and-neck region to prevent image reconstruction. They also recommend identifying localizations using the volumetric surface rendering software and replacing the name and surname with a pseudonym (pseudonymization) so as to enhance the security and privacy. However, in the developing world, this anonymization of data is lacking, and hence, there is a need for strict regulations pertaining to privacy and transparency.[1],[23] With the growing concern and the impeccable research contributions and monetary benefaction by the government of India, the day is not far when the field of medicine will witness the much-awaited improvement in health-care facilities with men not just behind the machines but an improvised version of a man synced with the machines.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.



 
  References Top

1.
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.  Back to cited text no. 1
  [Full text]  
2.
Moravčík M, Schmid M, Burch N, Lisý V, Morrill D, Bard N, et al. DeepStack: Expert-level artificial intelligence in heads-up no-limit poker. Science 2017;356:508-13.  Back to cited text no. 2
    
3.
Pendleton SD, Andersen H, Du X, Shen X, Meghjani M, Eng YH, et al. Perception, planning, control, and coordination for autonomous vehicles. Machines 2017;5:6.  Back to cited text no. 3
    
4.
Kolossváry M, Kellermayer M, Merkely B, Maurovich-Horvat P. Cardiac computed tomography radiomics: A comprehensive review on radiomic techniques. J Thorac Imaging 2018;33:26-34.  Back to cited text no. 4
    
5.
Aerts HJ, Velazquez ER, Leijenaar RT, Parmar C, Grossmann P, Carvalho S, et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat Commun 2014;5:4006.  Back to cited text no. 5
    
6.
Wu W, Parmar C, Grossmann P, Quackenbush J, Lambin P, Bussink J, et al. Exploratory study to identify radiomics classifiers for lung cancer histology. Front Oncol 2016;6:71.  Back to cited text no. 6
    
7.
Huynh E, Coroller TP, Narayan V, Agrawal V, Romano J, Franco I, et al. Associations of radiomic data extracted from static and respiratory-gated CT Scans with disease recurrence in lung cancer patients treated with SBRT. PLoS One 2017;12:e0169172.  Back to cited text no. 7
    
8.
Parmar C, Grossmann P, Bussink J, Lambin P, Aerts HJ. Machine learning methods for quantitative radiomic biomarkers. Sci Rep 2015;5:13087.  Back to cited text no. 8
    
9.
Orringer DA, Pandian B, Niknafs YS, Hollon TC, Boyle J, Lewis S, et al. Rapid intraoperative histology of unprocessed surgical specimens via fibre-laser-based stimulated Raman scattering microscopy. Nat Biomed Eng 2017;1. pii: 0027.  Back to cited text no. 9
    
10.
Baid U, Talbar S, Rane S, Gupta S, Thakur MH, Moiyadi A, et al. Deep learning radiomics algorithm for gliomas (drag) model: A novel approach using 3d unet based deep convolutional neural network for predicting survival in gliomas. In: International MICCAI Brainlesion Workshop. Cham: Springer; 2018. p. 369-79.  Back to cited text no. 10
    
11.
Clark K, Vendt B, Smith K, Freymann J, Kirby J, Koppel P, et al. The cancer imaging archive (TCIA): Maintaining and operating a public information repository. J Digit Imaging 2013;26:1045-57.  Back to cited text no. 11
    
12.
Ghosh A. Artificial intelligence using open source BI-RADS data exemplifying potential future use. J Am Coll Radiol 2019;16:64-72.  Back to cited text no. 12
    
13.
Dhara AK, Mukhopadhyay S, Dutta A, Garg M, Khandelwal N. Content-based image retrieval system for pulmonary nodules: Assisting radiologists in self-learning and diagnosis of lung cancer. J Digit Imaging 2017;30:63-77.  Back to cited text no. 13
    
14.
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.  Back to cited text no. 14
    
15.
Davatzikos C, Barnholtz-Sloan JS, Bakas S, Colen R, Mahajan A, Quintero CB, et al. AI-based prognostic imaging biomarkers for precision neurooncology: The ReSPOND consortium. Neuro Oncol 2020. pii: noaa045.  Back to cited text no. 15
    
16.
Singadkar G, Mahajan A, Thakur M, Talbar S. Deep deconvolutional residual network based automatic lung nodule segmentation. J Digit Imaging 2020.  Back to cited text no. 16
    
17.
Sapate S, Talbar S, Mahajan A, Sable N, Desai S, Thakur M. Breast cancer diagnosis using abnormalities on ipsilateral views of digital mammograms. Biocybernet Biomed Eng 2020;40:290-305.  Back to cited text no. 17
    
18.
Baid U, Talbar S, Rane S, Gupta S, Thakur MH, Moiyadi A, et al. A novel approach for fully automatic intra-tumor segmentation with 3D U-net architecture for gliomas. Front Comput Neurosci 2020;14:10.  Back to cited text no. 18
    
19.
Hambarde P, Talbar SN, Sable N, Mahajan A, Chavan SS, Thakur M. Radiomics for peripheral zone and intra-prostatic urethra segmentation in MR imaging. Biomed Signal Proce Control 2019;51:19-29.  Back to cited text no. 19
    
20.
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.  Back to cited text no. 20
    
21.
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 Univer Comput Informat Sci 2018.  Back to cited text no. 21
    
22.
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.  Back to cited text no. 22
    
23.
Mahajan A, Vaidya T, Gupta A, Rane SU. Authors reply to Pinninti et al., Niyogi and Baheti. Cancer Res Stat Treat 2020;3:136-7.  Back to cited text no. 23
  [Full text]  
24.
25.
Mahajan A. D.R. AI in Healthcare. Ai in Healthcare Artificial or Augmented Intelligence Choice is Yours. Available from: https://www.google.com/amp/s/www.express healthcare.in/amp/cancer-ca re/ai-in-healthcare-artificial-or-augm ented-intelligence-choice-is-yours-dr-abhi shek-mahajan/411998. [Last accessed on 2020 Feb 19].  Back to cited text no. 25
    




 

Top
 
  Search
 
    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

 
  In this article
References

 Article Access Statistics
    Viewed70    
    Printed0    
    Emailed0    
    PDF Downloaded17    
    Comments [Add]    

Recommend this journal