The lack of specialists is a major obstacle to reducing deaths from cancer, given the rising number of cases. The largest cancer hospital in India, Tata Memorial Hospital (TMH) in Mumbai, is using artificial intelligence (AI) to close this gap.
The hospital has created a “Bio-Imaging Bank” dedicated to cancer research, and it is using deep learning to create a customized algorithm that is specifically designed to detect cancer in its early stages. In the past year, it added information from 60,000 patients to the biobank.
This is all the information you require regarding the project.
How does artificial intelligence fit into this whole “Bio-Imaging Bank” thing?
The main objective of the project is to build a comprehensive archive of radiology and pathology images that is closely connected to clinical data, outcome information, treatment details, and other metadata. This all-inclusive resource is thoughtfully created for the rigorous testing, validation, and training of artificial intelligence systems.
By the time the initiative is finished, it will have more patient data committed for both head and neck cancers than it had for lung cancer, with at least 1000 patients for each cancer type. In addition to building a database, the project uses the collected data to train and test several AI algorithms that address medically relevant tasks like identifying lymph node metastases, classifying and segmenting nuclei, predicting biomarkers (like EGFR in lung cancer and HPV in oropharyngeal cancer), and predicting therapy response.
The Department of Biotechnology, in partnership with IIT-Bombay, RGCIRC-New Delhi, AIIMS-New Delhi, and PGIMER-Chandigarh, is funding the multi-institutional study.
How might AI aid in the early identification of cancer?
By simulating the information processing of the human brain, artificial intelligence greatly aids in the identification of cancer. Artificial intelligence (AI) is used in cancer detection to analyze radiological and pathological pictures. It learns from large datasets to identify distinctive traits linked to different types of cancer. Through the identification of tissue alterations and possible cancers, this technology aids in early detection.
The TMC Head of Radiodiagnosis, Dr. Suyash Kulkarni, described how the team uses AI in radiology. Longitudinal patient data produced by comprehensive imaging is helpful in analyzing behavior, response to treatment, illness recurrence, and overall survival. This data is used by AI and machine learning algorithms to create models that predict tumor survival and determine how aggressively to treat patients.
Segmenting and annotating photos, defining characteristics, and labeling tumors as malignant, inflammatory, or edematous are all steps in the process of creating a tumour image library. Images and clinical data are correlated with genomic sequences, histopathology, immunohistochemistry reports, and biopsy results to create a variety of algorithms.
With this strategy, TMH can create algorithms for various tumor types, evaluate treatment outcomes straight from picture data, and prevent needless chemotherapy for those who are expected to not respond, all of which have clinical value. Thousands of photos of breast cancer are used to create predictive and diagnostic models utilizing the biobank. AI and ML analysis is performed on the images with technical assistance from partners like IIT-Bombay.
Is this technology utilized right now?
Indeed. In the past year, TMH has started utilizing AI to lower radiation exposure for pediatric CT scan patients, adding the data of 60,000 patients to the biobank.
By using artificial intelligence algorithms to enhance photos, we were able to achieve a 40% decrease in radiation through an innovative approach. This guarantees a considerable reduction in children’s radiation exposure while preserving diagnostic quality without sacrificing it. Dr. Kulkarni stated, “This is an example of the impactful algorithms we aim to develop.”
A particular algorithm for thoracic radiology, which focuses on imaging and diagnosing diseases related to the thoracic portion of the body, notably the chest area, is also being used by the department in the ICU on a pilot basis. When physicians cross-check, the AI’s diagnosis, which was made instantly, turns out to be 98 percent accurate.
“Right now, we are validating a variety of AI algorithms, including the thoracic suit. This specialized instrument recognizes diseases such as pneumothorax and nodules by interpreting digital chest X-rays. For example, the AI program automatically delivers a diagnosis during an MRI in the intensive care unit, which is then confirmed by our radiologists,” he stated. “It saves time by assisting in early diagnosis,” he continued.
So, will AI contribute to a decrease in cancer deaths in the future?
AI has the potential to revolutionize cancer therapy in the future, especially in reducing mortality in rural India. AI has the ability to optimize therapeutic outcomes by customizing treatment techniques based on a variety of patient profiles.
The director of TMC, Dr. Sudeep Gupta, sees a time when AI can quickly identify cancer with a single click, doing away with the need for laborious testing and making it possible for ordinary practitioners to diagnosis complicated cancers. This technology has the potential to greatly improve cancer treatment precision. “AI improves accuracy through continuous learning, guaranteeing prompt cancer diagnoses, enhancing patient outcomes, and supporting healthcare professionals in decision-making processes,” the speaker stated.
However, the use of AI technologies sparks discussions about the possible replacement of human radiologists, which is subject to regulatory scrutiny and opposition from certain medical professionals and organizations.
