New Delhi: Artificial intelligence (AI), despite the obvious anxieties about its capacity to create job losses, is slowly and steadily gaining ground in the medical world. Doctors say AI has truly revolutionised the medical field and changed the way healthcare is approached.
Dr Rajeev Kumar, professor of urology and associate dean of academics at AIIMS Delhi, said that AI had future potential use in medicine and healthcare. “AI is already being used for interpretation of images like X-rays, CT scans, MRI, etc., because it is easy to train AI tools to read images,” he said.
“It is quicker and more consistent than humans at triage because humans might not appreciate all features of every image, particularly when the turnaround time has to be short, but AI has been trained to look at all those features.”
In fact, AI algorithms are often able to pick up features that humans cannot even read. “While evaluating images, there could be a difference of opinion among physicians. Machines, in some such cases, may be able to resolve the difference because of their consistent algorithms,” said Kumar.
He added, “As of now, we are not dependent on AI to make a diagnosis, but we use its features.” The only form of AI currently used in urology is in nomograms and algorithms to predict the possible outcomes of patients suffering from the disease. However, AI is expected to have a major role in the future, particularly in the reporting of MRI for prostate, bladder and kidney cancers, suggesting treatment options and predicting outcomes for diseases. “AIIMS Delhi has been designated a Centre of Excellence for AI in healthcare by the Union health ministry. The faculty of the departments there have started projects in three broad areas using image recognition: radiology (evaluating chest X-rays), dermatology (skin lesions) and ophthalmology (corneal and retinal lesions). We hope these tools will be ready soon for deployment,” said Kumar.
Dr Meinal Chaudhry, director, radiodiagnosis and intervention radiology, Aakash Healthcare, explained where AI has advantages. “With the help of AI, we can check the quality of an image whether it is reportable or not, which means even when the image is poor we don’t have to recall the patient again for another imaging. This is a huge value addition for the patient,” said Chaudhry.
The second advantage of AI, Chaudhry added, was the prioritisation of critical X-ray reporting. “AI can tell us which cases are critical and need immediate attention because they might be life threatening,” she said. She said AI can also analyse the trends of the previous X-rays. When a patient comes to the radiology department for X-ray, an AI software will see all images taken earlier too and prompt if something appears noticeable.
Explaining this, she gave the example of the detection of life-threatening pneumothorax, when the air collects in between the parietal and viscera pleura resulting in the lungs collapsing. Sometimes some air leaks into the pleural cavity and collapses the lung, so there is a very thin line between normal air intake and pneumothorax. “AI can prompt an alert to the radiologist if a pleura is there and get it diagnosed as soon as the X-ray is taken. It is definitely a lifesaving algorithm,” said Chaudhry.
AI is being used for preventive health check-ups like chest X-ray and spine MRI, said Kabir Mahajan, associate director of Mahajan Imaging. “If an X-ray report is normal and processed by AI, it automatically sends a report to the radiologist, who either accepts it or edits the report. But if AI detects something abnormal, it puts markers on the X-ray so that the radiologist knows what to look out for or consider,” said Mahajan. This helps the radiologist arrive at a more informed decision when writing his final report.
AI is also being used in CT scan, reducing the scan time to 30 per cent without compromising on the quality. “The scan is done within 10 mins, providing less discomfort to patients and also allowing the hospitals to conduct more scans in a day,” said Mahajan. “We have been doing this since 2019 but in different formats. Through AI, we are picking up more abnormalities but the price of tests remains the same for patients and they are getting more value for their money.”
Dr Mohinish Chhabra, director, gastroenterology, Fortis Hospital, Mohali, said that AI was useful in the screening of colorectal cancer, which is the fifth leading cancer in India and claims several lives every year. He said, “We were the first to launch an AI screening device to detect pre-cancerous growth called polyps/adenomas, which usually take 10-15 years to become cancerous. So, if polyps are detected and removed at an early stage, the risk of cancer decreases by up to 90 per cent and the probability of early death goes down.”
The gastroenterologist said that it had been estimated that 55-60 per cent of all cancer cases of this type occurred due to polyps missed by a colonoscopy. “AI is a boon in colon cancer screening as this tool doubles the detection of polyps measuring less than 10mm. They can be identified and removed during the colonoscopy procedure,” he said. There are a few cancers in the body that can be prevented. These include cervical and colon cancer. Other cancers can be detected early like breast cancer but only after it has developed. So the major difference is in detecting an early cancer and detecting pre-cancer which can later turn into cancer.
Dr Sajjan Rajpurohit, director medical, oncology, Max Superspeciality Hospital, Shalimar Bagh, said AI technology had made a significant transformation in precision medicine. “AI technology has made it simpler to identify the efficacy of drugs for individual patients. In the field of oncology drug discovery, AI has enabled the identification of patients with specific genetic mutations, allowing targeted drug treatments,” explained Rajpurohit.
He added, “Besides, AI tools aid doctors in predicting patient outcomes, enabling them to make tailor treatments for better results. Early detection is crucial in cancer treatment, and AI-enabled mammograms and metabolomics-based blood sample analysis are proving vital in detecting breast cancer early. With AI technology making healthcare more accessible, efficient and effective, the future of medicine looks brighter than ever before.”
Gynaecologist Rahul Manchanda at PSRI Hospital, said AI was being used in many fields including gynaecology and women’s health. “The data fed in the computer and machines helps in forming algorithms to diagnose complex problems and then treat them with precise management through minimally invasive surgical techniques and robotic surgery,” said Manchanda. ‘Since all endoscopic and robotic surgical techniques involve computers as an image transfer system, deep learning methods help with more precise surgery with these systems.”
AI incorporation helps in minimally invasive surgical techniques like laparoscopy and hysteroscopy being more precise. AI guides doctors in avoiding complications and hence promise better results of surgery for the patient and better patient outcomes. However, robotic surgery goes one step further with AI improving the range of movement and the preciseness. This decreases complications and blood loss further. Robotic surgery is, of course, ergonomic for the surgeon too because he or she can sit comfortably in front of a computer screen guiding the robot precisely.
