Controversy erupts after AIIMS New Delhi X ray report generated by 'AI Radiologist' goes viral

"Accordingly, neither the hospital nor Qure.ai makes no expressed or implied warranties or representations with respect to the merits, accuracy or the reliability of the report. The report should be correlated with patient history and relevant clinical findings in other medical tests and accordingly not be exclusively relied upon for arriving at a diagnosis, treatment plan, or other decision that may affect patient care," the disclaimer reads.  

Also read- AI-Powered ECG Model Predicts Postoperative Atrial Fibrillation (AF) After Cardiac Surgery

What sparked controversy and debate is that while some doctors called it a major technological leap, others raised concerns about patient safety and the future role of radiologists.

The doctor who shared the report said, "It’s a big leap for medical tech and a great workflow booster: faster reads, fewer misses, better triage. But the viral report itself reminds us — AI outputs are preliminary and must be clinically correlated. Radiologists interpret context, complexity, and patient nuance. The future is AI assisting radiologists, not replacing them."

However, Dr Suvarankar Datta, a radiologist currently training at AIIMS, New Delhi, and now an AI researcher, shared a detailed explanation of the AI report after receiving multiple queries from the public. 

Radiologist explains what the viral report really means

In a video statement, he explained that the report circulating online is part of a research study and cannot be treated as a final diagnostic report.

“AI reports today have absolutely no standing without a radiologist’s signature... Multiple laws in India prevent any autonomous diagnosis by software or AI. Every radiology or pathology report must be signed by a trained specialist,” he said. 

He further clarified that AI is widely used across countries and even in India by various vendors and labs to generate preliminary reads. In that context, the viral report is valid only as an initial AI-generated read before a radiologist signs it. It must be reviewed by a clinician or radiologist before being given to a patient as a final, verified diagnostic report.

Dr Datta added that while AI can generate preliminary reads, it cannot replace final human verification. He said such autonomous reports are only allowed under research protocols with proper consent and ethics approvals.

"So, if you ask me today what the standing of this report is, it's legally absolutely invalid. You can only consider it a report if it's signed by a radiologist. That's absolutely clear. ...Radiology residents who have reached out to me, you are not going to be replaced by AI anytime soon in the next couple of years," he said. 

AIIMS issues official clarification

Soon after the report went viral, the premier institution issued a statement on their 'X' handle confirming that the Radiology Department of AIIMS uses a clinically approved AI algorithm to triage chest radiographs to help manage the extremely high daily load of chest X-rays, which goes up to 1,000 radiographs a day.

"If any discrepancy is noticed in the AI results, a radiologist reviews the X-ray, along with the routine review of select positive radiographs. We manage very high volumes, up to 1,000 radiographs a day, and AI helps manage this load efficiently. The algorithm is tuned for a high negative predictive value (NPV), which reduces the chance of missing abnormalities while maintaining patient safety," the statement mentioned.