In a groundbreaking study conducted by experts at the Postgraduate Institute of Medical Education and Research (PGIMER) in Chandigarh and the Indian Institute of Technology (IIT), New Delhi, artificial intelligence (AI) has proven to be as effective as experienced radiologists in detecting gall bladder cancer (GBC) at a hospital in Chandigarh. The findings, published in The Lancet Regional Health – Southeast Asia journal, reveal the potential of AI in early cancer diagnosis, a significant breakthrough given the high mortality rate associated with GBC.
GBC is a highly aggressive form of cancer, but its early diagnosis is a major challenge. Often, benign gallbladder lesions can appear similar in medical images, making it difficult to distinguish cancer from non-cancerous conditions.
The team at PGIMER and IIT New Delhi set out to develop and test a deep learning (DL) model for GBC detection using abdominal ultrasound. Deep learning (DL) is a type of AI that helps computers process data in a way inspired by the human brain.
To test the AI model’s accuracy, the researchers collected abdominal ultrasound data from patients with gallbladder lesions at PGIMER between August 2019 and June 2021. They trained the DL model using data from 233 patients, validated it with 59 patients, and then tested it on 273 patients. The AI model’s performance was evaluated based on its ability to correctly detect GBC, measured by sensitivity, specificity, and an accuracy score known as the area under the receiver operating characteristic curve (AUC).
In the test group, the AI-powered model demonstrated a sensitivity of 92.3%, specificity of 74.4%, and an AUC of 0.887 in detecting GBC. Remarkably, these results were on par with the performance of two radiologists who independently reviewed the ultrasound images.
The DL-based approach also excelled in detecting GBC in cases with complicating factors such as the presence of gallstones, contracted gallbladders, small lesion sizes (less than 10 mm), and neck lesions. The AI model even outperformed one of the radiologists in identifying a specific type of GBC known as “mural thickening,” despite a slight reduction in specificity.
However, the study has its limitations, as it relied on data from a single center. The researchers emphasize the need for further studies involving multiple medical centers to validate these promising results on a broader scale.
This study highlights the potential of AI in revolutionizing cancer diagnosis in India and worldwide, offering hope for earlier detection and improved outcomes for patients with gall bladder cancer.
Read More: https://clipper28.com/en/uks-second-largest-city-birmingham-declares-itself-bankrupt/
 in Chandigarh){kind=link}