A recent study published in The Lancet Regional Health – Southeast Asia journal has highlighted the potential of artificial intelligence (AI) in Detecting Gallbladder Cancer (GBC) with diagnostic performance comparable to experienced radiologists. GBC is a highly aggressive cancer with a poor detection rate and high mortality. Early diagnosis is challenging due to the similarity in imaging features with benign gallbladder lesions.

The research, conducted by the team at the Postgraduate Institute of Medical Education and Research (PGIMER) in Chandigarh and the Indian Institute of Technology (IIT) in New Delhi, aimed to develop and validate a deep learning (DL) Model for GBC detection using abdominal ultrasound. They compared the DL model's performance to that of radiologists.

Deep learning, a subset of AI, involves training computers to process data in a manner inspired by the human brain.

The study used abdominal ultrasound data from patients with gallbladder lesions collected between August 2019 and June 2021 at PGIMER, a tertiary care hospital. The DL model was trained on a dataset of 233 patients, validated on 59 patients, and tested on 273 patients.

The DL model's performance was evaluated in terms of sensitivity, specificity, and the area under the receiver operating characteristic curve (AUC), a common measure of diagnostic test accuracy.

Two radiologists independently reviewed ultrasound images, and their diagnostic performance was compared to that of the DL model.

In the test set, the DL model demonstrated a sensitivity of 92.3%, specificity of 74.4%, and an AUC of 0.887 for Detecting Gbc. These results were comparable to those of the radiologists.

The DL-based approach exhibited high sensitivity and AUC for detecting GBC in the presence of various factors, including gallstones, contracted gallbladders, small lesion size (less than 10 mm), and neck lesions, all of which were also comparable to the radiologists.

Additionally, the DL model showed higher sensitivity in detecting the mural thickening type of GBC compared to one of the radiologists, although specificity was slightly reduced.

The study's authors noted that "the DL-based approach demonstrated diagnostic performance comparable to experienced radiologists in detecting GBC using ultrasound." They recommended further multicenter studies to fully explore the potential of DL-based GBC diagnosis.

However, they acknowledged some limitations, such as the study's reliance on a single-center dataset, emphasizing the need for broader validation through multicenter studies. Additionally, the study's knowledge cutoff date in 2021 means that subsequent developments in DL and GBC diagnosis may not be reflected in the findings.