A team of researchers led by Dr. Soon Ho Yoon, Ph.D., of Seoul National University College of Medicine, “trained a deep-learning algorithm on thousands of x-rays from patients with active and healed tuberculosis and found the model accurately differentiated between the two,” in a study published August 3, 2021, in Radiology, according to Will Morton of AuntMinnie.com.

Dr. Yoon and his team found:

• “In two test sets that included radiographs depicting active and healed tuberculosis (test set 1, n = 148; test set 2, n = 200), a deep learning model (AUCs, 0.83 and 0.84, respectively) differentiated active from healed tuberculosis on radiographs, comparable with the performance of expert readers (AUCs, 0.69-0.80 [P < .001 to P = .23] and 0.71-0.80 [P < .001 to P = .08]).
• A higher degree of sputum smear positivity increased model output (odds, 1.36 per 0.1 increase in the disease activity score), which decreased during treatment by 0.327 per month on average (P < .001).

“‘This deep-learning network may be advantageous in countries with a high burden of tuberculosis where spontaneously healed or previously treated patients with tuberculosis are prevalent,” the researchers concluded. “Those countries usually suffer from low incomes and a shortage of expert imaging professionals,'” reported Morton.

Sources:

Seowoo Lee, Jae-Joon Yim, Nakhon Kwak, Yeon Joo Lee, Jung-Kyu Lee, Ji Yeon Lee, Ju Sang Kim, Young Ae Kang, Doosoo Jeon, Myoung-jin Jang, Jin Mo Goo, Soon Ho Yoon Deep Learning to Determine the Activity of Pulmonary Tuberculosis on Chest Radiographs Radiology August 3, 2021. Available: https://pubs.rsna.org/doi/abs/10.1148/radiol.2021210063 Accessed: August 9, 2021.

Morton, Will. “AI helps triage tuberculosis patients on chest x-ray.” AuntMinnie.com August 5, 2021. Accessed: August 9, 2021