Deep learning predicts metastasis probability in patients with non-small cell lung cancer

While examining the possible benefits of deep learning for predicting N2 metastasis in patients with non-small cell lung cancer (NSCLS), researchers found important developments when implementing a new N2 signature model. 

Lung cancer staging is crucial, but it isn’t unheard of for clinicians to initially diagnose a patient’s nodal category incorrectly. In fact, PET/CT scans can only diagnose N2 disease with about 70-85% sensitivity. Correct staging and categorization are key in a treatment plan, and this new research suggests that deep learning can play a vital role.

In the study, a total of 3,096 patients with clinical stage 1 NSCLC were examined from May-October 2020 using three clinical models for predicting nodal involvement. One included an internal cohort to establish a deep learning signature, one external cohort that had been subsequently created to examine the proposed signature’s predictive effectiveness, and one multi-centered diagnostic trial. With N2 metastasis in mind, the team examined the ability of their deep learning signature to predict a prognosis.

The results showed the signature’s AUCs in each group (internal, external, and prospective) to be 0.82, 0.81 and 0.81, respectively. In the internal test set and external cohort, higher deep learning scores predicted poorer overall survival and recurrence-free survival. Of note, two of the models used pathologic, radiologic and clinical information, while the study’s signature model used only information from CT images, which showed higher predictive performance for N2 disease.

“Thanks to the simplicity of using only CT images, it is expected that the clinical applicability of this model would be better than that of conventional models," Chang Min Park, from the Department of Radiology at Seoul National University College of Medicine, and co-author, Jong Hyuk Lee, with the Department of Radiology at Seoul National University Hospital in Seoul, Korea wrote. 

With the aid of deep learning, the staging and categorization of non-small cell lung cancer proved to be more accurate than with the use of PET/CT scans alone in predicting N2 metastasis. The results of this study could be a valuable tool for treating patients with NSCLC and improving their prognoses, the researchers suggest. 

You can read the complete study here.

Hannah murhphy headshot

In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She joined Innovate Healthcare in 2021 and has since put her unique expertise to use in her editorial role with Health Imaging.

Around the web

The two companies aim to improve patient access to high-quality MRI scans by combining their artificial intelligence capabilities.

Positron, a New York-based nuclear imaging company, will now provide Upbeat Cardiology Solutions with advanced PET/CT systems and services. 

The nuclear imaging isotope shortage of molybdenum-99 may be over now that the sidelined reactor is restarting. ASNC's president says PET and new SPECT technologies helped cardiac imaging labs better weather the storm.