On January 5, 2022 Median Technologies (ALMDT:PA) reported outstanding performance for its iBiopsy Lung Cancer Screening (LCS) CADe algorithm in detecting potentially cancerous lung nodules (Press release, MEDIAN Technologies, JAN 5, 2022, View Source [SID1234598282]).

Schedule your 30 min Free 1stOncology Demo!
Discover why more than 1,500 members use 1stOncology™ to excel in:

Early/Late Stage Pipeline Development - Target Scouting - Clinical Biomarkers - Indication Selection & Expansion - BD&L Contacts - Conference Reports - Combinatorial Drug Settings - Companion Diagnostics - Drug Repositioning - First-in-class Analysis - Competitive Analysis - Deals & Licensing

                  Schedule Your 30 min Free Demo!

When screening for lung cancer, the entire lung must be scanned, which can generate hundreds of images for each patient to detect very small lung nodules. Time constraints and the very large numbers of images to review render it difficult for radiologists to exhaustively detect lung nodules without an automation tool. Without such automation tools, diagnostic errors may result when radiologists are dealing with fatigue. Automated detection tools can help radiologists read images and must offer high sensitivity to minimize false negatives and avoid missing nodules. These tools should also minimize the number of false positives to ensure that radiologists do not focus their attention and time on regions that are not clinically relevant, and will help to avoid lung biopsies as well as unnecessary follow up procedures for the patients.

iBiopsy LCS offers an integrated detection/diagnosis approach (CADe/CADx). The first step is detecting, as thoroughly as possible, all lung nodules in the CT scan images with minimal false negatives and false positives per scan. After automated detection, the diagnostic component (CADx) of iBiopsy LCS aims to reach high levels of sensitivity and specificity, accurately characterizing the presence of cancer while minimizing the false positive rate. False positives are one of the major barriers to adopting screening programs globally. It bears recalling that the outstanding characterization results of iBiopsy CADx were published on September 6, 2021 (performance characterizing cancerous nodules at all stages) and on November 23, 2021 (focus on stage 1 cancers).

The results released today by Median Technologies specifically focus on the lung nodule detection function, which applies Median’s proprietary deep learning algorithms to low-dose computed tomography (LDCT). They are based on a cohort of 888 patients from the LIDC/IDRI2 public database. The 10-fold cross-validation method was used for training (800 train/88 test patients). The overall result, obtained by aggregating the ten test fold results, showed a sensitivity of 94.9% for a false positive rate of 1 per CT scan.

According to publicly available data on the performance of lung CADe devices currently on the market, iBiopsy detection algorithms show the best combination of sensitivity/false positives per CT scan.

"After announcing iBiopsy’s first outstanding results for malignant vs. benign lung nodule characterization in 2021, we are proud to announce our detection algorithm’s performance," highlights Fredrik Brag, CEO and founder of Median. "These results give us reason to be optimistic as we enter the final phase of iBiopsy Lung Cancer Screening’s technological risk mitigation. We are now working on combining detection and diagnosis modules to determine integrated performance. We believe it is only by integrating these two functions in the same software as a medical device (SaMD) that we can remove the current barriers to implementing lung cancer screening programs. Large international trials have already shown the major impact of LDCT lung screening programs on mortality. In parallel, as demonstrated by studies such as the IASLC staging project and an I-ELCAP cohort study respectively, detecting and diagnosing the disease at its earlier stage, when lung nodules are still very small and manageable is key to drastically increasing the patient 5-year and 15-year overall survival rates, saving the lives of patients suffering from lung cancer," Brag added.