News | Artificial Intelligence | February 11, 2020

AI-assisted Radiologists Can Detect More Breast Cancer with Reduced False-positive Recall

A new study published in Lancet Digital Health shows a significant improvement in the performance of radiologists, before and after using AI

Mammograms of a 49-year-old woman with invasive lobular carcinoma on the right-side breast

Mammograms of a 49-year-old woman with invasive lobular carcinoma on the right-side breast. A small mass with micro-calcifications on the right-side breast was detected correctly by AI with an abnormality score of 96%. This case was recalled by 7 out of 14 radiologists (4 breast radiologists and 3 general radiologists) initially (without AI) and all 14 radiologists recalled this case correctly with the assistance of AI.

February 11, 2020 — A new study, published in Lancet Digital Health, shows the added value of AI-aided breast cancer detection from mammography images.

The study, conducted by Korean academic hospitals and Lunit, a medical AI company specializing in developing AI solutions for radiology and oncology, features large-scale data of over 170,000 mammogram examinations from five institutions across South Korea, USA and the UK, consisting of Asian and Caucasian female breast images. The dataset includes over 36,000 biopsy-proven, independent cancer positive cases — the largest scale of cancer data among mammography-related AI studies.

“It is an unprecedented quantity of data with accurate ground truth--especially the 36,000 cancer cases, which is seven times larger than the usual number of datasets from resembling studies conducted previously,” said Hyo-Eun Kim, the first author of the study and chief product officer at Lunit. “The quality of data has also been assured, with ethnic diversity, covering various imaging devices and scanning conditions. The marriage between the diversity of the dataset and the uniqueness of our algorithm, designed in interaction with one another, has been key to years of development of Lunit INSIGHT MMG since early 2016.”

The study shows a significant improvement in the performance of radiologists, before and after using AI. According to the study, the AI alone showed 88.8 percent sensitivity in breast cancer detection, whereas radiologists alone showed 75.3 percent. When radiologists were aided by AI, the accuracy increased by 9.5 percent to 84.8 percent.

One of the major findings also shows that AI, in comparison to the radiologists, displayed better sensitivity in detecting cancer with mass (90 percent vs 78 percent) and distortion or asymmetry (90 percent vs 50 percent). The AI was better in the detection of T1 cancers, which is categorized as early-stage invasive cancer. AI detected 91% of T1 cancers and 87% of node-negative cancers, whereas the radiologist reader group detected 74 percent for both.

Breast density is also an important factor in diagnosing mammograms, as dense breast tissues, mostly from the Asian population, are harder to interpret since dense tissue is more likely to mask cancers in mammograms. The findings show that the diagnostic performance of AI was less affected by breast density, whereas radiologists’ performance was prone to density, showing higher sensitivity for fatty breasts at 79.2 percent compared to dense breasts at 73.8 percent. When aided by AI, the radiologist’s sensitivity when interpreting dense breasts increased by 11 percent.

“One of the biggest problems in detecting malignant lesions from mammography images is that to reduce false negatives — missed cases — radiologists tend to increase recalls, casting a wider safety net, which brings an increased number of unnecessary biopsies,” said Eun-Kyung Kim, the corresponding author of the study and a breast radiologist at Yonsei University Severance Hospital. “It requires extensive experience to correctly interpret breast images, and our study showed that AI can help find more breast cancer with lesser recalls, also detecting cancers in its early stage of development.”

The study has been published online on 6 February 2020, in Lancet Digital Health. Lunit INSIGHT MMG is commercially available and is being used clinically, approved by Korea Ministry of Food and Drug Safety and pending approval by European CE within the first quarter and FDA clearance by later this year. It is available for free online demo at https://insight.lunit.io.

Related Content

Researchers from the School of Biomedical Engineering & Imaging Sciences at King's College London have automated brain MRI image labeling, needed to teach machine learning image recognition models, by deriving important labels from radiology reports and accurately assigning them to the corresponding MRI examinations

Getty Images

News | Magnetic Resonance Imaging (MRI) | July 28, 2021
July 28, 2021 — Researchers from the School of Biomedical Engineering & Imaging Sciences at...
Videos | Artificial Intelligence | July 22, 2021
This is an overview of trends and technologies in radiology...
An example of Viz.AI's pulmonary embolism AI application and mobile alert to the physician on-call. Viz.AI and Avicenna.AI Partner to Launch Artificial Intelligence Care Coordination for Pulmonary Embolism and Aortic Disease

An example of Viz.AI's pulmonary embolism AI application and mobile alert to the physician on-call. 

News | July 21, 2021
July 21, 2021 — Artificial int...
Registration is now open for the Radiological Society of North America (RSNA) 107th Scientific Assembly and Annual Meeting, the world’s largest annual radiology forum, to be held at McCormick Place Chicago, Nov. 28 – Dec. 2, 2021

Getty Images

News | RSNA | July 21, 2021
July 21, 2021 — Registration is now open for the Radiological Society of North America (...
Artificial intelligence-powered diagnostic tool spots asymptomatic prostate cancer in seconds

(L-R) Associate Professor Peter Brotchie (St Vincent's), Dr Ruwan Tennakoon (RMIT), Professor John Thangarajah (RMIT), Dr Mark Page (St Vincent's). Image courtesy of St Vincent's Hospital Melbourne

News | Prostate Cancer | July 19, 2021
July 19, 2021 — Prostate cancer is the most diagnosed
According to ARRS’ American Journal of Roentgenology (AJR), return to routine screening for BI-RADS 3 lesions on supplemental automated whole-breast US (ABUS) substantially reduces the recall rate, while being unlikely to result in adverse outcome

Normal right mediolateral oblique (A) and craniocarudal (B) view screening mammograms demonstrate density C breasts. Coronal (C), transverse (D), and reconstructed lateral (E) views from supplemental automatic breast ultrasound (ABUS) demonstrates 7 mm circumscribed slightly hypoechoic circumscribed lesion at 11 o’clock position in right breast. Lesion was classified as BI-RADS 3. Patient has undergone yearly mammograms for 4 years following the ABUS examination with no breast cancer diagnosis.

News | Breast Imaging | July 16, 2021
July 16, 2021 —...
Volpara now holds 98 granted patents across 27 countries
News | Breast Imaging | July 16, 2021
July 16, 2021 — Volpara has recently been granted another patent by the...
Interviews with the 1,000 at-risk patients also led to some interesting conclusions, Fine said. Even in the early days of 3D mammography, most women interviewed had at least heard about the technology.

Getty Images

News | Mammography | July 15, 2021
July 15, 2021 —  For any type of diagnostic test, technology that helps improve the reliability of test results is a
The Radiological Society of North America (RSNA), the American Society of Neuroradiology (ASNR), and the Medical Image Computing and Computer Assisted Interventions (MICCAI) society have announced the launch of the 10th annual Brain Tumor Segmentation (BraTS) challenge

Getty Images

News | RSNA | July 14, 2021
July 14, 2021 — The Radiological Society of North America (...