Screening News
Artificial Intelligence for Breast Cancer Detection in Screening Mammography: A Paired-readers Prospective Interventional Screen Positive Trial
Dembrower et al., The Lancet Digital Health DOI: 10.1016/S2589-7500(23)00153-X, 2023.
Mammography screening for breast cancer faces challenges due to variability in radiologist accuracy. Artificial intelligence (AI) shows promise as an independent reader for mammogram screenings. Although retrospective studies have provided claws of the diagnostic value of AI in reading mammograms, this study aimed to undertake a prospective trial to this effect. study involved a reading of 55,581 mammograms by one radiologist, by two radiologists, by one radiologist plus AI, or AI alone. Results indicated that AI combined with one radiologist or AI alone or alongside radiologists proved inferior compared to reading by two radiologists. The study suggests AI can be effectively integrated with risk management in real-world screening settings.
Early-stage Breast Cancer Detection in Breast Milk
Saura et al., Cancer Discovery OF1-OF12. doi: 10.1158/2159-8290.CD-22-1340, 2023.
Breast cancer during pregnancy (PrBC) and postpartum (PPBC) are typically detected at the aggressive stage with poor prognosis. To develop a screening test to recognize the development of PrBC and PPBC, using genomic and liquid biopsy technologies, the authors have unveiled the usefulness of breast milk from breast cancer patients in detecting circulating tumor DNA (ctDNA) indicative of breast tumors harboring tumor-specific variants. Interestingly, the authors also documented two cases of such detectable diagnostic changes in breast cancer 6 to 18 months before the standard diagnosis. This study holds significant promise in breast cancer detection and offers a potential avenue for improving outcomes and prognosis for postpartum breast cancer patients.
https://pubmed.ncbi.nlm.nih.gov/37704212/
Conformable Ultrasound Breast Patch For Deep Tissue Scanning And Imaging.
Du et al. Sci. Adv., 9(30):eadh5325. doi: 10.1126/sciadv.adh5325, 2023.
Researchers have developed a first-in-class wearable breast patch (cUSBr-Patch) that addresses challenges in integrating ultrasound and wearable technologies imaging large, curvilinear organs like the breast. The patch uses a nature-inspired design with a phased array for large-area, deep scanning, and multiangle imaging. In preclinical and clinical trials, cUSBr-Patch exhibited promising results, raising the possibility of using the device as an effective noninvasive tool for real-time monitoring changes in soft breast tissues.
https://pubmed.ncbi.nlm.nih.gov/37506210/
Alternative Methods to Measure Breast Density in Younger Women.
Lloyd et al., British Journal of Cancer, DOI: 10.1038/s41416-023-02201-5, 2023.
Most of our understanding of the diagnostic significance of breast density is derived from mammographic studies, which are generally not recommended for younger women under 40 years of age with no family history of breast cancer. This highlighted the need to develop alternative modes to quantitate breast density in younger women. This study examined the diagnostic value of Optical Breast Spectroscopy (OBS) and Dual X-ray Absorptiometry (DXA) in younger women (n = 539, age 18–49 years). The study reported an acceptability of over 93% for OBS and DXA methods to measure breast density in young women aged <40 years.
https://pubmed.ncbi.nlm.nih.gov/36828870/
Non-invasive Screening of Breast Cancer from Fingertip Smears – A Proof of Concept Study.
Russo et al., Scientific Reports 13(1): 1868, 2023.
Proteomics combined with mass spectrometry was performed using fingertips/fingerprint/ sweat deposits of patients with breast cancer, which indicated a high (97.8%) accuracy score when resulting peptides were segregated into the benign, early, or metastatic cancer subgroups.
https://pubmed.ncbi.nlm.nih.gov/36725900
Diagnostic Accuracy of A Three-protein Signature in Women with Suspicious Breast Lesions: A Multicenter Prospective Trial.
Lee at al., Breast Cancer Research 25(1):20, 2023.
To counteract the limitations of mammography in younger women with high breast density and to aid in population-based screening approaches, research groups and pharmaceuticals are consistently exploring noninvasive and/or biofluid-based diagnostic approaches. One such approach included using a 3-protein signature in a blood-based test in suspected breast cancer cases (n = 642). The three proteins are carbonic anhydrase I (CA1; role in microcalcification in breast cancer), apolipoprotein C-I (APOC1; downregulated serum marker in cancer), and neural cell adhesion molecule L1-like protein (CHL1; a tumor suppressor in breast cancer). This study concluded that the overall accuracy of the 3-protein signature in predicting breast cancer in suspected cases was 70.6%.
https://pubmed.ncbi.nlm.nih.gov/36788595/
Artificial Intelligence (AI) for Breast Cancer Screening: Breast Screen Population-based Cohort Study of Cancer Detection.
Marinovich et al., EBioMedicine 90:104498, 2023.
This study examined the effectiveness of using Artificial intelligence (AI) compared to radiologists for screening breast cancer in the context of cancer detection rates and workload distribution. The method was applied to a cohort of 108,970 consecutive mammograms in a retrospective study. The study reported a comparable AI sensitivity with radiologists but with a reduced AI specificity. Interestingly, AI picked internal cancers not reported by radiologists, raising new testable possibilities in future double-blinded studies.
https://pubmed.ncbi.nlm.nih.gov/36863255/
A Prospective Evaluation of Breast Thermography Enhanced by a Novel Machine Learning Technique for Screening Breast Abnormalities in a General Population of Women Presenting to a Secondary Care Hospital.
Bansel et al., Frontiers in Artificial Intelligence 5, 2022.
This study examined the feasibility of using a new device based on thermal imaging technology to detect suspected cases of early breast cancer in younger women. The authors compared the resulting data with the findings based on conventional mammography for the same cohort of women. The study reported an overall sensitivity and specificity of 95.24% and 88.58%, respectively, for Thermalytix. Interestingly, the sensitivity and specificity reached 100% and 81.65%, respectively, in a subset of 168 women with high breast density. The device could narrow the suspected women for a secondary screening with mammography and/or MRI for high breast density.
https://pubmed.ncbi.nlm.nih.gov/36686848/
Development and Validation of a Short-term Breast Health Measure as a Supplement to Screening Mammography.
Daily et al., Biomarker Research 10, 76, 2022.
This study tested the efficacy of using tear-based proteins to detect breast cancer while comparing the outcome with mammography using a large sample size of 847 tear specimens from women. Overall, using two diagnostic thresholds, the study reported a sensitivity range (52% to 90%) and specificity range (31% to 79%) of their proteomic approach and discussed the feasibility of using tears’ analysis to supplement mammography.
https://pubmed.ncbi.nlm.nih.gov/36284356/
News & Updates
Connect With Us
