January 4, 2021 | Updates from Ibex Medical Analytics and Paige on AI-assisted breast cancer diagnoses, using vocal biomarkers to diagnose pulmonary hypertension, whole genome sequencing in diagnoses, and more.
Ibex Medical Analytics, a pioneer in artificial intelligence (AI)-based cancer diagnostics and Institut Curie, France’s leading cancer center, have launched a research partnership aimed at improving diagnosis of breast cancer with AI. This research partnership, the first of its kind, will include a rich dataset of breast biopsy slides, digitized using a digital pathology scanner, and analyzed for cancer detection by Ibex’s Galen Breast solution. Independently, multiple pathologists from Institut Curie will diagnose the slides, followed by blinded analysis of the AI-solution’s performance. Galen Breast, the first AI solution used for detection of breast cancer in pathology, was developed utilizing state-of-the-art AI and machine learning techniques, and trained on hundreds of thousands of image samples. The solution is already deployed at the pathology institute of Maccabi Healthcare Services, Israel’s second largest HMO, where it is used as a second read application. Press release.
Novigenix SA Switzerland, a pioneer in Immuno-Transcriptomics for precision medicine today announced that a consortium, including PamGene International BV and Radboud University (The Netherlands), has been awarded a grant up to €1 million to develop a blood based multi-omics test to predict response to immune checkpoint inhibitors in cancer. The aim of the consortium is to develop a novel in vitro diagnostic (IVD) liquid biopsy test for accurate prediction of response to immune checkpoint inhibitors (ICIs), one of the main treatments for patients with metastatic cancers, leveraging a unique multi-omics approach of combining mRNA and kinase activity profiling. The biomarker development strategy is based on whole blood transcriptome and kinome profiling technologies. Moreover, the consortium will take advantage of the most advanced machine learning methods to develop and validate the AI based multi-omics signatures. Press release.
Mayo Clinic and Vocalis Health are collaborating to research and develop new voice-based tools for screening, detecting and monitoring patient health. The collaboration will begin with a study to identify vocal biomarkers for pulmonary hypertension (PH) which could help physicians detect and treat PH in their patients. In a previous trial with Vocalis Health, the Mayo research team established a relationship between certain vocal characteristics and PH. In this new collaboration, Mayo will conduct a prospective clinical validation study to further develop PH vocal biomarkers. The clinical validation will utilize Vocalis Health’s proprietary software, which can operate on any connected voice platform (mobile, computer, tablet, etc.), and analyzes patients’ health based on voice recordings. Following this initial phase, researchers will work to identify vocal biomarkers targeting additional diseases, symptoms and conditions. Press release.
The National Kidney Foundation and Renalytix AI are partnering to improve kidney disease diagnosis and assessment in the earliest stage of the disease that will ultimately guide clinical interventions and treatments for improved outcomes and quality of life. Renalytix AI’s KidneyIntelX is an in vitro diagnostic platform that uses a machine learning enabled algorithm to interpret proprietary biomarkers from a simple blood draw and electronic health record features to generate a patient specific risk score for disease progression. The KidneyIntelX risk score categorizes a patient as low, intermediate or high risk for kidney disease progression and kidney failure. The individual patient's score will be integrated with risk appropriate NKF recommended clinical actions and education resources as part of a comprehensive approach to improve outcomes. NKF and RenalytixAI will collaborate on future studies to assess current care standards in early stage kidney disease, identify barriers to better care and opportunities for improvement. Long-term initiatives will be developed as the partnership unfolds. Press release.
Illumina and Harvard Pilgrim Health Care have announced a risk-sharing agreement to make whole-genome sequencing (WGS) available to certain Harvard Pilgrim members, effective January 1, 2021. The program will leverage WGS to support faster diagnoses of genetic diseases in children, potentially eliminating the long, costly diagnostic odyssey experienced by many families, with the goal of improving patient outcomes. Through this agreement, Harvard Pilgrim and Illumina will work together to evaluate how insurance coverage of WGS impacts patient care and healthcare costs. To date, there are more than 20 peer-reviewed publications demonstrating the clinical utility of WGS in over 3,000 patients with suspected genetic diseases. Health economic models predict that implementing WGS earlier in the diagnostic workup is likely to be cost-neutral or even save payers money. During the term of the agreement, Harvard Pilgrim will cover WGS, through their network of lab providers, for pediatric patients meeting specific criteria. Illumina and Harvard Pilgrim will share the risk on genetic testing costs. Achieving a rapid diagnosis will prove most valuable for all—ending uncertainty for patient families and potentially halting unnecessary spending on the clinical side. Together, Harvard Pilgrim and Illumina will analyze the data, adjudicate the financials, and prepare a peer-reviewed study for publication. Press release.
Paige received a CE mark for Paige Breast, its breast cancer detection software that indicates suspicious areas for further review to pathologists evaluating breast biopsies. The company also received a CE mark for Paige Prostate Grading & Quantification, a digital diagnostic that offers slide level information for primary and secondary Gleason patterns and tumor size to inform treatment planning. With these CE marks, both products are now available for use in laboratories and hospitals in the European Economic Area, Switzerland and the UK. Paige Breast is designed to draw pathologists’ attention to concerning features on a breast tissue slide and provides slide and case level predictions about the presence of cancer. This functionality enables pathologists to more easily, efficiently and confidently identify small foci of cancer that can be easily missed. Paige continues to enhance functionality for Paige Breast, including quantification and subtyping functionality, for future deployment. Press release.