X-RAIS: The Third Eye
0 likes258 views
Laife is a data-driven company focusing on innovative healthcare solutions using AI, ML, and big data, including platforms like x-rais for medical image analysis and alfabeto for COVID-19 diagnosis support. X-rais, aimed at improving mammography analysis, is built on deep learning and seeks to enhance diagnostic accuracy and reduce clinical risks. Alfabeto provides a streamlined, AI-supported workflow for assessing patient conditions, ultimately aiming to reduce hospitalizations and improve treatment efficiency.
X-RAIS: The Third Eye
- 1. X-RAIS THE THIRD EYE Federico, Comotto | Data Scientist at Laife Reply [email protected]
- 2. Innovation for Healthcare LAIFE is the REPLY data company that combines the vertical expertise on HEALTHCARE & PHARMA with the technological specialization on AI, ML & BIG DATA. Medical Imaging, Radiomics, Natural Language Processing, Analysis of anatomical pathology reports, ChatBot, Behavior monitoring applications, Digital therapies are some of the use cases on which LAIFE works on with the largest public and private hospitals and the Pharma companies. #AI #ML #BIGDATA #DIGITALTHERAPEUTICS #RADIOMICS #ONCOLOGY #WELFARE #WELLBEING #CHATBOT #NLP For more information: [email protected]
- 3. AGENDA 1 Context 2 What is X-RAIS 3 ALFABETO 4 Conclusions
- 4. CONTEXT
- 5. mammograms are performed in U.S. yearly PAIN POINTS 40M of malignant tumors in mammograms are detected87% women suffer from breast cancer each year2.1M
- 7. X-RAIS is a medical platform (CE marking – IIa category in progress) based on Deep Learning, which can be specialized in different modalities and anatomical districts. Medical* AI Platform Layer Imaging Techniques Layer (XR, CAT, MRI, ...) * Medical certification in progress AI TO SUPPORT MEDICAL DECISIONS
- 8. OUR JOURNEY Jan 2018 Mar 2018 2019 April 2020 … X-RAIS was born with a prototype for analyzing medical images Sealed partnership with ICS Maugeri in mammography area Sealed partnership with Treviso Hospital to detect suspicious lesions in mammograms ALFABETO was born to support Covid-19 pandemic
- 9. 9 MAMMOGRAMS SPECIALIZATION Anomaly 1 Shape: irregular Margin: spiculated X-RAIS Deep Learning
- 10. 10 Partners involved Annotated by 25+ radiologist SOME NUMBERS 10 Server 50000 Iterations X-RAIS 20180 Images 19865 Anomalies 5350 Cases
- 12. ALFABETO
- 13. ALL FASTER BETTER TOGETHER X-RAIS SUPPORTING DIAGNOSIS DURING EPIDEMIOLOGICAL EMERGENCY X-RAIS
- 14. THE EXCELLENCE FOR A BETTER HEALTHCARE WHAT: ALFABETO is a platform based on AI which aims to reduce hospitalizations by supporting and speeding up diagnosis of Covid- 19.
- 15. DECISIONS ARE TAKEN AT HOME 1. Symptomatic patient asks for assistance 2. Medical visit at patient’s home (with the support of ALFABETO) 5. Reporting phase (with ALFABETO) 4. Severity assessment and progression estimate 3. Patient subjected to tests 6. Hospitalization vs Home stay
- 16. AI WORKFLOW: HOW ALFABETO WORKS Machine Learning Features extraction Vector CXR Clinical data + Machine Learning Classification Home Hospital Input Model
- 17. AI WORKFLOW: HOW ALFABETO WORKS X-RAIS Machine Learning Features extraction Vector CXR Clinical data + Machine Learning Classification Home Hospital Input Model Features extractor
- 18. AI WORKFLOW: HOW ALFABETO WORKS X-RAIS Machine Learning Features extraction Vector CXR Clinical data + Machine Learning Classification Home Hospital Input Model Features extractor Classifier
- 19. AI WORKFLOW: HOW ALFABETO WORKS X-RAIS Machine Learning Features extraction Vector CXR Clinical data + Machine Learning Classification Home Hospital Input Model Features extractor Classifier Doctor takes the last decision
- 20. AI WORKFLOW: HOW ALFABETO WORKS X-RAIS Machine Learning Features extraction Vector CXR Clinical data + Machine Learning Classification Home Hospital Input Model 1. Consolidation 2. Infiltration 3. Edema 4. Effusion 5. Pneumonia 6. Lung Opacity DenseNet Approved by a radiologist Classifier Features extractor
- 21. AI WORKFLOW: HOW ALFABETO WORKS X-RAIS Machine Learning Features extraction Vector CXR Clinical data + Machine Learning Classification Home Hospital Input Model 1. Cloud-based 2. Trained more than 50 models using AutoML 3. Cross validated • Ensemble model • 0,81 AUC Classifier
- 22. ADVANTAGES TA K I N G C A R E O F P AT I E N T S AT H O M E C L I N I C A L R I S K R E D U C T I O N M O R E P R E C I S E A N D F A S T E R T R E A T M E N T S T R A T E G Y D E F I N I T I O N R E D U C E D A C C E S S T O H O S P I TA L S I N F O R M AT I O N S H A R I N G A M O N G H E A LT H C A R E P R O F E S S I O N A L S A N D P AT I E N T S D Y N A M I C T R I A G E
- 23. CONCLUSIONS
- 24. An AI solution using Deep Learning to analyze medical images and support doctors in their everyday job X-RAIS 1 Inherently allows the inclusion of heterogeneous data to perform predictive analyses 2 Improves medical screening and reduce diagnosis time which means higher quality in medical treatments and reduction in clinical risk 3
- 25. LET'S AI SPEAK OUT • Artificial Intelligence is mature • Big Data & AI make Healthcare industry a better place • In Laife Reply, we develop innovative solutions based on AI and Big Data, e.g. X-RAIS and ALFABETO “disease costs more than health”
- 27. Feedback Your feedback is important to us. Don’t forget to rate and review the sessions.