AnemiaAssist combines XGBoost, CatBoost, RUSBoost, Logistic Regression, Gradient Boosting, Calibrated Linear SVM, LightGBM and AdaBoost as a transparent AI panel. Every model uses the same eight required CBC features only.
Used for WHO haemoglobin interpretation, report metadata and model-scope checks. These fields are not AI predictors.
Only these eight values are used by all eight AI models. Verify the values and units against the laboratory report.
Enter the eight required CBC values and select Analyze with 8 AI Doctors.
Each “doctor” below is a fictional AI persona mapped to one machine-learning model—not a real physician.
The vertical marker represents 50% Anemic probability. Bars show each AI model's current estimate.
Optional values do not alter the AI panel prediction.
Generate a structured A4 report with tables, AI Doctor opinions, probability bars, confidence, validation notes and clinical disclaimers.
CSV batch inference requires the same eight columns: WBC,RBC,HGB,HCT,MCV,MCH,MCHC,PLT. Other columns may be present and will be retained in exported results.
Choose a CSV file processed locally in this browser.
These names are intentionally human-readable personas for model outputs. They are AI models, not real doctors.
Different algorithms learn different decision boundaries. AnemiaAssist combines tree boosting, imbalance-aware boosting, probabilistic linear modeling and calibrated margin classification. The collaborative probability is a constrained weighted blend learned from cross-fitted development predictions, followed by probability calibration.
The panel is deliberately transparent: each model probability, opinion, confidence and ensemble weight is shown for every patient.
AnemiaAssist is a clinician-facing research decision-support system developed by Clariscan.AI. It is designed to assist doctors and other qualified healthcare professionals—not patients. The eight named AI Doctors are fictional AI model personas and are not real physicians.
Healthcare AI development focused on clinician-centered intelligent decision support, medical AI research and responsible translation of machine-learning methods into practical healthcare workflows.
Healthcare professionals are invited to share usability feedback, clinical concerns, model-error examples, validation opportunities and suggestions for improving AnemiaAssist.
The team brings together medical-AI research, machine learning, medical image analysis, academic research and healthcare technology commercialization.
Assistant Professor AI & Program Head, Department of Computer Science, Air University Karachi Campus. His work focuses on AI/ML, deep learning, medical imaging and clinical decision-support systems.
LinkedIn profile ↗AI startup strategist and healthcare-technology business specialist based in Riyadh, working on product strategy, commercialization, partnerships, growth and clinical-market translation.
LinkedIn profile ↗Assistant Professor at SINES, NUST. His expertise includes artificial intelligence, machine learning, deep learning, explainable AI, biomedical image processing and healthcare-focused data analysis.
LinkedIn profile ↗Assistant Professor at Hazara University with research interests in artificial intelligence, machine learning, medical image processing, computer vision and information retrieval, including AI research for anaemic RBC analysis.
LinkedIn profile ↗This system presents AI model outputs to support—not replace—clinical review. The names Dr. Xavier Green, Dr. Catherine Bose, Dr. Russell Boost, Dr. Logan Reeve, Dr. Grace Bostwick, Dr. Calvin Margin, Dr. Liam Greenfield and Dr. Adam Boober are fictional AI model personas, not real doctors.