A living systematic review of prevention, diagnosis and treatments for every disease.
Mission
We believe every human life has equal value and is worth saving. In the past few years, it has become effortless to spread unreliable, biased, outdated and incomplete health information. Simultaneously, it has become hard to synthesize reliable, unbiased, up-to-date and complete health information. The most optimal treatment for nearly all medical conditions is unknown due to an incomplete overview of the evidence. Even less is known about disease prevention and healthy living. The incomplete overview of evidence undoubtedly had resulted in non-optimal outcomes for many patients. The answer is living systematic reviews and (network) meta-analysis of treatments for every disease. Our mission is to facilitate the restoration of trust in medicine by building the technical infrastructure to make rigorous and up-to-date evidence synthesis possible.
How the systematic review software works
Pitts is an online platform for conducting (living) systematic reviews and machine learning research for (semi)automated literature screening and data extraction.
We facilitate the software to perform your systematic review as well as collaborations to set up, test, validate and deploy your own machine learning algorithms on the platform.
Automated Search
When conducting a living systematic review, it is essential to check if new evidence is published continuously. With Automated Search, your search strategy is automatically executed in bibliographic databases using application programming interface integration. Currently we have Automated Search integrated for PubMed and are working on integrating more databases. When new studies that fit your search strategy become available, they will automatically appear in the screening workflow.
Screening
Real-time collaboration with reviewers is critical when conducting a systematic review. When starting the review, you can configure customized exclusion criteria. You can choose to do the independent blinded screening. It is easy to resolve conflicts when they arrive. A machine-learning algorithm can be used to classify randomized controlled trials based on the abstracts.
Data extraction
It is often difficult to set up a data extraction form, especially for network meta-analysis. You can easily pre-configure the data extraction form in the web application to fit your needs. While conducting a living systematic review, it sometimes becomes necessary to add data due to unforeseen circumstances. You can edit the data extraction form during the living systematic review. You can also choose if you want to do data extraction with two reviewers independently with blinding.
Machine Learning Validation
A well-validated machine learning approach for semi-automated literature screening or data extraction combined with a reliable online web application is essential to reduce workload without compromising scientific rigor. Our vision is to create a platform with all the tools available to independently conduct your machine learning research project and write a paper about it. We are developing a dashboard with performance metrics like precision (i.e., positive predictive value), recall (i.e., sensitivity), F1 score and AUC-ROC curves. Using this dashboard, users will have a complete overview of the performance of the models on their datasets. We also plan to build functionality for setting up experiments to calculate kappa scores, compare machine learning model performance with human reviews and hybrid approaches, and do random or binomial proportion tests.
Customers
Jacob-Jan
Chief Technology Officer
Studies Computer Science at Utrecht University, passionate about helping the medical world.
Cornelis
Strategic Advisor
Broad experience in healthcare where he combines science, policy & entrepreneurship with the ambition to achieve qualitative good, accessible and affordable healthcare.
Rutger
Software Developer
Studies Computer Science at Utrecht University, likes optimizing processes with machine learning.
Piet
Chief Executive Officer
Studies computer science at the Open Universiteit. He loves entrepreneurship with a greater cause.
Our Team
We are a small but dedicated team of software engineers working on building the future of evidence-based medicine.