#Diagnostic tools for digital pathology
#Prognostic and predictive tools for oncology
#Quality control in digital pathology
#Algorithms for federated learning
#Clinical validation studies for digital pathology tools
November 16, 2024
Modelling prostate cancer growth and evolution
In our study published in Cell Systems, we present a comprehensive three-dimensional, realistic mathematical model of prostate cancer that integrates processes such as tumor growth, genetic evolution, and clonal competition. This model provides valuable insights into evolutionary trajectories and clonal hierarchies, offering potential clinical implications.
September 1, 2024
Computational platform for lung cancer
In our milestone study in Cell Reports Medicine, we develop a foundational computational pathology platform for non-small cell lung cancer with full explainability. We show how the platform could be effectively used for development of new diagnostic and prognostic algorithms. We suggest several new effective, fully explainable prognostic parameters based on joint evaluation of tertiary lymphoid structures (TLS) and tumor necrosis. These allow for effective patient stratification according to the risk of negative outcomes.
August 15, 2024
AI algorithm for virtual biopsies and optimization of targeted prostate biopsyIn our new study, we show how the highly accurate, precise diagnostic AI algorithm for prostate cancer detection and Gleason grading developed using a fast-track approach can be used to answer the important clinical question - what is the best setup for targeted prostate biopsy?
The results are relevant to both urologists and pathologists and can inform how MRI-targeted biopsies are performed in daily practice.
April 16, 2024
Detection of lymph node metastasis in colorectal cancerOur new paper to fast-track development of a clinical-grade tool for detection of lymph node metastasis was published in Modern Pathology.
In this study, we validate our tool on the histological slides from five pathology departments showing nearly perfect sensitivity and very high specificity for this diagnostic task.
September 29, 2023
Segmentation tool for colorectal specimens / Validation studyOur new paper to development and extensive clinical validation of the AI tool for segmentation / processing of colorectal specimens was published in Modern Pathology. Our algorithm allows for precise, fully quantitative analysis of different tissue classes in colorectal specimens.
We validate the algorithm on one particular clinical, diagnostic task of the colorectal biopsy pre-screening for invasive carcinoma using large multi-institutional case cohort scanned by three different scanning systems.
July 27, 2023
Validation study / Prostate cancer detection and grading toolOur new paper to one of the largest validation studies for AI tools in digital pathology published to date (for prostate cancer detection and Gleason grading) was accepted in the NPJ Precision Oncology.
Five pathology institutes and 11 pathologists from all over the world participated in this study.
We also release valuable material (slides of prostate biopsies from two departments with grading ground truth by 11 pathologists) as a public dataset.
March 10, 2023
Oesophageal cancer tumor detection and regression grading toolOur new paper about tumor detection and regression grading in oesophageal adenocarcinoma was published in the Lancet Digital Health journal.
We show how computational tool can substantially improve tumor detection and histological regression grading in resection specimens from patients with oesophageal adenocarcinoma and neoadjuvant radiochemotherapy.
February 4, 2023
SemiCOL Challenge for colorectal cancerWe organise SemiCOL computational Challenge together with European Society of Digital and Integrative Pathology (ESDIP) and a large number of partners from clinical and technical departments.
© tolklab.de 2023