AI Helping Predict Specific Mutations
At the University of Basel pathology department they are currently working on an AI algorithm to predict specific mutation based on Pap-stained virtual slides from lung cancer specimens. Once established, it would be necessary to validate this algorithm on Pap-stained NSCLC specimens (with known mutations) from selected other institutions (e.g. at least 20 such specimens per institution). This validation project is too early right know, but might become realistic sometimes early next year, depending on the performance of our AI algorithm.
Standardizing Immunocytochemistry
Currently, an ongoing study is in progress on standardization of immunocytochemistry protocols for ethanol-fixed Pap-stained lung cancer specimens. Once the optimal protocols are defined, they can be made them accessible to the community as a starting point to establish or improve ICC in their own laboratory. Laboratories that can engage in this project should get in touch. The goal would be to explore if the use of template protocols could be used to improve and increase the use of ICC in cytopathology labs. Again, this is also a bit to early, since our standardization study is still ongoing.
Reproducing a HGUC Diagnosis Using The Paris System?
Diagnosing HGUC is the main purpose of The Paris System, yet there might a certain interobserver variability in this diagnosis, specially, within the grey-zone of G2- tumors. The International Society of Urological Pathology is currently discussing a separation of G2 urothelial carcinoma into G2-low grade and G2-high grade. Such a development requires input from cytology. This could provide virtual slides (or still images) from patients with histological expert census diagnoses of G2-LG and G2-HG cases requiring cytology experts to review such cases after a virtual interactive instruction. Similar projects could be possible across other areas of cytology (e.g. cytological diagnosis of mesothelioma).
We want to hear from you!
