One model is all you need: Multi-task learning enables simultaneous histology image segmentation and classification
Data extraction in tissue sample is a critical task in computational pathology as a first step for further analysis, e.g. disease diagnosis or prognosis. The task of data extraction could be multiple: segmentation and classification of nuclei, glands, lumina and different tissue regions from multi-center images of different organs.
Graham et al. [1], propose a versatile model to tackle the multi-task learning and achieves comparable performances to that of single-task models. This model should speed up the analysis of tissue samples and help reveal new biomarkers for better patient treatment.
[1] Graham, S., Vu, Q.D., Jahanifar, M., Raza, S.E.A., Minhas, F., Snead, D., Rajpoot, N.: One model is all you need: Multi-task learning enables simultaneous histology image segmentation and classification. Medical Image Analysis 83, 102685 (2023)
2) This project can be split into several subprojects, each for one student. For example, one student can work on object segmentation and classification and the other can work on tissue type classification
