Mixing Histopathology Prototypes into Robust Slide-Level Representations for Cancer Subtyping
This work addresses computational bottlenecks in computational pathology for cancer subtyping, offering a more efficient alternative to existing methods, though it is incremental as it builds on prior techniques like MLP-Mixers and prototype-based approaches.
The paper tackled the computational expense of processing whole-slide images for cancer subtyping by proposing a method that uses feature embedding and clustering to create prototype representations, which are then input into an MLP-Mixer architecture, achieving comparable performance to state-of-the-art methods with lower training costs in terms of time and memory.
Whole-slide image analysis via the means of computational pathology often relies on processing tessellated gigapixel images with only slide-level labels available. Applying multiple instance learning-based methods or transformer models is computationally expensive as, for each image, all instances have to be processed simultaneously. The MLP-Mixer is an under-explored alternative model to common vision transformers, especially for large-scale datasets. Due to the lack of a self-attention mechanism, they have linear computational complexity to the number of input patches but achieve comparable performance on natural image datasets. We propose a combination of feature embedding and clustering to preprocess the full whole-slide image into a reduced prototype representation which can then serve as input to a suitable MLP-Mixer architecture. Our experiments on two public benchmarks and one inhouse malignant lymphoma dataset show comparable performance to current state-of-the-art methods, while achieving lower training costs in terms of computational time and memory load. Code is publicly available at https://github.com/butkej/ProtoMixer.