Learning biologically relevant features in a pathology foundation model using sparse autoencoders
- URL: http://arxiv.org/abs/2407.10785v3
- Date: Mon, 16 Dec 2024 21:02:15 GMT
- Title: Learning biologically relevant features in a pathology foundation model using sparse autoencoders
- Authors: Nhat Minh Le, Ciyue Shen, Neel Patel, Chintan Shah, Darpan Sanghavi, Blake Martin, Alfred Eng, Daniel Shenker, Harshith Padigela, Raymond Biju, Syed Ashar Javed, Jennifer Hipp, John Abel, Harsha Pokkalla, Sean Grullon, Dinkar Juyal,
- Abstract summary: We trained a Sparse Autoencoder on the embeddings of a pathology pretrained foundation model.
We found that Sparse Autoencoder features represent interpretable and monosemantic biological concepts.
- Score: 2.5919097694815365
- License:
- Abstract: Pathology plays an important role in disease diagnosis, treatment decision-making and drug development. Previous works on interpretability for machine learning models on pathology images have revolved around methods such as attention value visualization and deriving human-interpretable features from model heatmaps. Mechanistic interpretability is an emerging area of model interpretability that focuses on reverse-engineering neural networks. Sparse Autoencoders (SAEs) have emerged as a promising direction in terms of extracting monosemantic features from polysemantic model activations. In this work, we trained a Sparse Autoencoder on the embeddings of a pathology pretrained foundation model. We found that Sparse Autoencoder features represent interpretable and monosemantic biological concepts. In particular, individual SAE dimensions showed strong correlations with cell type counts such as plasma cells and lymphocytes. These biological representations were unique to the pathology pretrained model and were not found in a self-supervised model pretrained on natural images. We demonstrated that such biologically-grounded monosemantic representations evolved across the model's depth, and the pathology foundation model eventually gained robustness to non-biological factors such as scanner type. The emergence of biologically relevant SAE features was generalizable to an out-of-domain dataset. Our work paves the way for further exploration around interpretable feature dimensions and their utility for medical and clinical applications.
Related papers
- Causal Representation Learning from Multimodal Biological Observations [57.00712157758845]
We aim to develop flexible identification conditions for multimodal data.
We establish identifiability guarantees for each latent component, extending the subspace identification results from prior work.
Our key theoretical ingredient is the structural sparsity of the causal connections among distinct modalities.
arXiv Detail & Related papers (2024-11-10T16:40:27Z) - Deep Latent Variable Modeling of Physiological Signals [0.8702432681310401]
We explore high-dimensional problems related to physiological monitoring using latent variable models.
First, we present a novel deep state-space model to generate electrical waveforms of the heart using optically obtained signals as inputs.
Second, we present a brain signal modeling scheme that combines the strengths of probabilistic graphical models and deep adversarial learning.
Third, we propose a framework for the joint modeling of physiological measures and behavior.
arXiv Detail & Related papers (2024-05-29T17:07:33Z) - Tertiary Lymphoid Structures Generation through Graph-based Diffusion [54.37503714313661]
In this work, we leverage state-of-the-art graph-based diffusion models to generate biologically meaningful cell-graphs.
We show that the adopted graph diffusion model is able to accurately learn the distribution of cells in terms of their tertiary lymphoid structures (TLS) content.
arXiv Detail & Related papers (2023-10-10T14:37:17Z) - Biologically-informed deep learning models for cancer: fundamental
trends for encoding and interpreting oncology data [0.0]
We provide a structured literature analysis focused on Deep Learning (DL) models used to support inference in cancer biology.
The work focuses on how existing models address the need for better dialogue with prior knowledge, biological plausibility and interpretability.
arXiv Detail & Related papers (2022-07-02T12:11:35Z) - Learning multi-scale functional representations of proteins from
single-cell microscopy data [77.34726150561087]
We show that simple convolutional networks trained on localization classification can learn protein representations that encapsulate diverse functional information.
We also propose a robust evaluation strategy to assess quality of protein representations across different scales of biological function.
arXiv Detail & Related papers (2022-05-24T00:00:07Z) - PathologyBERT -- Pre-trained Vs. A New Transformer Language Model for
Pathology Domain [2.3628956573813498]
Successful text mining of a large pathology database can play a critical role to advance 'big data' cancer research.
No pathology-specific language space exist to support the rapid data-mining development in pathology space.
PathologyBERT is a pre-trained masked language model which was trained on 347,173 histopathology specimen reports.
arXiv Detail & Related papers (2022-05-13T20:42:07Z) - Mapping the landscape of histomorphological cancer phenotypes using
self-supervised learning on unlabeled, unannotated pathology slides [9.27127895781971]
Histomorphological Phenotype Learning operates via the automatic discovery of discriminatory image features in small image tiles.
Tiles are grouped into morphologically similar clusters which constitute a library of histomorphological phenotypes.
arXiv Detail & Related papers (2022-05-04T08:06:55Z) - Analyzing the Effects of Handling Data Imbalance on Learned Features
from Medical Images by Looking Into the Models [50.537859423741644]
Training a model on an imbalanced dataset can introduce unique challenges to the learning problem.
We look deeper into the internal units of neural networks to observe how handling data imbalance affects the learned features.
arXiv Detail & Related papers (2022-04-04T09:38:38Z) - Self-Supervised Vision Transformers Learn Visual Concepts in
Histopathology [5.164102666113966]
We conduct a search for good representations in pathology by training a variety of self-supervised models with validation on a variety of weakly-supervised and patch-level tasks.
Our key finding is in discovering that Vision Transformers using DINO-based knowledge distillation are able to learn data-efficient and interpretable features in histology images.
arXiv Detail & Related papers (2022-03-01T16:14:41Z) - A multi-stage machine learning model on diagnosis of esophageal
manometry [50.591267188664666]
The framework includes deep-learning models at the swallow-level stage and feature-based machine learning models at the study-level stage.
This is the first artificial-intelligence-style model to automatically predict CC diagnosis of HRM study from raw multi-swallow data.
arXiv Detail & Related papers (2021-06-25T20:09:23Z) - G-MIND: An End-to-End Multimodal Imaging-Genetics Framework for
Biomarker Identification and Disease Classification [49.53651166356737]
We propose a novel deep neural network architecture to integrate imaging and genetics data, as guided by diagnosis, that provides interpretable biomarkers.
We have evaluated our model on a population study of schizophrenia that includes two functional MRI (fMRI) paradigms and Single Nucleotide Polymorphism (SNP) data.
arXiv Detail & Related papers (2021-01-27T19:28:04Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.