IgCONDA-PET: Weakly-Supervised PET Anomaly Detection using Implicitly-Guided Attention-Conditional Counterfactual Diffusion Modeling -- a Multi-Center, Multi-Cancer, and Multi-Tracer Study
Shadab Ahamed, Arman Rahmim
eess.IV, cs.CV, cs.LG
Submitted: 2025-06-24
Updated: 2026-08-19
Comments: 48 pages, 13 figures, 4 tables
Journal ref: Computerized Medical Imaging and Graphics, 124, 102615 (2025)
DOI: 10.1016/j.compmedimag.2025.102615
Code: https://github.com/ahxmeds/IgCONDA-PET
License: http://creativecommons.org/licenses/by/4.0/
The gist: Minimizing the need for pixel-level annotated data to train PET lesion detection and segmentation networks is highly desired and can be transformative, given time and cost constraints associated with
Terminology
Abstract
Minimizing the need for pixel-level annotated data to train PET lesion detection and segmentation networks is highly desired and can be transformative, given time and cost constraints associated with expert annotations. Current unsupervised or weakly-supervised anomaly detection methods rely on autoencoder or generative adversarial networks (GANs) trained only on healthy data. While these approaches reduce annotation dependency, GAN-based methods are notably more challenging to train than non-GAN alternatives (such as autoencoders) due to issues such as the simultaneous optimization of two competing networks, mode collapse, and training instability. In this paper, we present the weakly-supervised I mplicitly- g uided CO u N terfactual diffusion model for D etecting A nomalies in PET images (IgCONDA-PET). The solution is developed and validated using PET scans from six retrospective cohorts consisting of a total of 2652 cases (multi-cancer, multi-tracer) containing both local and public datasets (spanning multiple centers). The training is conditioned on image class labels (healthy vs. unhealthy) via attention modules, and we employ implicit diffusion guidance. We perform counterfactual generation which facilitates "unhealthy-to-healthy" domain translation by generating a synthetic, healthy version of an unhealthy input image, enabling the detection of anomalies through the calculated differences. The performance of our method was compared against several other deep learning based weakly-supervised or unsupervised methods as well as traditional methods like 41% SUV max thresholding. We also highlight the importance of incorporating attention modules in our network for the detection of small anomalies. The code is publicly available at: https://github.com/ahxmeds/IgCONDA-PET.git.
Sources
- Comprehensive framework for evaluation of deep neural networks in detection and quantification of lymphoma from PET/CT images: clinical insights, pitfalls, and observer agreement analyses
- Weakly Supervised Anomaly Detection for Chest X-Ray Image
- Comprehensive Evaluation of Quantitative Measurements from Automated Deep Segmentations of PSMA PET/CT Images
- Adaptive Voxel-Weighted Loss Using L1 Norms in Deep Neural Networks for Detection and Segmentation of Prostate Cancer Lesions in PET/CT Images
- Anomaly detection using Diffusion-based methods
- Evaluation of pseudo-healthy image reconstruction for anomaly detection with deep generative models: Application to brain FDG PET
- DDPET-3D: Dose-aware Diffusion Model for 3D Ultra Low-dose PET Imaging
- Joint Diffusion: Mutual Consistency-Driven Diffusion Model for PET-MRI Co-Reconstruction
- Head and Neck Tumor Segmentation from [18F]F-FDG PET/CT Images Based on 3D Diffusion Model
- Hierarchical Text-Conditional Image Generation with CLIP Latents
- Diffusion Causal Models for Counterfactual Estimation
- Tutorial on Diffusion Models for Imaging and Vision
- Denoising Diffusion Implicit Models
- GLU Variants Improve Transformer
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Bridging the Gap: Addressing Discrepancies in Diffusion Model Training for Classifier-Free Guidance
- Classifier-Free Diffusion Guidance
- Generative AI for Medical Imaging: extending the MONAI Framework
- Use HiResCAM instead of Grad-CAM for faithful explanations of convolutional neural networks
- SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis
Related papers
- GBT-SAM: A Parameter-Efficient Depth-Aware Model for Generalizable Brain tumour Segmentation on mp-MRI
- PanoDiff-SR: Synthesizing Dental Panoramic Radiographs using Diffusion and Super-resolution
- Nonparametric Distribution Matching for Self-Supervised Whole-Slide Image Condensation
- Revisiting Integration of Image and Metadata for DICOM Series Classification: Cross-Attention and Dictionary Learning
- VesselSDF: Distance Field Priors for Vascular Network Reconstruction
- cSVR: Convolutional Slice-to-Volume Reconstruction