Publications

2025

1. arXiv

   NeuroLDM-3D: Enhancing Neurological Disease Detection by Leveraging Conditional Latent Diffusion for Brain MRI Synthesis

   Jay Sawant, Erik Kaestner, Ezequiel Gleichgerrcht, Leonardo Bonilha, Carrie R. McDonald, and Kyle Hasenstab

   2025

   Abs

   We introduce NeuroLDM-3D, a class-conditional latent diffusion framework for synthesizing realistic three-dimensional brain MRIs to improve model training for automated detection of neurological disease. The approach first trains a 3D variational autoencoder (VAE) to compress volumes into a smooth latent space that preserves neuroanatomy and pathological signatures, then learns a transformer-based denoiser (DiT-3D) to generate class-specific latents (Healthy vs. patient), which are decoded into full-resolution scans. Using a multi-site T1-weighted dataset of healthy controls (HC) and patients with temporal lobe epilepsy (TLE), NeuroLDM-3D produces anatomically coherent and class-consistent images that retain disease-relevant pathological cues. Compared with adversarial and voxel-space diffusion baselines, the proposed framework achieves higher generative fidelity, reflecting the benefits of latent-space modeling and transformer-based global context. When augmenting training sets, synthetic volumes improve downstream TLE classification performance in limited-data regimes and maintain performance when real data are abundant. Attribution analyses further show that models trained with only synthetic data identify the same medial-temporal and limbic structures associated with TLE, supporting the neurobiological plausibility of the generated images. Overall, these results demonstrate that targeted, class-aware 3D MRI synthesis using latent diffusion can effectively mitigate data scarcity, enhance diagnostic robustness, and enable scalable, anatomically grounded generative modeling for clinical neuroimaging applications.

2024

1. arXiv

   Semantic Segmentation Based Quality Control of Histopathology Whole Slide Images

   Abhijeet Patil, Garima Jain, Harsh Diwakar, Jay Sawant, Tripti Bameta, Swapnil Rane, and Amit Sethi

   2024

   Abs Bib

   We developed a software pipeline for quality control (QC) of histopathology whole slide images (WSIs) that segments various regions, such as blurs of different levels, tissue regions, tissue folds, and pen marks. Given the necessity and increasing availability of GPUs for processing WSIs, the proposed pipeline comprises multiple lightweight deep learning models to strike a balance between accuracy and speed. The pipeline was evaluated in all TCGAs, which is the largest publicly available WSI dataset containing more than 11,000 histopathological images from 28 organs. It was compared to a previous work, which was not based on deep learning, and it showed consistent improvement in segmentation results across organs. To minimize annotation effort for tissue and blur segmentation, annotated images were automatically prepared by mosaicking patches (sub-images) from various WSIs whose labels were identified using a patch classification tool HistoROI. Due to the generality of our trained QC pipeline and its extensive testing the potential impact of this work is broad. It can be used for automated pre-processing any WSI cohort to enhance the accuracy and reliability of large-scale histopathology image analysis for both research and clinical use. We have made the trained models, training scripts, training data, and inference results publicly available at this https URL, which should enable the research community to use the pipeline right out of the box or further customize it to new datasets and applications in the future.

   @misc{patil2024semanticsegmentationbasedquality,
     title = {Semantic Segmentation Based Quality Control of Histopathology Whole Slide Images},
     author = {Patil, Abhijeet and Jain, Garima and Diwakar, Harsh and Sawant, Jay and Bameta, Tripti and Rane, Swapnil and Sethi, Amit},
     year = {2024},
     eprint = {2410.03289},
     archiveprefix = {arXiv},
     primaryclass = {eess.IV}
   }

2023

1. JPI

   Efficient quality control of whole slide pathology images with human-in-the-loop training

   Abhijeet Patil, Harsh Diwakar, Jay Sawant, Nikhil Cherian Kurian, Subhash Yadav, Swapnil Rane, Tripti Bameta, and Amit Sethi

   Journal of Pathology Informatics 2023

   Abs Bib

   Histopathology whole slide images (WSIs) are being widely used to develop deep learning-based diagnostic solutions, especially for precision oncology. Most of these diagnostic softwares are vulnerable to biases and impurities in the training and test data which can lead to inaccurate diagnoses. For instance, WSIs contain multiple types of tissue regions, at least some of which might not be relevant to the diagnosis. We introduce HistoROI, a robust yet lightweight deep learning-based classifier to segregate WSI into 6 broad tissue regions—epithelium, stroma, lymphocytes, adipose, artifacts, and miscellaneous. HistoROI is trained using a novel human in-the-loop and active learning paradigm that ensures variations in training data for labeling efficient generalization. HistoROI consistently performs well across multiple organs, despite being trained on only a single dataset, demonstrating strong generalization. Further, we have examined the utility of HistoROI in improving the performance of downstream deep learning-based tasks using the CAMELYON breast cancer lymph node and TCGA lung cancer datasets. For the former dataset, the area under the receiver operating characteristic curve (AUC) for metastasis versus normal tissue of a neural network trained using weakly supervised learning increased from 0.88 to 0.92 by filtering the data using HistoROI. Similarly, the AUC increased from 0.88 to 0.93 for the classification between adenocarcinoma and squamous cell carcinoma on the lung cancer dataset. We also found that the performance of the HistoROI improves upon HistoQC for artifact detection on a test dataset of 93 annotated WSIs. The limitations of the proposed model are analyzed, and potential extensions are also discussed.

   @article{PATIL2023100306,
     title = {Efficient quality control of whole slide pathology images with human-in-the-loop training},
     journal = {Journal of Pathology Informatics},
     volume = {14},
     pages = {100306},
     year = {2023},
     issn = {2153-3539},
     url = {https://doi.org/10.1016/j.jpi.2023.100306},
     doi = {10.1016/j.jpi.2023.100306},
     author = {Patil, Abhijeet and Diwakar, Harsh and Sawant, Jay and Kurian, Nikhil Cherian and Yadav, Subhash and Rane, Swapnil and Bameta, Tripti and Sethi, Amit},
     keywords = {WSI pre-processing, Weakly supervised learning, Quality control, Generalization}
   }