NIH Research Festival
FARE Award Winner
Lymphoma is a type of malignant tumor that is often fatal to people of all ages. Positron Emission Tomography (PET) / Computed Tomography (CT) is the primary imaging method to assess lymphoma and monitor treatment response. As PET is sensitive to identify lymphoma regions while CT provides detailed anatomic information, the two imaging modalities can complement each other to enable improved diagnosis. However, automatic lymphoma segmentation is still challenging due to its substantial size and shape variability and limited datasets for training. To that end, we create a new whole-body lymphoma dataset. We design pre- and post- processing mechanisms particularly for lymphoma lesion segmentation according to medical knowledge priors. We designed an automatic lymphoma segmentation model with nnU-net as the backbone. Our pipeline incorporates pre- and post- processing mechanisms to remove regions of normally increased radiotracer uptake, and augment training with non-lymphoma samples. The proposed method was validated on the PET/CT scans of 46 patients. Experiments showed that by incorporating these additional steps, segmentation performance was further improved from 0.263 (LPMM-nsa) and 0.456 Dice (baseline nnU-Net) to 0.477 Dice (full pipeline). The results and analysis demonstrate the efficacy of the proposed method. Furthermore, the use of the AI model reduced the time required for segmentation by 55.7% compared to the reference manual segmentation. The mean Total Metabolic Tumor Volume (TMTV) calculated using the AI-predicted labels was not significantly different from those calculated using the manual labels (P = 0.153). This indicates that AI-based segmentation was comparable to manual segmentation.
Scientific Focus Area: Biomedical Engineering and Biophysics
This page was last updated on Monday, September 25, 2023