Skip to main content

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Summers – Mathai – 2023

Mentor: Ronald M. Summers, M.D., Ph.D. | rms@nih.gov
Mentor: Tejas Mathai, Ph.D. | tejas.mathai@nih.gov
Lab
Imaging Biomarkers and Computer-Aided Diagnosis Laboratory

Project #1 – Identifying Bone Metastasis in CT scans

Identifying bone lesions in patients is vital as it enables the assessment of metastatic risk and the subsequent course of patient therapy and management. Bone lesions are a common secondary lesion type, and they can be categorized into benign or malignant types. Metastatic tumors need to identified and distinguished as they indicate distant metastasis, i.e., the cancer originated from a different location (e.g., prostate) and has spread to the osseous regions. Furthermore, tumor recurrence after therapy can also be a reason for the formation of invasive tumors of the bone. Once the bony lesions are identified, their biological activity (destruction pattern) and peri-osteal reaction patterns can be analyzed. Radiologists currently find it difficult to distinguish new and recurring lesions from existing lesions, and it can be difficult for junior radiologists without sufficient experience to reliably identify such lesions in radiographic images (e.g. CT, PET-CT). Therefore, early detection and associated longitudinal assessment of bone lesions is crucial for successful treatment. The candidate selected for this project will aid the lab’s research thrusts in locating and correctly distinguishing the metastatic nature of bone lesions. He/she/they will be learn valuable skills regarding the human physiology, the presentation of bone lesions in CT studies, interact with practicing radiologist(s) to learn from their skills, and help with the development of computational tools rooted in artificial intelligence techniques to improve diagnosis and patient outcomes. 

Project #2 -  Quantification of Renal Structural Findings on CT/MRI

Morphological changes of the kidneys are important biomarkers for the assessment of renal diseases. In imaging studies (e.g. CT, MRI) of the kidneys in patients, identification of the kidney volumes play crucial roles in following the progression and treatment of polycystic kidney disease, evaluating candidate kidney donor suitability for renal transplants, and chronic kidney disease (CKD). Particularly, it is important to quantify the volumes of the renal cortices and medullae; for example, a lower volume of the cortex in relation to the medulla is indicative of poor glomerular function, CKD, aging, and potential nephrosclerosis. In patients receiving renal transplants, a smaller medullary volume was also indicative of a higher graft failure risk. Since it is cumbersome and time-consuming for radiologists to manually compute the volumetric measures during a busy clinical day, these measures are generally not available during donor evaluations. There is also an open question regarding the correlation between the kidney volumetric measurements and patient blood work. The incumbent candidate for this project will aid the lab’s research focus towards characterizing structural findings of the kidney on CT/MRI. He/she/they will be learn valuable skills regarding the human physiology, the presentation of kidney diseases in patient imaging (CT/MRI) studies, interact with practicing radiologist(s) to learn from their skills, and help with the development of computational tools rooted in artificial intelligence techniques to improve diagnosis and patient outcomes.