Browsing by Author "Zhang, Haoran"

Purpose: Accurately quantifying the extent vessel coverage in the pancreas is essential for determining the feasibility of surgery. The aim of this study is to train a segmentation model specialized for Pancreatic Ductal Adenocarcinoma (PDAC) imaging, focusing on delineating pancreas, tumor, arteries (celiac artery, superior mesenteric artery, common hepatic artery), and veins (portal vein, superior mesenteric vein) using an improved Attention Unet CNN approach.Methods: Data from 100 PDAC patients treated at the Ruijin Hospital between 2020 and 2022 were utilized. Using Synapse 3D software, masks of the tumor, arteries (including the celiac artery, superior mesenteric artery, and hepatic artery), and veins (including the superior mesenteric vein and portal vein) were generated semi-automatically and reviewed by radiologists. Standard image processing techniques, including adjustment of window level to 60 and width to 350 and histogram equalization were subsequently applied. Two types of CNN-based Attention Unet segmentation models were developed: (1) Unified Unet Model that segments all four components simultaneously, and (2) four Individual Unet Models that segments pancreas, tumor, veins, and arteries separately. The train-validation-test data assignment was set to 7:2:1. The segmentation efficacy was assessed using Dice similarity coefficient, with the Adam optimizer utilized for optimization. Results: The individual segmentation models achieve notable performance: pancreas (Accuracy: 0.84, IoU: 0.81, Dice: 0.76), tumor (Accuracy: 0.78, IoU: 0.77, Dice: 0.68), vein (Accuracy: 0.88, IoU: 0.86, Dice: 0.80), and artery (Accuracy: 0.91, IoU: 0.93, Dice: 0.93). However, the unified model demonstrates inferior performance with accuracy, IoU, and Dice coefficient scores of 0.61, 0.50, and 0.45, respectively. Conclusion: Accurate segmentation models have been developed for pancreas, pancreatic tumors, arteries, and veins in PDAC patients. This will enable the efficient quantification of vessel coverage in the pancreas, thereby enhancing the decision-making process regarding the feasibility of surgery for PDAC patients. The findings also demonstrate that Individual Models outperform the Unified Model in segmentation accuracy, highlighting the importance of tailored segmentation strategies for different anatomical structures in PDAC imaging.

Esophageal cancer is a very aggressive form of cancer, and in the past decade, the incidence rate of esophageal cancer is rising faster than any other malignancy in the U.S. Luckily, most precancers are preventable given timely surveillance and proper treatments. Despite the recent success of current screening methodologies, these techniques are still costly and limited. As an alternative, angle-resolved low-coherence interferometry (a/LCI) is an optical technique which enables depth-resolved measurements of nuclear morphology, a biomarker for precancer. In this dissertation, several advances in a/LCI technology were presented. First, computational analysis of a previous clinical a/LCI dataset was used to provide design guidance for future a/LCI designs. The impact of reductions in angular range and angular sampling frequency on the diagnostic performance of a/LCI was analyzed and discussed. Next, in order to improve the clinical utility of a/LCI, a novel processing algorithm based on deep learning was presented for identifying dysplasia from depth-resolved angular scattering scans collected by a/LCI with high accuracy and speed. Future development of this algorithm may open to possibilities for real-time clinical analysis of a/LCI data, and improve the clinical utility of the instrument during in vivo clinical trials for real-time screening of the tissue. In addition, instrumentational advances in a/LCI were also demonstrated. Development of the opto-mechanical instrumentation using a single multimode fiber was presented to overcome the limiting factor of using fiber bundles for a/LCI imaging, as these fiber bundles are fragile, expensive, and exhibits low optical throughput. The technique was validated using microsphere phantoms, and showed excellent agreement with the actual size. This technique was also insensible to the displacement of the fiber, and showed great potential for future endoscopic applications for medical diagnostics. Finally, a combined a/LCI and OCT imaging platform was developed and adapted for esophageal imaging, and a clinical study was performed to determine the effectiveness of using this combined instrumentation for screening dysplasia in patients with Barrett’s esophagus, a biomarker for dysplasia. Optical biopsies were taken from 50 distinct tissue biopsy sites and compared to histopathological analysis of co-registered tissue biopsies. Analysis of the a/LCI scans demonstrated perfect sensitivity (100%) for detection of esophageal dysplasia, and the increase in specificity (from 84% to 93%) compared with a previous clinical study demonstrated the ability of OCT in targeting potential diseased biopsies, suggesting that optical biopsy characterization using a/LCI nuclear morphology measurements with real-time OCT imaging guidance would aid the clinician in identifying dysplastic tissue sites in vivo, leading to improved screening protocols, and ultimately, better patient outcomes.