Advanced Medical Technologies 
Research Group
 
,AMT

 

 

 


Advanced Medical Technologies Research Group (AMT-Research Group)

 

Welcome to the Advanced Medical Technologies Research Group (AMT-RG) page!

 

The AMT- Research Group, headed by Dr. Sinan Onal, is an interdisciplinary research group in areas of medical image analysis and new product development with emphasis in medical devices. Our research focuses on developing new diagnosis and prediction models for quantitative analysis of illnesses using medical images. We use machine learning techniques for extracting potentially unpredictable patterns on medical images. We also focus on developing medical devices and apparatus to facilitate the surgical operations, medical treatment and medical training. In collaboration with medical doctors and surgeons from different medical schools, we observe surgical operations to understand the problems, create alternative design concepts using CAD systems, and select the best design based on our user needs. We facilitate physical prototypes using our rapid prototyping machines. Current collaborators of the AMT Research Group are from Washington University, College of Medicine- Department of Ophthalmology and Southern Illinois University-Edwardsville, School of Dental Medicine.

 

CURRENT PROJECTS

Image-based Ordinal Regression Models for Imbalanced Datasets to Predict Risk of Development for Pelvic Organ Prolapse

This project aims to predict the risk of development of multiple stages of pelvic organ prolapse through ordinal regression models that combine image and clinical data from highly imbalanced datasets. Pelvic organ prolapse (POP) is a major health problem that affects between 30-50% of women in the US. Although clinical examination is currently used to diagnose POP, there is still little evidence on specific risk factors that are directly related to particular types of POP and their severity or stages (Stage 0-IV). Data from dynamic MRI has the potential to improve POP prediction but it is currently extracted manually limiting its use to small datasets. Given the plethora of potential risk factors for POP, it is very likely that this condition is caused by a combination of risk factors that are patient-specific. Moreover, major challenges in the prediction of POP are that the stages are ordinal and their distribution is highly imbalanced. The hypothesis is that new ordinal regression models that combine automatically extracted MRI-based data and clinical data from large datasets can lead to the identification of women at risk of developing certain stages of POP to personalize the intervention strategy and help prevent further exacerbation of this condition.


 

Comparative Analysis of Peri-papillary Outer Retinal Layer for Diagnosis of Glaucoma

Glaucoma is a critical health condition affecting up to 44.7 million people world-wide and 2.8 million people in the U.S. resulting $2.5 billion cost. Glaucoma is a group of eye condition that results optical nerve damage which is a leading cause of blindness. It is currently diagnosed through clinical examination. Recently, imaging technologies have been used to analyze retina to provide structural measurement for diagnosis. However, these measurements have been found inadequate for timely diagnosis. Therefore, there is a major need for early diagnosis of glaucoma using new parameters. It is hypothesized that new topological parameters on OCT can be used to differentiate the healthy and glaucomatous eyes.



Automated Segmentation of Blood Vessels on Fundus Images for Diagnosis of Diabetic Retinopathy

It is approximated more than 10 million diabetic people have in the US. Diabetes is serious health condition in the world and it results serious diabetic eye disease.  Diabetic eye disease may include diabetic retinopathy, cataract and glaucoma. Diabetic retinopathy is the most common diabetic eye disease. Approximately 40% of diabetic patients have at least mild sign of diabetic retinopathy, and 3% have visual loss because of the diabetic condition. Diabetic retinopathy refers eye condition with damaged blood vessels in the retina. The disease signs can be seen in different conditions. In some cases, blood vessels may swell and leak fluid or abnormal blood vessels grow on the surface of the retina. These conditions can cause vision loss with different ways. Abnormal blood vessels can provide extra blood and cause leakage in to the center of the eye, or blood can leak into the center of the macula which gets swell. Diabetic retinopathy is diagnosed through comprehensive eye exams such as visual acuity test, dilated eye exam and tonometry. Retinal photography has become popular to diagnose the diabetic retinopathy. Fundus camera is used to provide detail of retinal structures especially blood vessels as seen in Figure 1. During the exams, the ophthalmologist checks blood vessels conditions. Therefore, blood vessels are detected in the fundus image for diagnosis of diabetic retinopathy. However, detection of blood vessels in the fundus image is time consuming and inconsistent process. We developed a new segmentation scheme to extract the vessels on fundus images automatically.

 


Hand Support System for Eye Surgeons during Cataract Surgery

Cataract surgery is one of the most common surgeries in the US. It is a procedure to remove the lens of eye and, in most cases, replace it with an artificial lens. Cataract surgery is used to treat the clouding of the normally clear lens of eye (cataract). Cataract surgery is performed by an ophthalmologist on an outpatient basis, which means patients don't have to stay in the hospital after the surgery. Although the cataract surgery is very common and is generally a safe procedure, there is a need for hand support system for ophthalmologists during the surgery. They normally perform 6-9 cataract surgeries in a day, so their fatigue increases the risk of the operation. In this research, a new and ergonomic hand support system will be designed and fabricated to facilitate the cataract surgery.