Sooyeul Lee

Glioma grading using apparent diffusion coefficient map: application of histogram analysis based on automatic segmentation

The accurate diagnosis of glioma subtypes is critical for appropriate treatment, but conventional histopathologic diagnosis often exhibits significant intra‐observer variability and sampling error. The aim of this study was to investigate whether histogram analysis using an automatically segmented region of interest (ROI), excluding cystic or necrotic portions, could improve the differentiation between low‐grade and high‐grade gliomas. Thirty‐two patients (nine low‐grade and 23 high‐grade gliomas) were included in this retrospective investigation. The outer boundaries of the entire tumors were manually drawn in each section of the contrast‐enhanced T1‐weighted MR images. We excluded cystic or necrotic portions from the entire tumor volume. The histogram analyses were performed within the ROI on normalized apparent diffusion coefficient (ADC) maps. To evaluate the contribution of the proposed method to glioma grading, we compared the area under the receiver operating characteristic (ROC) curves. We found that an ROI excluding cystic or necrotic portions was more useful for glioma grading than was an entire tumor ROI. In the case of the fifth percentile values of the normalized ADC histogram, the area under the ROC curve for the tumor ROIs excluding cystic or necrotic portions was significantly higher than that for the entire tumor ROIs (p < 0.005). The automatic segmentation of a cystic or necrotic area probably improves the ability to differentiate between high‐ and low‐grade gliomas on an ADC map. Copyright

The echotextural characteristics for the diagnosis of the liver cirrhosis using the sonographic images

Various statistical parameters have been tried for the computer-aided diagnosis of the liver fibrosis. The region of interest (ROI) for the liver and spleen parenchymas have been chosen, and the hepatolienal textural contrast for each ultrasound (US) image has been examined. The selectively chosen textural parameters are linearly combined with the pre-determined coefficients to give the computer-aided diagnostic parameter for the liver fibrosis, whose final stage is named as cirrhosis. From the comparison with the clinical diagnosis it is suggested that the proposed calculation scheme using the textural parameters show the quite promising classification performance for the computer-aided diagnosis of the liver cirrhosis.

Computer-assisted Sonographic Analysis of the Hepatorenal and Textural Features for the Diagnosis of the Fatty Liver

We investigated the B-mode ultrasound (US) images of abdomen by normalizing the gray level values of the liver with the representative gray levels (RGLs) of the liver parenchyma, renal cortex, subcutaneous fat, to quantify the fatty infiltration of the liver (FIL). The RGLs were compared with the clinical diagnosis of the FIL showing good correlation between them

PHASE DIAGRAM OF THE RESTRICTED SOLID-ON-SOLID MODEL COUPLED TO THE ISING MODEL

We study the phase transitions of a restricted solid-on-solid model coupled to an Ising model, which can be derived from the coupled XY-Ising model. There are two kinds of phase transition lines. One is a Ising transition line and the other is surface roughening transition line. The latter is a KT transition line from the viewpoint of the XY model. Using a microcanonical Monte Carlo technique, we obtain a very accurate two dimensional phase diagram. The two transition lines are separate in all the parameter space we study. This result is strong evidence that the fully frustrated XY model orders by two separate transitions and that roughening and reconstruction transitions of crystal surfaces occur separately.

A preliminary study on discrimination of osteoporotic fractured group from nonfractured group using support vector machine

Osteoporosis is characterized by an abnormal loss of bone mineral content, which leads to a tendency to non-traumatic bone fractures or to structural deformations of bone. Thus, bone density has been considered as a most reliable parameter to assess osteoporotic fracture risk. In past decades, by the way, bone texture measures have been studied to estimate other aspect of bone quality. Some studies have been performed on CT or MR images to assess bone quality using trabecular structure analysis. Other studies have been performed on plain x-ray images or ultrasound images to assess trabecular structure. However, most of the studies are focused on individual parameters to distinguish between osteoporotic fractured group and nonfractured group. In this preliminary study, we combine various texture parameters with bone density parameters using a support vector machine and point out the most promising combination of parameters to distinguish between osteoporotic fractured group and nonfractured group

Distal Radius Bone Mineral Density Estimation using the Filling Factor of Trabecular Bone in the X-ray Image

Osteoporosis is characterized by an abnormal loss of bone mineral content, which leads to a tendency to non-traumatic bone fractures or to structural deformations of bone. Thus, bone density measurement has been considered as a most reliable method to assess bone fracture risk due to osteoporosis. In past decades, X-ray images have been studied in connection with the bone mineral density estimation. However, the estimated bone mineral density from the X-ray image can undergo a relatively large accuracy or precision error. The most relevant origin of the accuracy or precision error may be unstable X-ray image acquisition condition. Thus, we focus our attentions on finding a bone mineral density estimation method that is relatively insensitive to the X-ray image acquisition condition. In this paper, we develop a simple technique for distal radius bone mineral density estimation using the trabecular bone filling factor in the X-ray image and apply the technique to the wrist X-ray images of 20 women. Estimated bone mineral density shows a high linear correlation with a dual-energy X-ray absorptiometry (r=0.87)

Digital image subtraction of temporally sequential chest images by rib image elimination

We studied image subtraction of temporally sequential chest images by rib image elimination. The image subtraction presents clinically useful information related to interval changes in chest images. Clinical chest images are not usually reproducible in terms of patient positioning, X-ray projection, inspiration, cardiac pulsation, etc. A proper image registration is needed prior to the subtraction. An image registration method by nonlinear warping was suggested [A. Kano et al., 1994]. In the method, template regions of interest and the corresponding search area regions of interest are selected in the lung areas on the two images. The method provides an improper image registration for some pairs of temporally sequential images. One of the reasons for the improper image registration is the difference between the shifts of lung and rib by inspiration. Some pairs of temporally sequential chest images show large difference of lung position in the pairs in spite of small difference of rib position in the pairs. In this case, the method does not provide a proper image registration. To get the proper image registration of the temporally sequential images with different position changes for the rib and lung, we eliminated rib images from the chest images assuming that the contribution of rib to the chest image is uniform. The nonlinear warping was then applied to the rib eliminating images leading to the properly registered images and enhanced interval changes.

Simplified false-positive reduction in computer-aided detection scheme of clustered microcalcifications in digital breast tomosynthesis

A computer-aided detection (CADe) system for clustered microcalcifications (MCs) in reconstructed digital breast tomosynthesis (DBT) volumes was suggested. The system consisted of prescreening, MC detecting, clustering, and falsepositive reduction steps. In the prescreening stage, the MC-like objects were enhanced by a multiscale-based 3D calcification response function. A connected component segmentation method was used to detect cluster seed objects, which were considered as potential clustering centers of MCs. Starting with each cluster seed object as the initial cluster center, a cluster candidate was formed by including nearby MC candidates within a 3D neighborhood of the cluster seed object satisfying the clustering criteria during the clustering step. The size and number of the clustered MCs in a cluster seed candidate were used to reduce the number of FPs. A bounding cube for each MCC was generated for each accepted seed candidates. Then, the overlapping cubes were combined and examined according to the FP reduction criteria. After FP reduction step, we obtained the average number of FPs of 2.47 per DBT volume with sensitivity of 83.3%. Our study indicates the simplified false-positive reduction approach applied to the detection of clustered MCs in DBT is promising as an efficient CADe system.

Sequential algorithm for the detection of the shockable rhythms in electrocardiogram

We suggest a sequential algorithm for the detection of the ventricular fibrillation (VF) and ventricular tachycardia (VT) of a rate above 180 bpm, so called shockable rhythms. The built-in algorithm for ECG analysis embedded in the portable bio-signal sensing module is aimed to discriminate between shockable and non-shockable rhythms and its accuracy is analyzed. An algorithm for VF/VT detection is proposed to analyze every 1 s ECG episode using the past 8 s episodes. The method is tested with 844,587 ECG episodes from the widely accepted databases. A sensitivity of 86.8 % and a specificity of 99.4 % were obtained and compared with the previous results.

Development of a patch type embedded cardiac function monitoring system using dual microprocessor for arrhythmia detection in heart disease patient

A patch type embedded cardiac function monitoring system was developed to detect arrhythmias such as PVC (Premature Ventricular Contraction), pause, ventricular fibrillation, and tachy/bradycardia. The overall system is composed of a main module including a dual processor and a Bluetooth telecommunication module. The dual microprocessor strategy minimizes power consumption and size, and guarantees the resources of embedded software programs. The developed software was verified with standard DB, and showed good performance.