Takeshi Iinuma

Image processing for computer‐aided diagnosis of lung cancer by CT (LSCT)

This paper reports a special-purpose CT for early detection of lung cancer and presents the basic idea. The parallel use of the following two systems is investigated as the diagnosis-support image display method needed in such a system, and a satisfactory result is obtained. (1) The maximum intensity projection (MIP) is applied as a means to realize the two-dimensional projection display from the three-dimensional information composed of 40 slices of lung tissue while retaining the information concerning the pathological shadows. When the simple MIP is applied, there is a large disturbance due to unnecessary organ signals. Consequently, a method is developed in which unnecessary information is deleted by thresholding; (2) The automatic recognition of candidates for the pathological shadows is applied to each slice. Only the CT image containing a shadow candidate is displayed on CRT, which helps to reduce greatly the number of cross sections to be displayed. In the automatic recognition of the pathological shadow, the quoit filter recently developed by the authors is employed.

Quoit filter-a new filter based on mathematical morphology to extract the isolated shadow, and its application to automatic detection of lung cancer in X-ray CT

We propose a new algorithm named Quoit filter (Q-filter) to extract the isolated but low amplitude shadow located in the background which has extremely high amplitude fluctuation. Q-filter is a kind of mathematical morphology and its formulation is quite simple. This simplicity brings about a unique merit that output from this filter is analytically expressive for the case of analytical input shapes like ball, cone, or rotation of cosine function, which have characteristics of rotation symmetry and monotonic decreasing from the origin. This Q-filter is composed of two sequential operations named Q Trans. and Q Inv. Trans., Q Trans. corresponds to extracting feature parameters like a matched filter from the input image having a nonideal isolated shadow, and Q Inv. Trans. corresponds to restoring isolated images using extracted feature parameters. This filter is applied to detecting the cancer candidate shadow automatically in the CT cross sections of lung areas, aiming to reduce drastically the number of cross sections to be diagnosed by the doctor.

Proposed System for Ultrafast Computed Tomography

Construction of a new ultrafast scanner system suitable for dynamic computed tomography is now under consideration, with development soon to begin. This system will be characterized by an electronically controlled X-ray generator with wide angle double diffraction, which will permit transverse axial images to be taken in times as short as 0.01 second.

Positologica: A Positron ECT Device with a Continuously Rotating Detector Ring

A rotary positron emission computed tomography device is developed for human brain and animal studies. The device utilizes 64 rectangular BGO detectors arranged at unequal spacing on a circular ring. The detector ring is continuously rotated at a constant speed of 60 rpm or less. This single continuous motion of the ring with the detector array provides excellent sampling characteristics and high detector redundancy. The device has a field of view 24 cm in diameter with a slice thickness adjustable from 1 cm to 2 cm. Measured width of system response to a 2 mm diameter line source is 5.8 mm FWHM at the center and less than 9 mm FWHM within a circle 16 cm in diameter. Measured sensitivity including scattered coincidence events is about 17 kcps/pCi/ml for a 20 cm diameter water phantom and 2 cm thick slice.

A detection method of ground glass opacities in chest x-ray CT images using automatic clustering techniques

In this paper, we described an algorithm of automatic detection of Ground Glass Opacities (GGO) from X-ray CT images. In this algorithm, first, suspicious shadows are extracted by our Variable N-Quoit (VNQ) filter which is a type of Mathematical Morphology filters. This filter can detect abnormal shadows with high sensitivity. Next, the suspicious shadows are classified into a certain number of classes using feature values calculated from the suspicious shadows. In our traditional clustering method, a medical doctor has to manually classify the suspicious shadows into 5 clusters. The manual classification is very hard for the doctor. Thus, in this paper, we propose a new automatic clustering method which is based on a Principal Component (PC) theory. In this method, first, the detected shadows are classified into two sub-clusters according to their sizes. And then, each sub-cluster is further classified into two sub-sub-clusters according to PC Scores(PCS) calcuated from the feature values of the shadows in the sub-cluster. In this PCS-based classification, we use a threshold which maximizes the distance between the two sub-sub-clusters. The PCS-based classification is iterated recursively. Using discriminate functions based on Mahalanobis distance, the suspicious shadows are determined to be normal or abnormal. This method was examined by many samples (including GGOs shadows) of chest CT images, and proved to be very effective.

Analysis of various malignant neoplasms detected by FDG-PET cancer screening program: based on a Japanese Nationwide Survey.

ObjectiveThe most distinctive feature of FDG-PET cancer screening program is the ability to find various kinds of malignant neoplasms in a single test. The aim of this survey is to clarify the range and frequency of various malignant neoplasms detected by FDG-PET cancer screening performed in Japan.Methods“FDG-PET cancer screening” was defined as FDG-PET or positron emission tomography and computed tomography (PET/CT) scan with or without other tests performed for cancer screening of healthy subjects. This survey was based on a questionnaire regarding FDG-PET cancer screening. We analyzed the situation of 9 less frequently found malignant neoplasms including malignant lymphoma, malignancy of head and neck, esophagus, hepatobiliary and gallbladder, pancreas, kidney, cervical and uterine, ovary, and bladder.ResultsThe detailed information of subjects with the suspected 9 kinds of malignant neoplasms mentioned above in the FDG-PET cancer screening program was studied in a total of 1,219 cases from 212 facilities. A statistical significance between PET/CT and PET was found in relative sensitivity and PPV for renal cell cancer. Malignant lymphoma was frequently of indolent type, suspected head and neck cancers had many false-positive results, and pancreatic cancer detected in this program was often in the advanced stage even in asymptomatic subjects. The recommendation of combined screening modality to PET or PET/CT was as follows: gastric endoscopy for assessing early esophageal cancer; abdominal ultrasound for screening hepatobiliary and gallbladder cancer; pelvic magnetic resonance imaging for assessing gynecological and pelvic cancers; and the CA125 blood test for screening ovarian cancer. Delayed image was helpful depending on the type of suspected malignant neoplasm.ConclusionWe analyzed various types of malignant neoplasms detected by the FDG-PET cancer screening program and presented recommended combination of examinations to cover FDG-PET and PET/CT.

Correction functions and statistical noises in transverse section picture reconstruction

Abstract An analysis of image reconstruction with one- and two-dimensional convolution methods is presented in an outline of the relationship among the correction functions, the point spread function and the statistical noise. The correction functions are derived which maximize the ratio of one-dimensional signal power to noise power for a given r.m.s. resolution width in a uniform image. The texture of the image noise is expressed by the autocovariance function. For emission images, the variance of noise is expressed by the convolution of the source distribution with the “error kernel” which is determined from the one-dimensional correction function.

Uptake and retention experiments of radioruthenium in man. I.

As a link in the pre-operational research program for the establishment of the first fuel reprocessing plant in Japan, the metabolism in man ofruthenium was studied in experiments involving the oral administration of metabolized tracer in shellfish. A whole body counter-scanner could trace in vivo the movement of ingested tracer from stomach to rectum through the large intestine within 24 hr. Whole body retention was best fitted by a three-component exponential function with time, the fastest component being caused by a large fecal excretion which amounted to about 95% ofadministered dose within 2 days. The second component could be attributable to the prolonged retention of radioruthenium inside the gastrointestinal tract, the biological half-life being 2.3 days. The slowest component, about 1 % of the administered dose, retained in the body with a half-life of 30 days, and this fraction was considered to be absorbed from the gastrointestinal tract. Successive experiments with non-metabolized tracer in the chemical forms of the chloro complexes of nitrosylruthenium(II1) and ruthenium(II1 and 1V)resulted in ahighergastrointestinal absorption of the former by a factor of about 3 than the latter. I t is suggested that changes in the chemical and physical states of ruthenium, of pH and of different periods of time following preparation until administration, as well as changes in residence time in the gastrointestinal tract may cause variability in absorption.

Measurement of Arterial Tracer Concentrations from Cardiac Pet Images

Objective Cardiac PET imaging offers the potential capability for the noninvasive measurement of arterial input function. This capability is somewhat limited, however, because of partial volume and spillover effects. The purpose of this study was to validate arterial tracer concentration measurements using PET images by assigning a region of interest (ROI) to the left ventricle, left atrium (LA), and ascending aorta. Materials and Methods We selected [11C]CO and [13N]ammonia as tracers, because the activity of [11C]CO is primarily in the blood pool and that of [13N]amtnonia primarily in myocardial tissue. Six [11C]CO and 11 [13N]ammonia PET scans were obtained in 17 subjects. Arterial tracer concentrations determined by PET were compared with those measured by well counting of arterial blood sampling obtained at the midpoint of each image acquisition. Results Arterial tracer concentrations as determined by LA-PET imaging correlated closely with those measured by blood sampling (r = 0.996, slope = 1.00 for [11C]CO; r = 0.841, slope = 0.974 for [13N]ammonia). Partial volume and spillover effects were minimized by assigning an ROI to the LA. Conclusion Arterial tracer concentrations can be measured noninvasively with cardiac PET. Index Terms Emission computed tomography—Heart—Tracers—Emission computed tomography, techniques.

Recognition of lung nodules from x-ray CT images using 3D Markov random field models

In this paper we propose a new recognition method of lung nodules from x-ray CT images using 3D Markov random field (MRF) models. Pathological shadow candidates are detected by our Quoit filter which is a kind of mathematical morphology filter, and volume of interest (VOI) areas which include the shadow candidates are extracted. The probabilities of the hypotheses that the VOI areas come from nodules (which are candidates of cancers) and blood vessels are calculated using nodule and blood vessel models evaluating the relations between these object models using 3D MRF models. If the probabilities for the nodule models are higher, the shadow candidates are determined to be abnormal. Otherwise, they are determined to be normal. Experimental results for 38 samples (patients) are shown.