Videotaped helical CT images for lung cancer screening

Risk Analysis for Pathological Changes in Pulmonary Parenchyma Based on Lung Computed Tomography Images

OBJECTIVE

The purpose of this study is to design a content-based medical image retrieval system, which helps excavate and assess pathological change of pulmonary parenchyma for risks analysis.

METHODS

A data set including lung computed tomography images obtained from 115 patients who experienced pathological changes in pulmonary parenchyma is used. Using morphological theory, images are preprocessed and decomposed into groups of pixel blocks (words), which construct vocabulary. A latent Dirichlet allocation (LDA) model is constructed to assess each image for risk analysis with the method of leave-one-out cross-validation. The precision and recall rate are used as the performance assessment criteria.

RESULTS

The LDA model generates a relevance rank of retrieval results from high to low. From the top 50 images, precision of identical tissue is 0.76 ± 0.031 and precision of each attribute of pulmonary parenchyma range from 0.776 ± 0.043 to 0.984 ± 0.008.

CONCLUSIONS

The study results demonstrate that the proposed LDA model is conductive to lung computed tomography image retrieval and has reliable efficacy on risk analysis about pathological changes of pulmonary parenchyma.

Computer-aided diagnosis of lung cancer: definition and detection of ground-glass opacity type of nodules by high-resolution computed tomography

PURPOSE

The ground-glass opacity (GGO) of lung cancer is identified only subjectively on computed tomography (CT) images as no quantitative characteristic has been defined for GGOs. We sought to define GGOs quantitatively and to differentiate between GGOs and solid-type lung cancers semiautomatically with a computer-aided diagnosis (CAD).

METHODS AND MATERIALS

High-resolution CT images of 100 pulmonary nodules (all peripheral lung cancers) were collected from our clinical records. Two radiologists traced the contours of nodules and distinguished GGOs from solid areas. The CT attenuation value of each area was measured. Differentiation between cancer types was assessed by a receiver-operating characteristic (ROC) analysis.

RESULTS

The mean CT attenuation of the GGO areas was -618.4 +/- 212.2 HU, whereas that of solid areas was -68.1 +/- 230.3 HU. CAD differentiated between solidand GGO-type lung cancers with a sensitivity of 86.0% and specificity of 96.5% when the threshold value was -370 HU. Four nodules of mixed GGOs were incorrectly classified as the solid type. CAD detected 96.3% of GGO areas when the threshold between GGO and solid areas was 194 HU.

CONCLUSION

Objective definition of GGO area by CT attenuation is feasible. This method is useful for semiautomatic differentiation between GGOs and solid types of lung cancer.

Computer-aided differentiation of malignant from benign solitary pulmonary nodules imaged by high-resolution CT

We investigated the possibility of using computer analysis of high-resolution CT images to radiologically classify the shape of pulmonary nodules. From a total of 107 HRCT images of solid, solitary pulmonary nodules with prior differentiation as benign (n=55) or malignant (n=52), we extracted the desired pulmonary nodules and calculated two quantitative parameters for characterizing nodules: circularity and second central moment. Using discriminant analysis for two thresholds in differentiating malignant from benign states resulted in a sensitivity of 76.9%, a specificity of 80%, a positive predictive value of 78.4%, and a negative predictive value of 78.6%.

Computer-aided diagnosis: a shape classification of pulmonary nodules imaged by high-resolution CT

We investigated the possibility of using computer analysis of high-resolution CT images to radiologically classify the shape of pulmonary nodules. Using a combination of circularity and second moment as quantitative measures we were able to classify pulmonary nodules in each shape group as effectively as could a radiologist. We found that pulmonary nodules with circularity < or =0.75 and second moment < or =0.18 were very likely to reveal lung cancer.

Small-Bowel Obstruction in a Phantom Model of ex Vivo Porcine Intestine: Comparison of PACS Stack and Tile Modes for CT Interpretation

PURPOSE

To compare computed tomographic (CT) image interpretation with picture archiving and communication systems (PACS) stack and tile modes for speed and accuracy of transition zone localization in small-bowel obstruction by using ex vivo porcine specimens.

MATERIALS AND METHODS

Twenty-five small-bowel obstruction phantom models made of ex vivo porcine intestines from a slaughterhouse were imaged at CT. One was used for observer training, and 24 were used for experimentation. At 20-cm intervals throughout the intestines, metallic markers were placed in the mesenteries immediately adjacent to bowel. One obstruction was made in each intestine, midway between markers, by ligating intestine with a 3-0 silk suture to simulate mechanical small-bowel obstruction. The lumen proximal to the ligation site was distended with air and a soybean oil-iodized oil mixture until at least two-thirds of the proximal intestine exceeded 2.0 cm in transverse diameter. Dilated segments were 310-550 cm in length. Soybean oil and a mixture of soybean and iodized oil were used to simulate differences in attenuation among bowel wall, intraluminal fluid, and extraluminal abdominal fat. Four experienced abdominal radiologists independently determined the transition zone by using stack mode (cine viewing of stacked images) and, at least 2 weeks later, tile mode (side-by-side image display). Accuracy and degree of error in counting markers were evaluated, and speed of interpretation was recorded. Statistical analysis was performed with the McNemar and Wilcoxon signed rank tests.

RESULTS

For all observers, accuracy of transition zone localization tended to be better with stack mode (63%-83% [15-20 phantoms]) than with tile mode (50%-63% [12-15 phantoms]), but the differences were not significant. For each observer, mean counting error was lower in stack mode (range, 0.96-2.48) than in tile mode (range, 1.74-3.22), with significance for three observers (P < .01, P < .01, and P = .04). Interpretation was significantly faster with stack mode by a factor of two to three for all observers (P < .01).

CONCLUSION

Stack mode evaluation for identification of the transition zone in obstructed small bowel is faster than evaluation with tile mode. Accuracy is not significantly different between modes, although there is a tendency toward better results with stack mode.

Solitary pulmonary nodules

The purpose of this study was to determine an optimal slice thickness that was efficient in differentiating malignant from benign solitary pulmonary nodules (SPNs) on high-resolution computed tomography (HRCT) images. For a total of 92 SPNs, four radiologist indicated their confidence level for the malignant or benign SPN on the CT images presented in 1-, 3-, and 5-mm slice thickness. HRCT could be used to differentiate more accurately the malignant nodules from the benign ones using 1-mm-thick sections than 3- or 5-mm-thick sections.

Metastases of colorectal carcinoma: comparison of soft- and hard-copy helical CT interpretation

PURPOSE

To compare soft- and hard-copy computed tomographic (CT) image interpretation with regard to evaluation time and detection rates for hepatic and extrahepatic colorectal metastases in candidates for liver surgery.

MATERIALS AND METHODS

In 20 patients with a history of colorectal carcinoma, two radiologists independently evaluated CT data sets. Focal hepatic lesions were characterized as benign or malignant by using a five-point scale. In each patient, soft-copy readouts and hard-copy printouts were compared for nonenhanced hepatic, contrast material-enhanced hepatic, and contrast-enhanced extrahepatic data sets. A stopwatch was used to document evaluation time. Ninety-two hepatic metastases and six extrahepatic metastatic recurrences were detected with the standard of reference--surgical, intraoperative ultrasonographic, and histologic findings.

RESULTS

Both observers evaluated the contrast-enhanced hepatic data set significantly faster (P =.026 and.009) by using soft-copy readouts. The contrast-enhanced extrahepatic data set was also evaluated significantly faster (P =.010 and.006) with soft-copy readouts. Detection of hepatic and extrahepatic tumor with soft-copy readouts is not significantly superior to that with hard copies. Detection rates of hepatic metastases for nonenhanced and contrast-enhanced CT for both observers ranged from 50%-80% (46-74 of 92) for soft-copy readouts and 46%-75% (42-69 of 92) for hard copies. Interobserver agreement was highest for contrast-enhanced soft-copy readouts for hepatic metastases.

CONCLUSION

Soft-copy readouts of contrast-enhanced CT data sets for the detection of hepatic metastases and extrahepatic metastatic recurrences were evaluated significantly faster than were hard copies, with at least equal sensitivity and with excellent interobserver agreement.