Computer aided diagnosis of parotid gland lesions using ultrasonic multi-feature tissue characterization.

In this article, an ultrasound based system for computer aided characterization of biologic tissue and its application to differential diagnosis of parotid gland lesions is proposed. Aiming at an automated differentiation between malignant and benign cases, the system is based on a supervised classification using tissue-describing features derived from ultrasound radio-frequency (RF) echo signals and image data. Standard diagnostic ultrasound equipment was employed to acquire ultrasound RF echo data from parotid glands of 138 patients. Lesions were manually demarcated as regions-of-interest (ROIs) in the B-mode images. Spectral ultrasound backscatter and attenuation parameters are estimated from diffraction corrected RF data, yielding spatially resolved parameter images. Histogram based statistical measures derived from the parameters distributions inside the ROI are used as tissue describing features. In addition, texture features and shape descriptors are extracted from demodulated ultrasound image data. The features are processed by a maximum likelihood classifier. An optimal set of 10 features was chosen by a sequential forward selection algorithm. The classifier's performance is evaluated using total cross validation and receiver operating characteristic (ROC) analysis. As a reference method, postoperative pathohistologic analysis was conducted and proved malignancy or prospective malignancy in 51 patients. The classification using the proposed system yielded an area under the ROC curve of 0.91, proving significant potential for differentiating between malignant and benign parotid gland lesions.

Investigation of the influence of blood flow rate on large vessel cooling in hepatic radiofrequency ablation.

Radiofrequency (RF) ablation using high-frequency current has become an important treatment method for patients with non-resectable liver tumors. Tumor recurrence is associated with tissue cooling in the proximity of large blood vessels. This study investigated the influence of blood flow rate on tissue temperature and lesion size during monopolar RF ablation at a distance of 10 mm from single 4- and 6-mm vessels using two different approaches: 1) an ex vivo blood perfusion circuit including an artificial vessel inserted into porcine liver tissue was developed; and 2) a finite element method (FEM) model was created using a novel simplified modeling technique for large blood vessels. Blood temperatures at the inflow/outflow of the vessel and tissue temperatures at 10 and 20 mm from the electrode tip were measured in the ex vivo set-up. Tissue temperature, blood temperature and lesion size were analyzed under physiological, increased and reduced blood-flow conditions. The results show that changes in blood flow rate in large vessels do not significantly affect tissue temperature and lesion size far away from the vessel. Monopolar ablation could not produce lesions surrounding the vessel due to the strong heat-sink effect. Simulated tissue temperatures correlated well with ex vivo measurements, supporting the FEM model.

Sonohistology for the computerized differentiation of parotid gland tumors.

A system for the computerized differentiation of parotid gland tumors is proposed. The parotid gland is the largest of the salivary glands. It is found in the subcutaneous tissue of the face, overlying the mandibular ramus and anterior and inferior to the external ear. The classification system is based on a multifeature tissue characterization approach, using fuzzy inference systems as higher-order classifiers. Baseband ultrasonic echo data were acquired during conventional ultrasound imaging examinations using standard ultrasound equipment. Several tissue-describing parameters were calculated within numerous small regions of interest to evaluate spectral and textural tissue properties. The parameters were processed by an adaptive network-based fuzzy inference system, using the results of conventional histology after parotidectomy as the "gold standard." The results of the classification are presented as a numerical score indicating the probability of a certain tumor or alteration for each parotid gland. The score can be presented to the physician during examination of the patient to improve the differentiation between various types of parotid gland tumors. The system was evaluated on n = 23 cases of patients undergoing radical parotidectomy. The receiver operating characteristic curve area is A(ROC) = 0.95 +/- 0.07 when using fourfold cross-validation over cases and differentiating between various benign parotid gland tumors and monomorphic adenoma.

Ultrasound elastography for the age determination of venous thrombi. Evaluation in an animal model of venous thrombosis.

The exact age determination of venous thrombi is important if thrombolytic therapy or surgical thrombectomy is considered. Clinical symptoms as well as duplex-ultrasound and phlebography are unreliable in this respect and do not allow an exact age estimation. Ultrasound elastography can provide information about the elastic properties of thrombi. Since thrombus elasticity decreases with age due to the organisation process, it should be possible to use elastography to stage the degree of organisation and thereby determine the age of venous thrombi. Experimental venous thrombi aging 1, 3, 6, 9, 12 and 15 days were created in a porcine model by laparoscopic ligation of the infrarenal Vena cava in combination with transfemoral infusion of thrombin. The thrombosed iliac veins were explanted and embedded in gelatine, after that they underwent examination by ultrasound elastography. In addition, histological evaluation of the thrombi was performed. Elastography demonstrated a decline in thrombus elasticity between days 6 and 12 with the 12-day-old thrombi being about 3 times harder then the 6-day-old thrombi. This correlated with the histological findings, which demonstrated a marked increase in fibroblast and collagen production in the clots during this time, with the 12- and 15-day thrombi showing signs of advanced organisation. In conclusion, in an experimental setting, ultrasound elastography was helpful in determining the exact age of venous thrombi. The differences in elasticity were most pronounced between days 6 and 12, which is also the most relevant time frame when considering invasive therapies in human venous thrombosis.

A tutorial on the use of ROC analysis for computer-aided diagnostic systems.

The application of the receiver operating characteristic (ROC) curve for computer-aided diagnostic systems is reviewed. A statistical framework is presented and different methods of evaluating the classification performance of computer-aided diagnostic systems, and, in particular, systems for ultrasonic tissue characterization, are derived. Most classifiers that are used today are dependent on a separation threshold, which can be chosen freely in many cases. The separation threshold separates the range of output values of the classification system into different target groups, thus conducting the actual classification process. In the first part of this paper, threshold specific performance measures, e.g., sensitivity and specificity, are presented. In the second part, a threshold-independent performance measure, the area under the ROC curve, is reviewed. Only the use of separation threshold-independent performance measures provides classification results that are overall representative for computer-aided diagnostic systems. The following text was motivated by the lack of a complete and definite discussion of the underlying subject in available textbooks, references and publications. Most manuscripts published so far address the theme of performance evaluation using ROC analysis in a manner too general to be practical for everyday use in the development of computer-aided diagnostic systems. Nowadays, the user of computer-aided diagnostic systems typically handles huge amounts of numerical data, not always distributed normally. Many assumptions made in more or less theoretical works on ROC analysis are no longer valid for real-life data. The paper aims at closing the gap between theoretical works and real-life data. The review provides the interested scientist with information needed to conduct ROC analysis and to integrate algorithms performing ROC analysis into classification systems while understanding the basic principles of classification.