Methods based on Sobel edge detection offer more objective tools than thresholding methods for segmenting objects similar to lymph nodes in computed tomography images. Both methods, Sobel/watershed and interactive deformable contour algorithm, are fast and have simple user interfaces.

Forty-three patients with advanced, locally accessible neoplasms were treated in a Phase I clinical trial employing hyperthermia alone or hyperthermia combined with either high-dose-rate external beam or low-dose-rate interstitial radiotherapy (interstitial thermoradiotherapy). All patients had failed previous conventional therapeutic attempts, including various combinations of surgery, chemotherapy and radiation therapy. Many had received tolerance or near tolerance levels of prior radiation that restricted dose prescriptions in this trial to subcurative values. A number of tumors with different histologies were treated, including squamous cell carcinoma (14), adenocarcinoma (14), melanoma (8), malignant fibrous histiocytoma (2), and sarcoma (5). The response evaluation criteria used included no response (NR-less than 50% decrease in tumor volume), partial response (PR-50% I tumor volume reduction < 100%) and complete response (CR-complete tumor disappearance). For all tumor types, hyperthermia therapy alone resulted in a total response rate of 45% (27% PR, 18% CR). Hyperthermia combined with high-dose-rate external beam radiotherapy yielded a total response rate of 80% (53% PR, 27% CR). Seventeen patients treated with interstitial thermoradiotherapy displayed a 100% total response rate (29% PR, 71% CR). By tumor histologies for all treatment groups, total response rates have ranged from 50 to 79% for all types except melanoma, which has shown a 100% (8/8) response rate to date. Response durations have varied from one to 24 months. Twelve of the 43 patients remain alive; three have no evidence of disease (NED) while nine have either stable local disease or are NED in the treated volumes but have metastatic disease. Complications have been minimal and have included one third-degree burn and three second-degree burns from fringing R F fields, one vaginal-rectal fistula, a superficial focal soft tissue necrosis, and some minor blistering. The results of this Phase I trial demonstrate that hyperthermia alone or combined with radiation can be safely applied in the treatment of malignant disease. Most importantly, the data suggest that hyperthermia, especially when combined with interstitial thermoradiotherapy, can yield remarkable results in the eradication of local cancers. Cancer 49205-216,1982. H E RATIONALE for the use of hyperthermia in the T treatment of human malignancies is based on a spectrum of biologic and physical research, much of the evidence accumulating over the past ten years. Hyper

The introduction of local hyperthermia as a method of cancer therapy implies the necessity of quantitative measurements of the thermal dose. Our intention is to describe the nature of the problem, both physically and physiologically, with illustrations drawn from thermographic measurements in phantoms and in animals. The characteristics of a thermometry calibration facility are described. Some measurement problems associated with conventional thermometer probes are mentioned and several new thermometers which were developed for use in the electromagnetic fields are reviewed. We present some of the concepts that will guide the development of noninvasive thermometry. Systemic hyperthermia is not considered. We recommend that other reviews specifically directed toward localized hyperthermia be prepared on the methods of heating and on thermal physiological problems.

Measurements of the speed of ultrasound as a function of temperature in the 35 °C to 45 °C range were made on mammalian (canine) tissue both in vitro and in vivo at 5 MHz using a pulsed transmission technique. Cored samples coinciding with the ultrasonically irradiated tissue were obtained and assayed for water and total lipid content. Results indicate about 2 percent agreement between in vivo and in vitro data, and that, on the basis of kidney data, lipid content and temperature coefficient of ultrasound speed are more highly correlated (0.76 correlation coefficient) than are lipid content and speed of sound (0.17 correlation coefficient).

No abstract